Framerates Matter

An anonymous reader writes "As more and more games move away from 60fps, the myth of the human eye only being able to detect 30fps keeps popping up. What's more, most people don't seem to realize the numerous advantages of a high framerate, and there's plenty of those."

Comments

Motion blur and bloom effects

[sopssa]

The article notes about motion blurring, and links to NVidia's page about it's technology. The last figure shows a terrain with full-screen motion blur effect, which in my opinion is pretty important in games to create that feeling of speed. People usually object against this and bloom effects and just want a sharp picture, but maybe some games have taken it too far. It's important none the less, even if it's not all sharp picture, because your eye picture isn't all that sharp either and you experience the same blur.

Re:Motion blur and bloom effects

[Locke2005]

Doom always appeared to draw textures with much lower resolution while you were moving, and only display the full texture when you stopped and looked directly at an object, as a way of speeding up rendering. This gave the appearance of "motion blur" without a lot of additional processing required.

Re:Motion blur and bloom effects

[Hatta]

It's important none the less, even if it's not all sharp picture, because your eye picture isn't all that sharp either and you experience the same blur.

If my eye creates the blur, why do I need artificial motion blur?

Re:Motion blur and bloom effects

[Anonymous Coward]

Because you aren't moving past the computer screen as fast as the scenery in the game is.

Re:Motion blur and bloom effects

[sopssa]

Because you're still looking at a single object, your monitor, and the picture and movement in it is artificially created. If you look at real objects moving or move or shake your head you'll notice theres huge motion blur effect. If you do it in game that has no motion blur effect, you notice how it instantly jumps to where you want to look at.

Re:Motion blur and bloom effects

[spun]

Just a guess, but perhaps because the frame rate isn't high enough for your eye to generate the blur? That is to say, if the scene were real, the frame rate would be well-nigh infinite, and your eye, capable of only a certain frame rate, would blur together all the frames. With discrete frames, you need to put in the blur the eye would generate from the frames in-between.

Or something like that.

Re:Motion blur and bloom effects

[Shin-LaC]

Your eyes introduce blur due to the reaction time of the light-sensitive cells in the retina. Fortunately, the image processing area in your brain treats blur introduced by the eyes and blur built into the frame more or less the same, so you can use blur to give the impression of smooth motion with a lower frame rate than would otherwise be necessary. This is used to good effect in cinema, where the camera's exposure time naturally introduces blur that is quite similar to the one introduced by your eye.

In the case of video games, however, it is not so clear that rendering effctive artificial motion blur saves much processing time compared to simply rendering more frames. Then again, there is a limit to how fast your monitor can update its image, so rendering more frames is no longer an option past that point.

Re:Motion blur and bloom effects

[Shadow of Eternity]

That just means we should strive for a higher framerate until our eyes blur things on their own. Reality is not inherently blurry (unless you need glasses...), our eyes and brain do that internally.

Making movement in a game inherently blurry when your head is already going to blur it for you internally is just a shortcut to motion sickness for a whole lot of people.

Re:Motion blur and bloom effects

[Monkeedude1212]

Essentially people want these effects to be done by their eyes though, not the game. Why can't the game/computer/monitor produce fast enough frame-rates that its my eyes that are creating the blur, not the Post Rendering effects?

Don't get me wrong, I like the realism that these effects give, but some people see them as kind of fake and it draws away from their experience. Perhaps some people's eyes can percieve frame-rates slightly faster than others and thus don't actually see as much blur when moving fast as other people do.

Re:Motion blur and bloom effects

[LordKazan]

Why can't the game/computer/monitor produce fast enough frame-rates that its my eyes that are creating the blur, not the Post Rendering effects?

Physics.. monitors cannot change fast enough and in the right way to do this. they simply don't work that way.

Speaking of Physics - the properties of a game's physics engine have the properties of a Riemann sum where n=fps. so the higher your FPS the more accurate your physics simulation, even if your monitor cannot discretely display all those frames.

[note: only applies in games where physics ticks/sec are tied to framerate... which is almost all games]

Re:Motion blur and bloom effects

[poetmatt]

what are you talking about?

ever waved your hand so fast back and forth that it creates a blur? Ever seen those little things that go back and forth back enough to display an image?

Reality is indeed inherently blurry. It's just hard to accurately portray blur when you're staring at something that's not moving.

Re:Motion blur and bloom effects

[Chris Pimlott]

Perhaps you're thinking of mipmapping, which was implemented at least as early as Quake 1.

Re:Motion blur and bloom effects

[Anonymous Coward]

The human eye does not work with frames. It is a continuous, dynamic system.

The only way to achieve the blurring effect without artificially rendering it would be to make the object move on the screen at the same velocity it would in normal life. Needless to say, making a car move at 300 km/h in the virtual world and showing it with the real speed on the screen would simply make the game unplayable even excluding the practical difficulty of driving at that speed, because today screens simply don't have the necessary response times.

Re:Motion blur and bloom effects

[sopssa]

I dont think we will get to a point that the framerate would be fast enough. The 3D monitors only generate up to 120fps too, and there's still lots of hardware limits to generate framerates over that with current games on good resolutions. And there is no framerate in real world; you're taking in images in realtime. Some argue (like the battle between 30fps vs 60fps) that human eye can't process more than certain amount of "frames" per second. The natural motion blurring effect and it's absence with video games perfectly shows that it can. While you see a smooth movement, you're still missing extra things like that generated by brain.

Re:Motion blur and bloom effects

[sopssa]

Also bloom and lighting effects you still have to do in game because they rely on game world, can hide objects behind that bloom or make other objects dark, and because monitor just shows the color data you give it.

Re:Motion blur and bloom effects

[i_ate_god]

what are you talking about?

ever waved your hand so fast back and forth that it creates a blur? Ever seen those little things that go back and forth back enough to display an image?

Reality is indeed inherently blurry. It's just hard to accurately portray blur when you're staring at something that's not moving.

That's your brain doing that, I'm sure there are creatures who will see the hand clearly as it moves.

Re:Motion blur and bloom effects

[DavidTC]

More to the point, the eye does not work with frames. The eye itself has no framerate.

Rods and cones individually update at about 15 times a second, but each individual one is entirely asynchronous from all the others. One update, another update, another update, etc. Your entire eye is not read 15 times a second, each individual light sensor 'trips' 15 times a second, semi-randomly, and sends the current light level. (1)

While each rod and cone only sends one signal, and then nothing, until it resets and sends another, our brains seems to assume that the light and color levels have remained the same.

Hence we get a 'blur', as objects move, and our brain assumes that said object is also in the old position until all rods and cones have updated.

1) And even that's not entirely right. Each rod and cone is actually sending a sorta average of the light it received since in the last update. You don't have to receive a photon exactly as it updates.

Re:Motion blur and bloom effects

[Moryath]

Not quite.

The eye blur happens for two reasons. The first is the fact that the human eye is "assembling" an analog reading of the light taken over a specific time, very similar to how a camera exposure works. We aren't "digital" beings, in the sense that there is allowance forward and back in our visual processing, but we DO assemble "frames" for the rest of our brain to analyze.

The second is focusing. A fast-moving object moves into, and out of, the focused field quite quickly. Either we keep tracking it (in which case the unfocused foreground and background areas alter) or we don't, and it goes out of focus. We mentally render this as blurring. Directors in 2D movies use depth-of-field to do a quick transition between two speaking characters and ensure the right one has prominence, by keeping the speaker in focus and then quickly shifting focus in/out to bring the other to prominence when the dialogue turns.

The real sin, and unalterable problem currently, for 3D technology is that everything renders in-focus. Motion blurs work to some degree, but a large-scale image with "background" objects sharply in focus gives us headaches. We follow the other visual cues, try to "focus" to distance, try to "refocus" for the fuzziness it causes, and then wobble back and forth till we have sore, tired eye muscles.

The 3D Brendan Frasier Journey to the Center of the Earth was the closest done so far, because they did introduce some background blur, but it still had problems should the viewer decide to focus on something other than what the director wanted them to focus on, visually. Avatar commits the same sin as well, and doesn't even try to do it properly. It's like watching some big pixely, perfect-focus-for-miles video game.

As for the other items they mention - "The framerate of a game is usually directly tied to the processing of its logic." Not true. Indeed, only true if you've got shoddy programmers (the fix for one of the most notorious examples, the jumping-height differences of various iterations of the Quake engine, was to simply lock the calculations to assume a static framerate; the id software programmers, who chose to instead discard "erroneous" round-up errors, wound up widely criticized for STILL making the jumps somewhat randomly framerate-dependent. The truth is that the visual rendering framerate of a game simply does not have to be the same as the internal calculation "frame" rate.

As for input lag... the difference in "lag" between a 30-fps framerate and a 60-fps framerate is 16 ms. Even if you get to 120-fps and have a monitor capable of doing so at your chosen resolution, your difference is 25 ms. Human reaction to visual stimuli is generally in the neighborhood of 150-300 ms.

Even playing on a LAN in the same building, you're looking at random lag times longer than the difference between 120fps and 30fps.

Re:Motion blur and bloom effects

[Monkeedude1212]

Well yes, its become an art to program things so that they appear no different than reality, at least to the user, no matter what is actually happening behind the scenes. I personally would prefer it if computers were stong enough to calculate a photon hitting a material, reflecting its non-absorbed light into a "camera" object in game and taking the rendered picture and sending it to the monitor, thus creating a more realistic lighting effect, but we just aren't there yet.

Re:Motion blur and bloom effects

[spun]

Then how is it, in games like Gran Turismo with real world cars and tracks, that actual race car drivers turn in near-identical times on virtual and real tracks with the same car?

A car traveling 300 Km/h travels 300,000 m/h, or 83.3 m/s. At 30 frames per second, that's around 3 meters per frame. Are you saying this isn't possible to calculate?

Re:Motion blur and bloom effects

[karnal]

Even playing on a LAN in the same building, you're looking at random lag times longer than the difference between 120fps and 30fps.

If it's designed right, you're not. Network stack and physical "move the bits across this copper" lag on a proper LAN should be 1ms. Ping testing can sometimes get sketchy - routers can slow down a ping if you address a ping packet to them and their cpu is a little loaded. Packets going through them though - even pings - will get as high a priority as the traffic they're classed in.

Re:Motion blur and bloom effects

[Nicolay77]

OK reality is blurry. I see a fly in the air and it has blurry wings.

However, if reality is blurry, why can a high speed camera capture the wings moving without any blur at all?

It seems that experimental evidence totally contradicts your statement.

Re:Motion blur and bloom effects

[takev]

But, if you follow the hand with your eyes, your hand will appear sharp. You'll be supprised how quickly and stable eyes can track moving objects.

The BBC has been experimenting with fast frame rate TV, running at 300 frames-per-second. Moving objects will appear much sharper with such a broadcast compared to the standard 50 frames-per-second (not fields). They showed a side by side example, both were 1080 progressive scan. Great for sports broadcasting.

Also Silicon Graphics (when they were called that) have done test with fighter pilots when designing flight simulators. Motion sickness is a problem with those flight simulators, compared to an actual jet plane. When they got a constant frame rate above 80 frames (160 frames per second when doing stereo imaging) per second the motion sickness was greatly reduced. They solved the processing power problem by being able to reduce the rendering resolution on each frame.

Re:Motion blur and bloom effects

[Anonymous Coward]

Reality is indeed inherently blurry. It's just hard to accurately portray blur when you're staring at something that's not moving.

OMG, did you really just suggest that objects in the real world physically blur themselves when they move fast??

This has got to be the stupidest claim I've seen on Slashdot in weeks, and that's quite an achievement!

Re:Motion blur and bloom effects

[Shadow of Eternity]

No, I'm not. If someone shoots past me at Mach Fuck in Tribes I don't see a perfectly sharp player slide across my screen, I see a slightly blurred SOMETHING shoot past because it was displayed moving at a speed I can't see clearly.

An image on a screen moving and a real object moving will still blur for you if they're moving fast enough because as far as your body cares it's pretty much the same input.

Re:Motion blur and bloom effects

[Twinbee]

[quote]Physics.. monitors cannot change fast enough and in the right way to do this. they simply don't work that way.[/quote]
OLED TV will change that. LG have already brought out a 15" TV finally in Korea, so the tech is taking off...

Re:Motion blur and bloom effects

[Blakey Rat]

Doom? No it didn't. It was extremely basic, super simple texture-mapping with absolutely no dynamic effects like that. (I'd say Marathon-like, but I guess more people have played Doom than Marathon. ;)

I do know what you're talking about-- some original Xbox games have done that in recent years, although I can't think of a specific one off the top of my head. (Maybe Chronicles of Riddick?)

A decent amount of Xbox 360 games also have some form of delayed-texture-loading which might be what you're seeing-- the level will initially load relatively crude textures, then replace them with the detailed textures once they load. (See Borderlands for an example. Or the Halo games.) But once the detailed texture is in the scene, it stays there until the level changes... not the same as the effect you're describing.

Re:Motion blur and bloom effects

[smitty97]

Because the framerate is high.

There, i've taken it full circle.

Re:Motion blur and bloom effects

[Anonymous Coward]

Blur occurs when the light reflected at a given instant is low. Much like a camera in low light. Notice that those LED tail lights leave lots of individual dots in your vision when you move your head while watching them... they register as individual points of light, rather than a blur due to their high intensity and short duty cycle. The case of a monitor displayed image is similar to the LED tail light... its a high intensity image registering on the retina... the residual image will be crisp even as the next one draws... it fails to create a realistic motion blur at almost any framerate -- thus the motionblur needs to be simulated in software... blurring the actual drawn image.

Re:Motion blur and bloom effects

[Ohrion]

Lol, I was thinking the same thing. The guy totally misunderstood the argument, and made an asinine statement as a rebuttal.

Re:Motion blur and bloom effects

[Anonymous Coward]

It is of course completely possible to calculate, and the games actually do it.

So, your calculations are correct. But they don't take into account the fact that the scene viewed on the screen is not to scale with reality. Therefore, the real speed on the monitor is less, and the blur effect is therefore less pronounced, hence the need for artificial blurring.

Re:Motion blur and bloom effects

[spun]

Sorry, but to me, that doesn't make nearly as much sense as some of the other explanations given here.

Re:Motion blur and bloom effects

[Anonymous Coward]

Wrong. It is not the same at all.

Imagine each frame displayed is a picture of the world. While it is taking the picture, the world stands still. Contrast that to a camera or an eye. Light is allowed to hit the film for X milliseconds. If the light is changing as an object moves, the film/retina will blur because different intensities/colors were hitting various parts of the photosensitive material. You get a smooth blur.

Now here is where the problem comes in. Computers just can not generate these still frames fast enough and update the geometry to get the same smooth motion. The faster the object on the screen is moving, the more you miss between frames. I show a frame where the character is at position 0,0 and then the next he is at 400,0 you will not see blur, you will see a jump. That is why we apply motion blurring. Saying "just raise the framerate" is like saying not to anti-alias because if the resolution was high enough we wouldn't need it.

Re:Motion blur and bloom effects

[Hurricane78]

The difference is, that your eyes have a freedom to focus anywhere on the screen. Which means that the place you look at has to be free of blur, just as in nature. But that in-game blur usually forces you to not move your eyes, which is incredibly annoying. Also it pointlessly takes a good load of GPU power. It’s always the first thing I disable in games.

Re:Motion blur and bloom effects

[merreborn]

The point TFA was making, was 24 FPS film works, is because 24 FPS film is, as you call it "inherently blurry".

Therefor, for 24 FPS games to work equally well, they, too, must be "inherently blurry"

Re:Motion blur and bloom effects

[beelsebob]

The really silly thing about this article is that it's Just Plain Wrong(TM). 1 and 2 on it's list of reasons for wanting a high frame rate are "games tie the physics engine refresh rate to the screen refresh rate", which is simply *not true*. Any game that did this in fact would be utterly unplayable. Most games run their physics engines at *many* times more iterations per second than the renderer is running at.

The point about consistency is to some extent true – except that it again falls for this falacy that games are tied to a certain simulation rate, and a certain frame rate. A game is not a "30 fps game" -- it's a game that averages 30fps no that particular hardware. Which pretty much negates his point about "most of the time it'll be running at 20-30fps"... no it won't it's averaging 30fps.

The point about motion blur is rubbish as well, he's right - humans *do* use motion blur to see motion accurately... But that doesn't immediately lead to "so the frame rate has to be higher", it leads to "so the engine should try to do motion blurring accurately.

In short, four reasons – all bollocks.

Re:Motion blur and bloom effects

[brianosaurus]

I agree with you...

Some argue (like the battle between 30fps vs 60fps) that human eye can't process more than certain amount of "frames" per second.

Isn't the reason movies use 24 fps (and similarly TV uses ~30fps) because of historical technical limitations? That is right about the minimum rate where your eyes and brain can smooth out the annoying flicker. 30fps isn't the upper limit that the eye can process, but rather a lower limit that makes the image sequence appear as motion without causing stutter, headaches, or otherwise detract from the visual experience. Its a compromise to allow movies to fit on reasonable sized rolls of film, and for TV to have been able to fit "good enough" video quality into the available bandwidth at the time, and to not have frequency beating artifacts due to lights running on 60Hz AC power (or 50Hz & 25fps in Europe, etc).

For an easy example that 30fps isn't enough, run iTunes, play some music and turn on the "iTunes Classic Visualizer" full screen. Hit "F" to display the frame rate, then use "T" to toggle the 30fps limit on and off. Tell me you don't see a big difference.

I'm sure there's an upper threshold where you can't distinguish a difference as frame rate increases, but its much higher than 30 or 60 fps, and as the parent said it is probably higher than we can achieve in hardware for the near future.

Re:Motion blur and bloom effects

[Anonymous Coward]

More to the point, the eye does not work with frames. The eye itself has no framerate.

Rods and cones individually update at about 15 times a second, but each individual one is entirely asynchronous from all the others. One update, another update, another update, etc. Your entire eye is not read 15 times a second, each individual light sensor 'trips' 15 times a second, semi-randomly, and sends the current light level.

This description still sounds like "frames" in the sense of the rod having an output value that stays fixed until some discrete point, at which it takes on another output value. Cells are analog, and the notion of "15 times per second" comes from a time constant around 60 ms, which seems appropriate for the time it takes a rod action potential to propagate through a few layers of neurons into the layer someone has designated as "conscious awareness". The real light processing happens in the rod cell itself, independent of action potentials. It's mostly a competition between a photon-driven phosphodiesterase and a calcium driven guanylate cyclase. That competition can vary the charge inside the rod continuously. The charge determines the rod's firing rate. Everyone likes to think about action potentials in neurons like the digital signals in a computer, but they're much more like a PWM signal that effects a continuous change in charge on the downstream neuron.

Re:Motion blur and bloom effects

[LaminatorX]

At very short exposure times, the length of the blur due to motion becomes smaller than the circle of confusion of the reproduced image, eventually falling beneath even the circle of confusion of the image capture medium. Generally, though, if you increase the magnification enough, you still see blur.

For reference, when examining negatives under a microscope Ansel Adams could no longer detect a difference between a handheld shot and a tripod shot of the same scene at exposures shorter than 1/500 of a second with a 50mm lens. The motion blur from his hands at that speed was smaller than his film and lens could resolve.

However, with a 300mm lens, he'd have had to shoot much faster to achieve the same equivalence, due to the higher lens magnification.

Re:Motion blur and bloom effects

[Anonymous Coward]

Simply stated, motion blur is anti-aliasing in the temporal domain. Anti-aliased text and pictures look much nicer even though their spatial resolution is unchanged. Why should temporal resolution behave any differently?

Re:Motion blur and bloom effects

[Anonymous Coward]

For an easy example that 30fps isn't enough, run iTunes, play some music and turn on the "iTunes Classic Visualizer" full screen. Hit "F" to display the frame rate, then use "T" to toggle the 30fps limit on and off. Tell me you don't see a big difference.

Okay. I don't see a big difference.

Re:Motion blur and bloom effects

[stevenvi]

For an easy example that 30fps isn't enough, run iTunes, play some music and turn on the "iTunes Classic Visualizer" full screen. Hit "F" to display the frame rate, then use "T" to toggle the 30fps limit on and off. Tell me you don't see a big difference.

This example is flawed, because the update speed is regulated by the framerate, and not a timer. So of course there will be a difference, it's moving faster. This does not prove or disprove that the intermediate frames are actually seen.

Re:Motion blur and bloom effects

[Anonymous Coward]

Uh, no it didn't.

Re:Motion blur and bloom effects

[poetmatt]

ah, I see what you mean.

my bad on the argument here.

Re:Motion blur and bloom effects

[sperxios10]

Speaking of Physics - the properties of a game's physics engine have the properties of a Riemann sum where n=fps. so the higher your FPS the more accurate your physics simulation, even if your monitor cannot discretely display all those frames.

[note: only applies in games where physics ticks/sec are tied to framerate... which is almost all games]

Actually all decent FPS engines have geometry/physics engines quite distinct from the graphics-pipeline!

The geometry/physics engines work on body bounding-boxes and their respective velocity-vectors describing their trajectories, and they try to solve the intersection-problem among all bodies with regard to time, by responding with a timestamp - the collision-timestamp - to questions like this:

"When is body A going to hit body B?"

And on that collision-timestamp an event is scheduled, for the game-logic to kick-in, to calculate the new body-trajectories, or deaths, new body births, sarpnels, whatever.

The physics/geometry usually runs on the game-server *simultanesous* with the clients to avoid sending back-and-forth excessive info into the network. The server is only authoritative for the game-logic decisions. Yet the client runs additionally the graphics-pipeline which uses the next-frame's timestamp to calculate the body-positions on the 3D space.

But sometimes there is a slight delay between the collision-timestamp and the response from the server about what to do next (the game-logic's decision), that may allow a body to be drawn past its collisions point, and this is what make us think that FPS affects physics.

To sum it up, fps has nothing to do with physics, even if some times it seems that way.

Re:Motion blur and bloom effects

[Khyber]

Smaller lenses make for higher magnification.

For example, the macro mode on my 62mm lens is much less powerful than the macro on my 50mm lens. the 50mm lens also has a longer zoom range.

Re:Motion blur and bloom effects

[Anonymous Coward]

I thought that while the eye sends periodic updates, They still have chemical inertia (for lack of a better wording) meaning that if it went from full black to a little bit of light. The rods and cones would take a small amount of time to fully stabilize. This would also give a motion blur because the rodes and cones wouldn't instantly see a sharp picture of moving light because they were still effected by the objects last positions.

Re:Motion blur and bloom effects

[vikstar]

Why do we see the wagon wheel effect if updates are asynchronous? There must be some synchronization going on down the line.

Re:Motion blur and bloom effects

[Shadow of Eternity]

A 24fps film works because it's film, which is utterly different from an interactive media with an unstable framerate. You may as well try to claim that 60hz on an LCD is the same as 60hz on a CRT. Sure they're both "60hz" in name but they're pretty much incomparable outside that.

Make a 24fps game blurry and all you'll have is a difficult to play laggy slideshow that's also blurry on top of everything else.

Re:Motion blur and bloom effects

[FlyingBishop]

You're missing the point. A frame has a point value for every pixel at a given point in time. The eye has no such thing. Each pixel is transmitted individually and asynchronously to the brain, so even though an individual 'pixel' has a rate of refresh, you could conceivably have a different refresh time for every cone on the eye.

Re:Motion blur and bloom effects

[pluther]

If you're talking about this, then it's because you are getting what you see on TV confused with what you see directly with the naked eye.

You don't get the wagon wheel effect with the naked eye.

You also can't actually outrun an explosion or dodge helicopter blades.

Re:Motion blur and bloom effects

[dzfoo]

There's also a matter of scale. Take, for example, a car racing game: The monitor cannot represent the actual full view of a driver sitting at the cockpit in its true size, so it scales it down so as to include all the necessary information in a smaller viewport. However, portraying this accurately, e.g. scaling down the effect of physics, such as parallax displacement and speed, would make the game "feel" as if you were farther away, and not really sitting in the cockpit; thus breaking the illusion or at least making it seem wierd by entering the "uncanny valley".

      -dZ.

Re:Motion blur and bloom effects

[vikstar]

You didn't even read your own link. So for the benefit of people who may stumble upon your misinformed post let me say that the wagon wheel effect is visible with the naked eye under continuous illumination, which happens to be mentioned in your own link.

Re:Motion blur and bloom effects

[phoenix321]

Human reaction is less than 100ms in trained people when muscle contractions have only short distances to travel, like in gamers pressing a button.

Human perception, on the other hand, is a lot faster than 100ms. Humans just can't output a useable signal (pressing a button, stomping on the brake pedal, talk etc.) fast enough.

It's mostly about visual smoothness and the naturally fluid looking motion it produces. Maybe movement prediction works better then, but it certainly feels more immersed. Quake 1 and a recent computer are capable of reaching insane frame rates and hard limited 60fps certainly look a lot better than hardlimited 30fps.

Re:Motion blur and bloom effects

[Anonymous Coward]

Why do we see the wagon wheel effect if updates are asynchronous? There must be some synchronization going on down the line.

Do you see the wagon wheel effect under sunlight or artificial light? I've seen it a lot under street lights.

Re:Motion blur and bloom effects

[dzfoo]

>> it leads to "so the engine should try to do motion blurring accurately.

But it can't, at least not in any reasonably realistic way. Your eyes blur images that are moving relative to your eye movement. This means that if an object moves and you follow it with your eyes, the image will not be blurred, as long as you approximate its speed while tracking it.

On the other hand, a video game would blur the images unconditionally, regardless of your focus point or eye tracking movements.

This ultimately does not solve the problem, and may even seem more wierd than the alternative flickering.

The blurring illusion would only work if you keep your eyes centered on the screen and resist any natural urge to focus on or track individual moving objects. Otherwise, it's just annoying.

        -dZ.

Re:Motion blur and bloom effects

[phoenix321]

I always thought all cones and rods update independently of each other, giving a constant signal flow from all parts of the retina which is averaged and stitched into a comprehensive image. A bit like interlacing works, but with randomly dispersed imaging. That would perfectly explain motion blur, as a moving light makes several neighboring rods fire with the brain stitching together a blurred image.

We could artificially recreate the effect by assigning all pixels randomly to a group between 1 and, say, 5, and then render the scene at 300fps, but only updating pixel group 1 in the first frame, pixel group 2 in the second and so on (with pixel group 1 again rendered in frame 6 etc.) A moving light that is1 pixel wide would then light up several pixels in different groups.

That way, we would have recreated a like cones and rods perceive light and automatically created motion blur without computing anything artificial. We would just have to update the geometry, game logic and scene at a 300fps rate, and somehow only render the appropriate pixels which is probably more expensive than to just render everything traditionally at 100fps...

Re:Motion blur and bloom effects

[neokushan]

Or perhaps you simply misunderstood his statement? He said reality is blurry. Reality, as it happens, is your perception of the world around you. Your own perception, with your own eyes, will indicate that anything moving at any particular speed is "blurry". He wasn't indicating anything of the sort and anyone with half a brain could see that.

Re:Motion blur and bloom effects

[osu-neko]

Why do we see the wagon wheel effect if updates are asynchronous? There must be some synchronization going on down the line.

When you see the "wagon wheel effect", there is synchronization going on, but it's external to you, before the light even gets to your eyes. You don't see it watching real wagons moving in sunlight. You do see it on TV, in movies, or under artificial lighting which often has a stroboscopic effect. This last fact comes as a surprise to a lot of people, since their lights don't look like they're strobe-lights, but that's for the same reason a movie doesn't look like what it is: a slide-show. Pretty much any light on a dimmer, or for that matter running on alternating current, is going to have a strobe effect you can't see directly, but potentially strong enough to trigger the "wagon wheel effect" illusion under the right circumstances.

Re:Motion blur and bloom effects

[PaladinAlpha]

And if you would have read that section before jumping down the guy's throat, you would have noted that the section exists specifically to discredit the idea that the wagon-wheel effect is observable under continuous illumination, instead being in all cases attributable to factors that reproduce an externally stroboscopic effect. You are the one guilty of not reading the source material and your hostility is therefore even further to your discredit.

Re:Motion blur and bloom effects

[Anonymous Coward]

Uh, no...he did not say that reality is perception, which is a debatable philosophical position. You're putting that statement in his mouth.

He was clearly implying that the blurring is actually *out there* in the world, and not in the eye or visual cortex, which is a stupid claim.

Re:Motion blur and bloom effects

[vikstar]

Sigh, you didn't read it either, so for your benefit let me do the work for you...

"In 2008, Kline and Eagleman demonstrated that illusory reversals of two spatially overlapping motions could be perceived separately, providing further evidence that illusory motion reversal is not caused by temporal sampling [9]. They also showed that illusory motion reversal occurs with non-uniform and non-periodic stimuli (for example, a spinning belt of sandpaper), which also cannot be compatible with discrete sampling. Kline and Eagleman proposed instead that the effect results from a "motion during-effect," meaning that a motion after-effect becomes superimposed on the real motion"

So, by the latest study the included, you do see the wagon wheel effect under continuous illumination, although it is not due to "temporal sampling". The reason behind the effect is still not fully understood and the authors have proposed the ffect results from a "motion during-effect".

Re:Motion blur and bloom effects

[ericcj]

You don't get the wagon wheel effect with the naked eye.

Sure you do. Have you ever looked at the wheels of a car traveling next to you?

If our rods and cones were precisely synchronized, we would see perfectly sharp discrete changes in the rotation (like a stroboscope would produce). But we don't, it's blurry. The _net_ rotation appears to be clockwise, counterclockwise, or stationary, but there's some fuzziness in our perception of the current state of the wheel.

Re:Motion blur and bloom effects

[adamdoyle]

It is of course completely possible to calculate, and the games actually do it.

So, your calculations are correct. But they don't take into account the fact that the scene viewed on the screen is not to scale with reality. Therefore, the real speed on the monitor is less, and the blur effect is therefore less pronounced, hence the need for artificial blurring.

So you're saying a 5" x 5" x 5" object going 25 mph moves at a different velocity than a 1" x 1" x 1" object going 25 mph?? I'm sorry, but Physics 101 teaches that velocity has absolutely nothing to do with size.

Re:Motion blur and bloom effects

[Imrik]

Your monitor can't actually render something moving at 300km/h so instead it blurs the object to give the appearance of it moving at 300km/h. If you had a high enough frame rate you could render it properly and let the blur effect happen inside the viewers' heads like it would if they were actually driving.

Re:Motion blur and bloom effects

[AvitarX]

Scale.

The car next to you is a few inches long on screen, in real life it appears much bigger.

So if the car moves x number of car lengths a second, in real life you may get blur, but on screen your eye can cope with it.

I am willing to be that no amount of frame-rate will get natural motion blur in a game.

A car crossing your entire screen is not crossing your entire field of vision.

I don't know if it used motion blur, but F-zero GX did feel really fast when playing it, most games not as much to me.

Re:Motion blur and bloom effects

[Imrik]

The main reason it seems fake is that while it looks good when you're looking where you're supposed to, if you try to track things with your eyes you end up seeing objects that are blurry when they should be clear and other things are clear when they should be blurry.

Re:Motion blur and bloom effects

[Zeussy]

New Scientist ran a really good article on this, a couple of months ago. It can be found here: http://www.newscientist.com/article/mg20427311.300-timewarp-how-your-brain-creates-the-fourth-dimension.html
A 2006 experiment has put the rate of human vision at about 13 fps. People can see the wagon wheel effect in real life, without the aid of Strobing lights, television etc. After I read this article I did manage to observe this effect outside in sunlight, while travelling parallel to a car travelling at about 50km/h. Very surreal

Re:Motion blur and bloom effects

[geekoid]

Make the picture sharp, the human brain will blur what the person isn't looking at.

Re:Motion blur and bloom effects

[spun]

Why can't it render something moving at 300km/h? Could it show a movie of something filmed moving at 300km/h? If so, then it could render it. I'm sorry, but this explanation makes no sense.

Re:Motion blur and bloom effects

[spun]

What? Sorry, this makes no sense. Can your monitor display a movie of a car moving at 300km/h? Then it could render the same thing in a game.

Or are you saying, "A movie of a car is not the same thing as a car?" Because if you are saying that, then thank you captain obvious.

Re:Motion blur and bloom effects

[DM9290]

"They also showed that illusory motion reversal occurs with non-uniform and non-periodic stimuli (for example, a spinning belt of sandpaper), which also cannot be compatible with discrete sampling. "

according to wiki they were concluding that illusory motion reversal in non-uniform and non-periodic stimuli disprove the theory of discrete sampling.

You asked
"Why do we see the wagon wheel effect if updates are asynchronous? There must be some synchronization going on down the line. "

You are arguing from ignorance..

And you cite an article that contradicts you as if it was in agreement with you.

finally you say

"The reason behind the effect is still not fully understood and the authors have proposed the effect results from a "motion during-effect"."

which has absolutely nothing to do with your original position.

There is a huge difference between "the reason behind the effect is still not fully understood " and " There must be some synchronization going on down the line. ".

Re:Motion blur and bloom effects

[rjstanford]

And if they are going fast enough, they never appear on your screen at all, whereas in reality there'll always be something that interferes with your vision, assuming that your eyes are open and facing the correct direction.

Re:Motion blur and bloom effects

[AvitarX]

I'm saying that things on the monitor are scaled down when displayed to you.

So even though the car moves at 300km/h at scale, in reality is is moving by much slower on your monitor.

If we measure things as apparent inches/second at 1 foot away, the monitor has things moving at less of them than real life.

In a racing game, take the width of your windshield as a reference (http://img186.imageshack.us/i/streetchllngeexv8e2d462ic2.jpg/), on my monitor that would make the width of my hood 12 inches or so if my face was a foot away (small monitor). In real life the width of my hood is wider than my entire field of vision. So a rapidly moving piece of debris flying over my hood is moving across my field of vision much faster in real life than in a game.

If things are measured in fields of vision, games are scaled down very much from real life, thus no motion blur.

It probably still only makes sense in my head though.

Re:Motion blur and bloom effects

[RJFerret]

(and similarly TV uses ~30fps)

Actually, NTSC TV doesn't use 30 frames per second, it uses 60 fields per second! This allows smoother movement as well as something revolutionary, color! ;-)

Movies (at 24 fps) look like they stutter to me, especially any sweeping pan motion with the camera.

I used to do 3D animations in video production, and let me tell you, the difference between rendering an animation at 30 frames per second versus 60 fields per second is VERY noticeable (as you presumed) and interestingly, was noticeable for sub-pixel movements too. (Which someone earlier presumed wouldn't matter.)

Persistence of vision allows that rate to not appear to flicker, but nobody ever suggested 30 fps was enough for smooth fast motion.

That being said, market economics have dictated cartoons at 15 fps for broadcast...

As the expression goes, money talks.

Re:Motion blur and bloom effects

[OldSoldier]

Here's a question...

If the graphics card puts out say... 120 FPS while the monitor only handles say 30FPS could motion blur be made to happen because of that difference? With both 120 FPS card and monitor the blurring occurs in the eye. At the other end of the spectrum, some graphics cards may compute the blurring directly in the video card and output a blurred image to the monitor. I'm wondering if there's a 3rd way, either now or possible in the future, for an intentional mis-match between graphics card FPS and monitor FPS such that some "buffer" between GPU and monitor could average 2-4 frames worth of graphics card generated high-frame-rate/non-blurred images to present a blurred image on the monitor.

(For that matter, normal 'spatial' antialiasing could be done by generating 3200x1800 images and averaging 4 pixels down to 1 at 1600x900 resolution).

 

Re:Motion blur and bloom effects

[LordKazan]

the graphics engine still has to acquire locks on the data to render it. which is going to prevent the physics engine from altering it - thus introducing the properties i described.

Re:Motion blur and bloom effects

[spun]

Let's take a step back here. My post, the one you answered, reads like this:

"Then how is it, in games like Gran Turismo with real world cars and tracks, that actual race car drivers turn in near-identical times on virtual and real tracks with the same car?

A car traveling 300 Km/h travels 300,000 m/h, or 83.3 m/s. At 30 frames per second, that's around 3 meters per frame. Are you saying this isn't possible to calculate?"

Now, here's your reply to that:

Scale.

The car next to you is a few inches long on screen, in real life it appears much bigger.

So if the car moves x number of car lengths a second, in real life you may get blur, but on screen your eye can cope with it.

I am willing to be that no amount of frame-rate will get natural motion blur in a game.

A car crossing your entire screen is not crossing your entire field of vision.

I don't know if it used motion blur, but F-zero GX did feel really fast when playing it, most games not as much to me.

Can you explain how what you wrote in reply applies to what I wrote? It isn't even answering the question I asked.

But let me just ask you this, to perhaps clear things up: do you think a movie of a car moving at 300km/h will show motion blur in each individual frame? What if the movie were shot using a high speed camera, say, 100,000 frames per second? Would individual frames show any motion blur then? I don't think they would. Yet, when shown, I do believe the eye would see motion blur.

Okay, so then, what if we simulated a car moving at 300km/h, at 24 frames per second. Obviously, if the 24 frame per second movie contains motion blur in each frame, our simulation would too. We don't have the tech to do 100,000 frames per second, but if we did, then I don't think we'd need to show motion blur either, the eye would create it.

Re:Motion blur and bloom effects

[McGiraf]

"Movies (at 24 fps) look like they stutter to me, especially any sweeping pan motion with the camera."

Yes, I'm not the only one! I find this very anoying.

Re:Motion blur and bloom effects

[sjames]

Because part of it is that the object reflecting light in real life actually passed through the area you see a blur in (and so reflected light from those points). That has to be at least approximated in a realistic frame based animation.

Re:Motion blur and bloom effects

[Anonymous Coward]

Reality is not blurry, but the human eye and brain make it appear so.

Play around with the shutter settings on a video camera and you'll understand more of this.

Re:Motion blur and bloom effects

[Nicolay77]

My post was not a circular argument, but rather a crude reductio ad absurdum to disproof GP post.

Technically, the framerate is not just high, because it is not a quantifiable quantity, unless we can manufacture a sensor that can detect a particle moving a Planck length over a Planck time. The framerate of reality is more likely describable as unmeasurable or close to infinity.

Re:Motion blur and bloom effects

[Imrik]

A movie filmed of something moving at that speed will have a blur effect of its own. Both film and our eyes are effectively analog devices recording everything that happens between frames (which gets more complicated in our eyes which are asynchronous analog devices lacking distinct frames) Computer monitors, however, display a single instant for each frame and most cannot display enough frames per second for smooth movement at that speed without fudging it.

Re:Motion blur and bloom effects

[sinrakin]

What bothers me about even 60 fps is when you move quickly in games. If I spin quickly in place, I can make, say, a small object (a bird in the sky, a tower, or whatever), move across my screen a distance of maybe 15 inches in about a quarter of a second. That's 60 inches a second. At 60 fps, every time that object gets redrawn on my screen, it's hopped across an inch of space. So even at 60 Hz, I get the sense that it was HERE, and HERE, and HERE, but not at any of the places in between, because those pixels never got turned on. Not sure if that's the same as blur (would blur just draw a blurred line across the entire screen?). But it's why I feel like I want more fps.

Re:Motion blur and bloom effects

[electrons_are_brave]

I didn't think he was saying that objects blur themselves when they move fast. I thought he meant "my hand looks blurry when I wave it in front of my face".

Re:Motion blur and bloom effects

[Anonymous Coward]

There's no rule that says reality must make sense.

Re:Motion blur and bloom effects

[electrons_are_brave]

I remember reading in psych about these people who couldn't see motion at all because they had damage to some area of the brain. They only saw frames.

Re:Motion blur and bloom effects

[DavidTC]

Heh, you actually came up with the same example I almost used in another post here.

Yeah, it's essentially like you described, but the entire system is also somewhat random (It is a biological system), so imagine that pixels drift slowly between groups. One might start as group 1, but next might update in group 2, then group 3, then group 2 again, etc.

Although you need a good deal more than 300fps to pull it off, and certainly more than 5 groups.

Computerized motion blur essentially works by calculating how each pixel 'actually' looks, and then averaging it with previous pixels.

Which sounds correct, and is mostly correct, but we do notice something is slightly off about it we can't explain.

That's because things are actually motion blurred, it's our brain creating the 'persistent' effect, and something in our brain knows that the 'old pixels' that haven't updated yet aren't 'real', or at least are 'less real'. With computer motion blur, they are 'real', it's coming on new updates, and we can tell there's something slightly wrong, somehow.

Re:Motion blur and bloom effects

[Anonymous Coward]

The synchronization is probably pattern matching going on inside your brain.

The brain can't operate entirely asynchronously. At some point it has to say 'spoke of a rotating wheel' by looking at some total 'field of vision', no matter how that picture got created.

According to the article linked to by 'Zeussy', below, scientists 'found a specific rhythm in the right inferior parietal lobe (RPL)' that is about 13 times a second. That seems to be it, in fact. Your brain runs 'What am I seeing?' 13 times a second.

But I wasn't really talking about what happened in the brain, I was just explaining how the eyes got data to the brain.

Re:Motion blur and bloom effects

[spun]

But we're talking about models of reality. Theories. Those have to make sense.

Re:Motion blur and bloom effects

[LaminatorX]

That is only true when the subject distance small relative to the focal length:

Magnification=Focal_length/(Focal_length-Subject_distance)

Your second sentence is contradictory. Zooming is re-aligning lense elements to change the aggregate focal length.

Re:Motion blur and bloom effects

[GigaplexNZ]

So you want ray tracing. Why didn't you just say that?

Re:Motion blur and bloom effects

[DavidTC]

Going from darkness to light is different.

The eyes essentially reprogram themselves in low light to up the 'gain' of rods and to ignore the cones, which don't function. It's called scotopic vision, and it does, in fact, take some time to undo. Along with the other changes of the eye, like undoing opening the iris and the altering of the composition of the stuff in the eyeball and the stuff on the rods and cones to block less photons.

It's fun to pay attention to this the next time you have to make a brightness transition. Instantly, your irises will change, which you can't really notice. Next, your rods and cones will start working differently and your color vision will change as a result. Then, over the next minute or so, the composition of the light-blocking material in your eyes and on your rods and cones will change so it will slowly get brighter or darker, and finally you'll be perfectly fine.

But none of that really has anything to do with normal, same-light vision. I'm sure there is some slight inertial effect there, but it's probably not anywhere near as relevant as the persistence of vision caused by the brain 'storing' each value as it comes in until it updates again.

Re:Motion blur and bloom effects

[Anonymous Coward]

What about the simulated depth-of-field that's used in some video games nowadays? Essentially it intentionally blurs objects in the distance. It's still far from perfect, but it does seem to improve things quite a bit.

On the other hand, I dislike simulated motion blur in games. Feels really fake, and a lot of games do it excessively and don't let you control how much blur there is.

Re:Motion blur and bloom effects

[the biologist]

The fly's wings aren't blurry. Your perception is blurry. Perhaps you should upgrade your eyes.

Re:Motion blur and bloom effects

[Dutch Gun]

BTW, 24 or 30 FPS will cause horrible flicker. That's why film projectors flashes each frame twice on the screen, and TV broadcasts interlaced the image, showing half of each frame in 1/60th second increments.

I think we need to make a distinction between the ability to see flicker versus the ability to perceive smooth motion. And again, both film and TV make use of the camera's natural ability to preserve blurred motion, which helps fool our eyes. An algorithmically generated visualizer without that blurring will naturally look worse.

Re:Motion blur and bloom effects

[Sark666]

Well maybe my eyes are shit but for me I can't tell anymore once I reach the early 70's. I did this test awhile back, so it would be useless with current lcds if your refresh maxes at 60.

I launched quake 1 on a system that could easily render anything in that game. I think it was something like 800x600 with the monitor at 85hz refresh, vsync on.

Launched a map and used the arrow keys to rotate around, with it capped at various limits. 30, choppy as hell, 45 a little better but still easily noticeable, 50 and so on.

At 60 I perceived that smooth as glass feeling but when I went to about 72, I noticed it being a little smoother still. Any higher than that and I couldn't tell a difference.

Re:Motion blur and bloom effects

[the biologist]

Except you're wrong.

The macro mode designed into your lenses don't have anything to do with the power of the focal length. The macro mode on my 300mm rig is no doubt much more powerful than your 55mm lens (of course my lens is no longer stock).

The main difference in macro ability across lenses with different focal lengths is that the longer lenses focus further away from the lens. With some 50mm lenses, the close focus range falls to within the lens itself. (you can image a fingerprint on the front surface of the lens.)

And the zoom range... um no. 2 ft to infinity (50mm lens) is not a greater zoom range than 5 ft to infinity (62mm). It looks like more, but you're contradicting a mathematical fact.

Re:Motion blur and bloom effects

[twidarkling]

Having actually run projectors, both ones 2 decades old, and 5 years old, I have to say, you're full of shit. Film projectors in movie theatres do NOT show each frame twice. How do I know this? I've hand-cranked them to ensure they were threaded correctly. Frame is shown while shutter is open, frame moves while shutter closes. This allows it to not be a smear across the screen while the film moves. Showing the same frame twice in a row wouldn't do shit but *decrease* the frame rate, since you'd be showing 12 frames in the space of what should have been 24.

Now, if you mean "each frame is duplicated on a reel, making it twice as long as it would have been had each frame only been present once," again, my time splicing reels together to run on those same projectors proves you an idiot, not to mention naive at best. Film is expensive to make, and costly to transport (and needs to be transported securely). You really think they'd have designed a system that takes up twice as many resources as they could otherwise get away with?

Re:Motion blur and bloom effects

[GrumblyStuff]

Maybe the sun is just blinking really fast?

Re:Motion blur and bloom effects

[Vastad]

Either we keep tracking it (in which case the unfocused foreground and background areas alter) or we don't, and it goes out of focus. We mentally render this as blurring. Directors in 2D movies use depth-of-field to do a quick transition between two speaking characters and ensure the right one has prominence, by keeping the speaker in focus and then quickly shifting focus in/out to bring the other to prominence when the dialogue turns.

The real sin, and unalterable problem currently, for 3D technology is that everything renders in-focus. Motion blurs work to some degree, but a large-scale image with "background" objects sharply in focus gives us headaches. We follow the other visual cues, try to "focus" to distance, try to "refocus" for the fuzziness it causes, and then wobble back and forth till we have sore, tired eye muscles.

This explains an annoying game engine "feature" in Dragon Age: Origins. During in-game engine scripted cutscenes, they would do that character-centric depth-of-field technique of 2D directors by having the character in perfect focus and the background fuzzed out.

Problem is it was done in the cheapest clumsiest way possible. It was like a pseudo-green screen effect except done with a lame alpha layer that was roughly the outline of the character speaking. Everything else looked like it had frosted glass in front of it.
But the alpha shadow wasn't perfect, particularly around the hair. You could see like a clear "window" through the "frosted glass" behind the character, revealing a perfectly rendered background. Furthermore, the transition from focus to blur was sudden, not gradual, further highlighting the alpha shadows presence. It totally destroys all cinematographic merit in all the cutscenes.

The irony is that when you leave the cutscene and go back into exploration mode, the blur disappears and you have perfect focal clarity in the game world at all distances. Probably one of the most stupid game design decisions I have ever encountered.

Re:Motion blur and bloom effects

[Anonymous Coward]

Take another look. He was refuting the statement that "reality is not inherently blurry (unless you need glasses...), our eyes and brain do that internally," which logically implies that if the blur is not in our eyes/brain, it must be "in the world" itself.

He also said "it's just hard to accurately portray blur when you're staring at something that's not moving." Well, how do you "accurately portray blur"? It depends entirely on the viewer and not the subject. WTF does "accurate" mean, if a high-speed camera can take pictures of that waving hand with no blur?

Re:Motion blur and bloom effects

[romiz]

Not all movie projectors work the same way: depending on the projector, a frame will be displayed one, two or three times before changing to the next frame. See the shutter description in the relevant article for more information.

Re:Motion blur and bloom effects

[noname444]

Having actually run projectors, both ones 2 decades old, and 5 years old, I have to say, you're full of shit. Film projectors in movie theatres do NOT show each frame twice.

I'm no projector operator but I've heard, like the OP stated, that the shutter is operating at 48 Hz in movie theaters to reduce flicker (ie. "show each frame twice", since the film is captured at 24 fps).

Turns out they do this and more since some even run their shutters at 72Hz ("show each image three times").

http://en.wikipedia.org/wiki/Movie_projector#Shutter

Re:Motion blur and bloom effects

[delt0r]

In sunlight i have never seen this. Under street lights that have a 50Hz cycle... well yea you can see it.

Re:Motion blur and bloom effects

[xouumalperxe]

Uh, no...he did not say that reality is perception, which is a debatable philosophical position. You're putting that statement in his mouth.

In fact, in countering "Perception is blurry" with "no, it's reality that's blurry", he pretty much assumed quite the opposite!

Re:Motion blur and bloom effects

[kcitren]

Maybe he's talking about Heisenberg? http://en.wikipedia.org/wiki/Uncertainty_principle Nah.

Re:Motion blur and bloom effects

[AvitarX]

But let me just ask you this, to perhaps clear things up: do you think a movie of a car moving at 300km/h will show motion blur in each individual frame? What if the movie were shot using a high speed camera, say, 100,000 frames per second? Would individual frames show any motion blur then? I don't think they would. Yet, when shown, I do believe the eye would see motion blur.

We disagree here, but it doesn't mean your wrong.

If said movie was played on a 20 inch display at 2 feet away i don't think it would. If it were played at 100,000 FPS on a screen significantly wider than my field of view, I believe it would.

Re:Motion blur and bloom effects

[Acaeris]

This isn't an XBox 360 thing. Borderlands is built on UT3 which shows whichever version of a texture loads first then any more detailed versions as they finish loading.I presume a similar arrangement is implemented by Bungie in their own engine.

Also, you'll occasionally notice in any UT based game that if you are moving quickly (via vehicle for example) it still goes through this process even if the textures are already ready (you'll quickly see the front of a vending machine in Borderlands flick through different detail levels for example).

Re:Motion blur and bloom effects

[Nicolay77]

Now that is amazing.

Re:Motion blur and bloom effects

[Blakey Rat]

This isn't an XBox 360 thing.

I didn't say it was "an Xbox 360 thing," I said it was visible in a good amount Xbox 360 games. Which is true, as a good amount of them use either the Halo or the Unreal3 engine, and those engines both exhibit this behavior.

But thanks for "correcting" me, Mr. Can't Read Fucking English.

Re:Motion blur and bloom effects

[Nicolay77]

My comment is just a reductio ad absurdum attempt to disproof of the GP premise. It doesn't have anything to do with my eyes.

As I have been downmodded and flamed for my comment, I assume my attempt is flawed.

The reductio ad absurdum works like this:

1. You assume the premise is true.
2. You then logically reach a contradiction.
3. Premise is disproved.

Finally, I don't need to upgrade anything in my eyes, at least for one year, as I have new eyeglasses.

Re:Motion blur and bloom effects

[squiggleslash]

Unless your monitor is an exact multiple of 30fps (which it probably isn't, even an NTSC TV signal is 29.97ffps, not 30) you're going to see jerkiness with your example.

This, incidentally, is why, for example, the "downward pan" at the beginning of Star Wars (to name a famous pan) appears to be jerky when you watch it on TV, no matter how good your progressive-scan DVD player is, yet appears perfectly smooth when seen at the cinema. Incompatible frame rates are a real problem, and that 30/1.001 crap was one of the worst technical decisions made in TV even if it did appear to make perfect sense at the time (the reason was that TVs would have needed to be slightly more expensive to filter out various frequency artifacts had the framerate been exactly 30ffps.)

(ffps = full frames per second, in case anyone thinks I typo'd - full frame = two fields.)

Re:Motion blur and bloom effects

[Anonymous Coward]

well, you sure got your ass handed to you. who's the idiot and full of shit now, you over-aggressive cunt?

Re:Motion blur and bloom effects

[phred75]

Actually that was the Unreal 1 engine that did that.

Re:Motion blur and bloom effects

[Dutch Gun]

Well, I was a projectionist at a 5-plex (about 20 years ago). No, the frames were not duplicated. As you pointed out, that would be ridiculous. The films area already huge and cumbersome to transport and maintain. And yes, I'm well familiar with the star-cam and shutter mechanism in projectors. On the projectors I ran, the shutter opened twice on the same frame for each full revolution of the cam.

Here's a link to a patent that describes a particular star-cam mechanism.

A quote of interest from that article (emphasis mine):

The reason that the shutter must close during pull down is that the projected movie image would be degraded if the moving film were projected onto the screen. Therefore, the projected movie image necessarily "flickers" as the shutter opens and closes. It has been found that a flicker rate of 24 Hz produces a noticeable flicker and is objectionable to the audience. This problem is much less noticeable at a flicker rate of 48 Hz. For this reason, it is common to use a shutter which closes again while the film frame is motionless in the projection gate. From the standpoint of flicker, this results in a good quality movie projection.

Another important aspect of movie projection quality is screen brightness. While closing the shutter twice per frame is good from the standpoint of flicker, it is bad from the standpoint of screen brightness. To achieve high screen brightness while still having a shutter rate of 48 Hz, the duration of the time the shutter is closed in comparison to the time that it is open should be as short as possible. But the length of time the shutter is closed is determined by the time required for film pull down. So screen brightness can be improved by reducing the film pull down time.

I know it's fun to jump on someone you think is wrong, but at the very least, please make sure you're actually correct before you do so.

Re:Motion blur and bloom effects

[Khyber]

You are wrong. The smaller 50mm lens, with EVERYTHING from 2 feet to infinity, has a greater zoom range than the 62mm, which starts out at 5ft and goes out to infinity, by 3 feet - 2ft vs 5ft out from the lens. Your math sucks.

I'm correct. Smaller lenses = greater magnification at closer and further distances with proper lens distance. This is why microscopes use TINY LENSES, and telescopes (my 14" is a good example) use tiny lenses for the eyepiece - the one I'm using right now is 6mm. Makes for AWESOME views of Saturn's rings compared with the default 15mm lens that comes standard with the telescope.

Try again when you actually take a class on the subject, mmmkay?

Re:Motion blur and bloom effects

[Khyber]

No matter the magnification device, telescope, microscope, a smaller objective ocular piece will make for a more magnified image, sure there is less light imput but the image is always larger with smaller lenses. This doesn't hold true for anything that uses a mirror as a magnification device, only glass prisms, but even mirror-based telescopes (like my 6" meade) have eyepieces with smaller optics for higher magnification.

Re:Motion blur and bloom effects

[Acaeris]

I apologise but your message came across as 'A decent amount of XBox 360 games' (because they on the XBox 360) 'have some form of delayed-texture loading'

Additionally, I have only seen this artifact on UT3 based games. There are only 2 game running on the Halo engine (both internal) and I haven't seen this on Halo 3.

Re:Motion blur and bloom effects

[Blakey Rat]

Additionally, I have only seen this artifact on UT3 based games. There are only 2 game running on the Halo engine (both internal) and I haven't seen this on Halo 3.

First of all, there are 4 games using the Halo engine, and one in development. Secondly, what does "internal" mean in this context? It sounds like you're suggesting that Halo 3 was never released outside Microsoft. But, see, instead of assuming that and calling you wrong, I post a question to clarify. Amazing.

Re:Motion blur and bloom effects

[Acaeris]

I'm actually amazed how easily I managed to annoy you completely unintentionally.

First, to the best of my knowledge, the Halo 3 engine (i.e. the one for the XBox 360) is internally developed and, unlike Unreal Engine, not licensed out to third parties. Thus, the only games developed with that engine that have seen public release are Halo 3 and Halo:ODST. I am well aware that Halo:Reach is in development.

Second, at best, all I have done is mis-interpret your use of 'XBox 360 games' in a sentence that could have easily just used 'games'. I've even apologised for the mistake and clarified my position (twice now).

If you don't mind me saying, you appear to have misinterpreted my own posts so it's not as though this is a one way thing. So, I will clarify what and why I posted:

I posted a simple statement clarifying that this is usually an engine based thing and usually not an improvement on the standard method of loading textures. I had read your comment referencing XBox 360 games as though it was inherent to the XBox 360 and thus replied, not with the intention to shoot down your post but to add clarification to it.

Now, maybe, you can see my reasoning, attitude and point for the post, though I get the feeling I should expect another slap across the face for apologising on slashdot.

Re:Motion blur and bloom effects

[Blakey Rat]

First, to the best of my knowledge, the Halo 3 engine (i.e. the one for the XBox 360) is internally developed and, unlike Unreal Engine, not licensed out to third parties.

Why is that relevant to this conversation?

BTW, there's just one Halo engine, under continuous development. Halo 2 actually runs in the "Halo 3 engine" if you run it in an Xbox 360. Thus, 4 games have been made with it.

Now, maybe, you can see my reasoning, attitude and point for the post, though I get the feeling I should expect another slap across the face for apologising on slashdot.

Yup!

Re:Motion blur and bloom effects

[Acaeris]

Why is that relevant to this conversation?

I get the feeling you are just reading the replies and forgetting the context of them. You stated that there had been 4 games developed for the Halo engine. I was replying to that statement. Technically, you are right, there are 4 games developed for the Halo engine, just as their are technically 6 games developed for the GTA engine. However, I highly doubt Bungie or Rockstar consider the ongoing developement of those engines to be one version of the engine. For example Halo runs on Halo 1.0, Halo 2 on Halo 2.0 and Halo 3 and ODST on Halo 3.0 and 3.1 or something along those lines. We consider Unreal Engine to have 3 distinct versions yet that too has been a continual developement, so why is the Halo engine different in this circumstance?

The fact it is internally developed and used and not licensed out means that logically the only games that could have been developed for the third generation of the Halo engine were Halo 3 and ODST

Also, Halo 2 runs in a patched version of Halo 2's Engine on an Xbox 360, it does not suddenly use the features that were added in Halo 3.

So, we've taken a detour over the existence of a console name in sentence that didn't require it. We're now discussing what makes the Halo engine development cycle any different to the development cycle of Unreal Engine. Where would you like to go next?

Re:Motion blur and bloom effects

[the biologist]

2ft to infinity = infinity.
5ft to infinity = infinity.

You're stating, "infinity > infinity", which is not the case.

If the cardinality of these two scenarios was different, you would have a bit of wiggle room in the argument (as infinity of an larger cardinality could indeed be referred to as larger than infinity with a basal cardinality). Not the case here.
--

The focal length of a lens is the distance at which parallel light (ie, rays from infinity) focuses past the lens. Long focal length lenses have a greater magnification because they make a small real image into a larger virtual image. Short focal length lenses have a lesser magnification because they make a large real image into a smaller virtual image.

Small lenses have a large depth of field, but allow in less light due to the smaller aperture.

In most microscopes, you need a very small diameter lens to get a useful depth of field comparable to a cell depth. The problem of low light is compensated for by a powerful light source.

In telescopes, you need a large diameter lens to allow in as much light as possible. The very small depth of field resulting from this lens is of little importance, since the telescope is focused at infinity.

Your 6mm focal length telescope eyepiece does improve magnification of the system. The overall focal length is found by dividing the main optic focal length by the objective focal length (decreasing the objective focal length increases overall focal length). This is not the same as with camera lenses which calculate out more like simple theoretical lenses.

Magnification does not relate to size of lens and relates to focal length of a lens in the reverse of how you described.

The telescope objective example fails because you're modifying the overall optic system opposite to the way you think. (Is this where your intuition comes from?)

Re:Motion blur and bloom effects

[Anonymous Coward]

lawlz

Really?

[non0score]

Not sure how much it matters to sales, though.

Re:Really?

[Sockatume]

Graphics are sold by screenshots and by box shots. YouTube and so on might make a difference, but ultimately you'll get more players to swoon with half the framerate and twice the geometry, than vice versa.

Re:Really?

[Narishma]

Aren't youtube videos limited to 30fps?

Re:Really?

[Anonymous Coward]

Thought this would come up and well from their perspective if sticking to a solid 60fps is costing them money to do and then getting less sales in return, well then I understand dropping that standard.

At least they are being honest about it, unlike all the faking of screenshots and trailers going on these days. Say if you compare the Forza 3 trailer to the Gran Turismo 5 one (may not be true for all their trailers). The former just use ingame footage, the latter are rendered at slower than realtime with lots of AA and interframe blending making a beautiful video, but ultimately something the PS3 can't actually do. (To its credit GT5 does run at 1080p60 though).

Its like when we get sent Wii screenshot at well above 480p resolution, they actually look pretty good but there is no way in hell the actual game will look like that.

Re:Really?

[Anonymous Coward]

The poll in that article is constructed badly. There is no choice that expresses "framerate is important to me", other than the extreme "60 FPS or I won't buy!". The high number of respondents that pick this ridiculous extreme, leads the author to dismiss the results as invalid.

Picking out 60 FPS as a magic number is fallacy. Most of us who know about framerates know this; what is needed or worthwhile varies by game.

Framerate helps determine how 'fun' a game is, just like better graphics. And arguably, framerate is more important for long-term enjoyment. On games with adjustable graphics, I might like turning things way up once in a while, but if I spend long at it, I'll want to tweak towards faster rendering, not for the performance boost this yields in some games, but just because it's more pleasant to play through.

Where it matters most.

[Icegryphon]

In fighting games you need 30FPS period.
There are books for Tekken and the like that have frame data for every move.
Input any lag into the equation and what might be safe of block, might not, costing you the game.

Re:Where it matters most.

[Anonymous Coward]

Huh. And here I thought that those fighting games just randomly picked a move as the players sat there randomly bashing the controller buttons as fast as they could. Leave it to Koreans to prove me wrong.

Re:Where it matters most.

[91degrees]

It will work just as well at 25fps, 60fps, even 42.77654fps. Maybe you'll have trouble below 24fps but the main importance is consistency.

Re:Where it matters most.

[Icegryphon]

No you are thinking smash brothers.
Real fightans are a song an dance of multiple move and mind games against another person.
CPU's are to stupid to play with your head, they typically follow a pattern.

Re:Where it matters most.

[Anonymous Coward]

For games where timing and targeting prescision is required, you need AT LEAST 30 fps. Less than that leads to noticeably worse performance. The open question is, if more than 30 fps does make any difference. It probably won't improve accuracy by much, but if you reduce your monitor rate to 60 Hz, you'll pretty soon notice that the eye can detect a lot more than 30 pictures per second. It does just not affect our sense of motion by much.

Re:Where it matters most.

[Anonymous Coward]

Tekken? 30fps? Real fighting games run on 60fps.

Re:Where it matters most.

[MozeeToby]

I think what you're getting at is that consistancy matters more than maximum frame rate. For different reasons than the one you state, I'd rather play a game at a constant 20 hz than at 30 (or even 60) hz most of the time but dropping down to 15 during the most intense moments. It's the large changes in framerate that are noticable, your brain can fill in the missing pieces if the framerate is constant.

Re:Where it matters most.

[spun]

You don't understand how frame rate works, do you? The pictures drawn on the screen aren't the real model the game uses. Adding frames in between other frames won't generate lag (if the processing speed is high enough) So, if activating a block at a given frame works with 30fps, it will work with 15fps, 60fps, or 300fps. The frames aren't the 'real thing,' the game's unseen internal model is the real thing. The frames are drawn as often as is possible, given the hardware, and are drawn to comply with the current state of the internal model.

HTH.

Re:Where it matters most.

[SendBot]

I thought I was super badass at street fighter 2 in middle school, then I went to the arcade and saw older kids getting the insane combos on killer instinct. First thing I thought was... wow, you really have to study this stuff to know what you're doing. If only there was some sort of global information network where I could quickly and easily find out what all those moves are.

Re:Where it matters most.

[Icegryphon]

Sorry but, you are dead wrong. Input will lag behind with any other lag.

Re:Where it matters most.

[maxume]

That isn't always the case, I recall a game in the past where gravity had less effect on players that had faster hardware. Or something like that. Anyway, the logic was mixed in with the rendering, so frame rate had an impact on what the player could do.

Re:Where it matters most.

[bigstrat2003]

If you think smash bros is nothing but button mashing, you've obviously never played against someone really good (and I happen to have some friends who are really good, much to my dismay as I'm not all that great). At lower skill levels, smash bros is a lot of button mashing, just like any fighting game is at low skill levels. With highly skilled players, though, smash bros is very much a "song and dance of multiple moves and mind games".

Re:Where it matters most.

[Speare]

In many embedded apps, like coin-op arcade games, the "model" is indeed tied to the frame rate. The main loop assumes a fixed dt, and pipelines the input, update, render tasks. Often this is done without threading, just while (!dead) { do_input(); do_update(); do_render(); } in the main function. Even with threads or co-processors, they often tie the rates 1:1:1. Some have no room for adjustment, and some will at least update their dt if the render took too long.

Re:Where it matters most.

[hardburn]

Yes, that's entirely possible if it's programmed so that you fall x meters for each frame rendered. What should be done is to say you fall x meters per second, taking into account how long it's been since you last calculated the value.

(I'm simplifying the effect of acceleration above--many games could get along without it and produce a decent result, though it's not hard to factor in if you want.)

Re:Where it matters most.

[asills]

When I was 13 (1992) I manually compiled moves learned while playing and discussing Mortal Kombat and sold them for $3-$5 a piece. Who needs the internet an enterprising little kid is destroying you in MK then offers to sell you the list of moves he knows?

Re:Where it matters most.

[spun]

How so? What lag are you talking about? What, in your theory, does 'lag' mean?

Heck, could you rephrase the sentence, "Input will lag behind with any other lag," so it actually makes sense?

Re:Where it matters most.

[Anonymous Coward]

A higher framerate is better -- your moves will be more responsive and render faster.

Re:Where it matters most.

[Monkeedude1212]

You don't understand the games he's talking about.

For something like Street Fighter at EFO, they take extra steps to make sure that the framerate is consistant across all play-times, times when the players are just standing there, and times when players are attempting to break blocks for their Hypercombofinishes.

Like many flash games - there is code that is actually executed ON THE FRAME. It is done as the frame is being rendered. When you get intensive moments that have people putting alot of input, lots of stuff to draw on screen, and whatever else, there is always that chance that latency will show up, slowing the frame-rates, which ultimately changes the rest of play - because now you would usually let go of your block at a precise moment, but because the game is slightly slower, your opponents initial attack is still flinging at you, forcing you to hold your block a bit longer.

Consistancy is what they are getting at. It needs to remain at the same FPS at all times for games where code is executed on rendering.

Re:Where it matters most.

[thetoadwarrior]

Smash Bros is just as much a real fighter as any other and likewise is has button mashers as you'll find for Street Fighter, Tekken, etc.

Re:Where it matters most.

[spun]

Ah, so faster hardware will actually update the model more quickly. But does this change the way the model acts? In some physics models, I guess it would. More frames would, in fact, be more accurate. But in most simple models, would calculating more time-slices actually change anything? I kind of doubt it, so even though you are right, and visual frame rate (in a non-threaded game) is tied to model frame rate, more frames would not change the outcome.

Basically, the original poster was making it sound as if the internal time of fighting games was tied to frames, so that, if a move took 12 frames to complete at standard speed, adding more frames per second would make it complete faster. I just don't see modern games being made this way.

Re:Where it matters most.

[Icegryphon]

I was joking with the old meme,
Smash brothers gets not repsect on the Fightan scene,
this is probably because of all the Casuals/Button mashers who play it as well
There are skilled players in any game that would whoop my butt, even good old Pacman!

Re:Where it matters most.

[flows]

Killerinstinct's FAQ was the first big file (over 60k iirc ) i ever got from the internet, November 1995, and yes, i know what you're thinking, that was BEFORE searching for porn. Really. It was.

Re:Where it matters most.

[Monkeedude1212]

Which is what the parent was getting at, alot of fighting games go by frames, not by seconds. Sounds ridiculous but it makes for easier programming and its alot less resource intensive.

Re:Where it matters most.

[Kreigaffe]

It's way, way way more than that.

The old HL engine -- at least in Natural Selection, but most likely any game on that engine -- your framerate didn't just effect your gravity (which made it so that at certain framerates you could literally jump further, which meant BHopping was sicker)..

it also changed the DPS of weapons. Yep. Weapon firing rate was tied to FPS in a very very odd way. Some dudes did too much testing. Insane.

And you can, visually, tell a difference between 100fps and 50fps and 25fps. Very easily. Takes a few minutes of playing, but there's a clear difference and anybody saying otherwise eats paint chips.

Graphics don't make games good. Graphics can cripple good games. Graphics never make bad games good.

Re:Where it matters most.

[Sir_Lewk]

I know for a fact that Quake 3 did that (125fps is ideal for trickjumping mods like Defragged). The previous Quakes no doubt did it too.

Re:Where it matters most.

[lagfest]

Most games on the Quake 3 engine have that. 125 fps was the sweet spot.

Re:Where it matters most.

[Anonymous Coward]

The problem is that even in higher level play button mashing wins.

Re:Where it matters most.

[sopssa]

No it doesn't. Modern day games process games in different threads, so "lag" in drawing the screen doesn't "lag" things like input processing, physics etc.

Re:Where it matters most.

[Spazmania]

The pictures drawn on the screen aren't the real model the game uses.

That's not necessarily true. There's a long history of games relying on the graphics processor to determine when two objects overlap or otherwise meet specific conditions relative to each other. Goes all the way back to the 8-bit days when the graphics processor could tell you whether the non-transparent parts of two sprites overlapped.

Re:Where it matters most.

[TheRaven64]

Mod parent right up. The eye can only see something like 30fps (actually a bit less), but it will notice if the frame rate drops below this for even one frame. 60fps is useless if it's 59 frames in 0.5 seconds and then one frame in 0.5 seconds. The jitter is far more important than the throughput.

I didn't RTFA, but it's also useful having the physics engine running as fast as possible, even if it isn't tied to the graphics engine. For other outputs, it can be even more important. Haptics generally need at least 1Hz or they start to feel weird.

Re:Where it matters most.

[Tacvek]

In reality different game engines work differently.

There are some games where the engine cycles and game framerate are tied together, so if game calculations take to long, the game will slow down, not outputting a frame when it was supposed to. Further in this type of game rendering taking too long will slow down everything (one type of 'lag').

In most modern 3D games, if the engine takes to long to calculate to make its ideal cycle-rate, the game may slow down ('lag'), but this only impacts the framerate in so far as a framerate higher than the engine's cycle rate does no good, as there is no new data to render. If the engine slowdown was due to CPU overload, then it is possible (perhaps even likely) that the framerate may also be hurt, as that may slow rendering if any part of rendering at all is using the CPU.

Then you also have the case where the game is still running full speed, but the rendering takes too long to maintain the desired framerate, so the visible framerate drops. The game is still running as normal, so the only 'lag' this can cause is a delay in getting new information from the game engine. Somebody with good timing can overcome any 'lag' of this type in many games, since the game is still running at full speed.

Re:Where it matters most.

[Thalagyrt]

Modern games generally use the same cycle of input, update, render, repeat. The difference is instead of ticks being based on that loop's iteration the ticks are are based on system time. The end result is that a higher frame rate gives you less input lag and a more accurate simulation.

Some games, such as Supreme Commander, take multithreading to the extreme and will throw the AI and physics on one core then do the input and rendering on the other. In the case of a quad core it takes full advantage and has a core for general purpose, one for AI, one for physics, and one for GPU calls. This is an exception, not a rule, but in the future we'll be seeing more of it.

Re:Where it matters most.

[HaZardman27]

Smash Bros does not take nearly the amount of muscle memory training or attention to detail that real fighting games require to master.

Re:Where it matters most.

[Anonymous Coward]

I didn't RTFA,

Obviously.

The eye can only see something like 30fps (actually a bit less),

Had you read TFA, you might've seen a rebuttal to this point.

Re:Where it matters most.

[Anonymous Coward]

All of these "the eye can only see X fps" statements are retarded and based upon a gross ignorance of how the human visual system works. Read a textbook on perception before "weighing in" with your "insight".

Re:Where it matters most.

[bigstrat2003]

Ah, fair enough. Sorry, I couldn't tell you were joking.

Re:Where it matters most.

[PGOER]

Very correct. The word frames refers to the frames of a TV screen, not the visual acuity responce of the eye which changes person to person. I think average frequency responce threshold, for a stationary point is 20 to 25 Hz, the responce to changes in location is much higher, but varies from person to person.

Re:Where it matters most.

[Bluebottel]

Surely you must be joking. The golden standard for fighting games is 60 and nothing else.
Heres a faq for Tekken 5 Also have a look here and here

Re:Where it matters most.

[Anonymous Coward]

It sure can make mediocre movies "groundbreaking" though.

Re:Where it matters most.

[DemonBeaver]

I remember a LAN game of SupCom, with 8 players, and a 5k unit cap... after about 3 hours of playing, the whole game world suddenly froze... the graphics would work without stuttering, but it seemed as though the input was completely ignored... is that a side-effect of messing with multi-threading?

Re:Where it matters most.

[dangitman]

Fightans?

Re:Where it matters most.

[Anonymous Coward]

I think what he means is that input will lag because reacting to something on screen (which happens all the time in fighting games) will be delayed because the on-screen drawing is lagging what is actually happening.

An example of this is in online play with Street Fighter Alpha 3, where Zangief's jumping splash move (or what I call the "Hairy Chest") is rendered a lot safer because the network induced lag in displaying the image prevents players from reacting with a counter move that would hit Zangief out of the splash. By the time a player attempt to react to the visual cue, it's too late, so they either eat the hit, or are forced to block. In either case, they are placed in a disadvantageous situation.

Re:Where it matters most.

[Khyber]

The original arcade Tekken machines were 30FPS.
60FPS Tekken didn't happen until Tekken 5, where they moved to LCD screens and full progressive scan in the arcade machines. Until then, All Tekkens up to 4 were on CRT screens in the arcade, and thus did 30FPS interlaced.

Re:Where it matters most.

[Toonol]

I disagree. An expert player at SSB will beat a merely good player just as consistently as in, say Street Fighter IV... and will benefit just as much from extended training.

Re:Where it matters most.

[Thalagyrt]

I'd guess so. That game is so massively parallel that nothing would surprise me. If the input thread came to a halt, since it's not one big loop, the game should just carry on in the background not knowing what happened. I'd be willing to bet that the graphics could freeze and you could still play the game if you somehow knew exactly where all of your things were. That input glitch probably was fixed with a patch later on. It's been a while since I played it... I need to reinstall that game.

Re:Where it matters most.

[Khyber]

The human eye can detect FAR MORE than 30FPS.

And here's a simple way to prove it - find yourself some 60Hz fluorescent lighting. Look up into the light, wave your hand in front of it. Note the strobe effects, and if you're good enough you can count the different hand images and do some math to figure out your eyes average response time/FPS. Do the same thing in front of an incandescent light bulbs, notice you don't get a blur.

The average calculated human response is approximately 72 FPS.

You also 'predict the future' as it takes about 1/10 of a second for the signal from your eyes to be processed by the brain. When you play baseball and make a swing, your brain is automatically doing lots of lag compensation so you can actually hit such a fast moving object.

Re:Where it matters most.

[Khyber]

Any multiple of 3 was the sweet spt - 90 FPS, I'd always set my max limit at 333 FPS (which was nothing for a GeForce 4 ti4600 to do in Q3) and bunnyhop like mad around the map. Also, playing extreme arena, I'd just use the plasma gun alt fire to navigate a level and rain death as I propelled myself at insane speed across the map.

Re:Where it matters most.

[Khyber]

No way, Jose.

Until Tekken 5, all other prior Tekken games rendered at 30 FPS.

Because it was on a 60Hz CRT that didn't do progressive scanning.

Teken 5 introduced the LCD screen with their machines, and thus made 60FPS the standard.

Killer Instinct only rendered at 12FPS.

Street Fighter only did 8FPS.

I do make videos of each sprite and rip it out for use in MUGEN, this is how I know. 800+ characters and still expanding. Later street fighter type games moved to 15 FPS, like Marvel vs Capcom or Capcom vs SNK, but that was the era of porting from arcade to SNES/Genesis, so it made sense to keep games at that speed.

Re:Where it matters most.

[Skillet5151]

Gravity and movement in general are based entirely on time in HL as in just about every comparable game engine.  I can imagine that there might be some weird bugs with fast firing weapons at slow frame rates but you make it sound like the engine was designed by morons.

For some fun, have a look here:
http://svn.alliedmods.net/viewvc.cgi/trunk/hlsdk/pm_shared/pm_shared.c?revision=2980&root=amxmodx&view=markup&pathrev=2984

void PM_AddCorrectGravity ()
{
    ...
    pmove->velocity[2] -= (ent_gravity * pmove->movevars->gravity * 0.5 * pmove->frametime );
    pmove->velocity[2] += pmove->basevelocity[2] * pmove->frametime;
    pmove->basevelocity[2] = 0;
    ...
}

"frametime" being the key value.

Re:Where it matters most.

[Anonymous Coward]

It's way, way way more than that.

The old HL engine -- at least in Natural Selection, but most likely any game on that engine -- your framerate didn't just effect your gravity (which made it so that at certain framerates you could literally jump further, which meant BHopping was sicker)..

it also changed the DPS of weapons. Yep. Weapon firing rate was tied to FPS in a very very odd way. Some dudes did too much testing. Insane.

And you can, visually, tell a difference between 100fps and 50fps and 25fps. Very easily. Takes a few minutes of playing, but there's a clear difference and anybody saying otherwise eats paint chips.

Graphics don't make games good. Graphics can cripple good games. Graphics never make bad games good.

I was also going to post about framerate dependent rate of fire/jetpack fuel regeneration in Natural Selection, I miss the days of "guys with a good video card, weld the hive!"

Re:Where it matters most.

[Idiomatick]

To be fair he meant brawl not melee.

Re:Where it matters most.

[bmorton]

I had a friend who would destroy people in Mortal Kombat. He would then offer to sell the a move sheet for $5. They would often oblige. They wanted to play just as well.

He didn't stick around long enough for them to figure out that move sheets had all bogus information.

Now THAT is enterprising.

Re:Where it matters most.

[thetoadwarrior]

Indeed but a lot of people like to dismiss Smash Bros. because of the characters.

Re:Where it matters most.

[SendBot]

I like to brag about how I was able to get e-porn in `91 ;D

the KI file would have been nice though

Re:Where it matters most.

[Anonymous Coward]

It's way, way way more than that.

The old HL engine -- at least in Natural Selection, but most likely any game on that engine -- your framerate didn't just effect your gravity (which made it so that at certain framerates you could literally jump further, which meant BHopping was sicker)..

it also changed the DPS of weapons. Yep. Weapon firing rate was tied to FPS in a very very odd way. Some dudes did too much testing. Insane.

That's because a simple game loop design can cause physics problems at extremely high or extremely low framerates. See http://dewitters.koonsolo.com/gameloop.html for details.

Re:Where it matters most.

[AK Marc]

You also 'predict the future' as it takes about 1/10 of a second for the signal from your eyes to be processed by the brain. When you play baseball and make a swing, your brain is automatically doing lots of lag compensation so you can actually hit such a fast moving object.

My understanding is that it doesn't. It does comparisons in a look-up chart of brain and muscle memory, but not actual calculations. But things like, say, it being impossible to teach a dog calculus but have the dog be able to catch a ball with ease after sufficient practice indicate that the calculations aren't what is being done, but predictions based on experience. This is also why people good at math are no better at swinging a bat than those demonstrably worse, calculations are irrelevant to the action. It's about repetition to create the lookup tables and muscle memory, as opposed to calculations.

Re:Where it matters most.

[UnknownSoldier]

Another trick I used to do back in the day of CRTs...

On either the left side or right side, look slightly past the edge of a monitor. If your CRT has a low vsync, say around ~60 Hz, you should noticing flicker. Increase the vsync until you no longer notice the flicker.

I preferred 100 Hz as the image was rock solid steady, even though I lost brightness.

Re:Where it matters most.

[Anonymous Coward]

I don't see how gravity can be caused by framerate.

Re:Where it matters most.

[cliath]

If you jump 1 inch per frame, and you get 60 frames per second, you jump 60 inches per second... if you're running at 30 fps you jump 30 inches..

Re:Where it matters most.

[cliath]

This happens in the Source Engine as well, at least in Counter-Strike: Source. Except, its not actually tied to your FPS, its tied to your network update rate, but your network update rate has a maximum of whatever your FPS is. If you set your cvar, cl_updaterate to 100 (the absolute maxium) and your game runs at 60 FPS, your actual update rate will be around 60 ticks per second. You can also set cl_updaterate to 30, and if you're running at 60 FPS, your actual update rate will be around 30 ticks per second. At an update rate around 30 ticks per second there is a very noticeable difference in the rate of fire vs anything over 60 ticks per second.

Re:Where it matters most.

[DemonBeaver]

Me too

Re:Where it matters most.

[indiechild]

People claim that the same thing happens with the Modern Warfare series, which runs a heavily modified Quake III engine. Your weapons' rate of fire can vary quite a bit, especially the ones with an extremely high ROF.

Re:Where it matters most.

[HaZardman27]

Where an expert SFIV player wouldn't even let a moderate player touch them, two players of the same varying skill levels in SSB are still going to trade a lot of blows.

Re:Where it matters most.

[Anonymous Coward]

Quake 3 also had this issue. Physics [including rocket jumping] was tied to FPS. Ideal FPS in Quake 3 was 125, and it wasn't until later Point releases [1.32?] that they liberated the physics from being calculated by FPS.

More important to have a robust webserver

[Anonymous Coward]

./'d already...

LCDs = need even higher FPS

[Anonymous Coward]

LCDs are hold-type displays which create motion blur when you follow a moving object on the screen. This can be avoided by modulating the backlight or by increasing the number of frames per second, if the LCD can keep up with the frame rate. Some high-end TVs already interpolate video frames four-fold, i.e. they create 3 interpolated frames for every actual frame delivered by the video source. This technique combined with backlight modulation creates very noticeably smoother motion. Unfortunately this technique is not suitable for interactive sources due to the unavoidable delay created by the interpolation. In conclusion: 60fps? Give me 120fps and we can start talking about finally replacing the CRT on my desktop.

Re:LCDs = need even higher FPS

[IndieKid]

...and some high-end TVs have a 'game mode' that amongst other things switches the interpolation off to avoid the delay you speak of. Specifically, I think some Samsung models have this feature.

There is a related point though which is the fact that a number of TVs/LCD Displays claim to be 100Hz or even 120Hz but can't actually accept a 100/120Hz input. Supposedly the coming generation of '3D ready' displays will rectify this since for a comfortable 3D viewing experience 60 FPS to each eye is required.

Counter-Strike...

[Manip]

I myself used to play Counter-Strike (classic), and I can tell you both FPS and Ping made a HUGE difference in that game to the point that my score would increase as I connected to servers closer to home and used OpenGL instead of DirectX (since OpenGL almost doubled the FPS at the time).

Now, I wasn't an expert but I did play a whole lot. I think you ask most serious players and they would agree the impact of both...

Re:Counter-Strike...

[Anonymous Coward]

yeah I always seemed to kick more ass with 60 fps than 30. Also, Avatar at 24fps didn't cut it for me with scenes featuring a lot of motion.

Re:Counter-Strike...

[hitmark]

i wonder how much that had to do with the engine design. As in having the render engine and the game logic joined at the hip so that higher fps meant more repeats of the game logic pr second.

Re:Counter-Strike...

[Anonymous Coward]

+1 I'm no phile (I'm the kind of person who listens to 96kbps and never bothered with powerstrip) but i would notice if the game wasn't running near 100FPS!

Re:Counter-Strike...

[Monkeedude1212]

Agreed. Most players would notice a difference in a few miliseconds of network Latency more than a dozen frames per second, but its undeniable that extra Frames per second give you a distinct advantage.

If I see you and you see me, and you're running at twice my frames per second, You will have a smoother "turn and shoot" motion than me, which means you'll either notice your reticule over my head a slight bit faster than me, or you won't make the mistake of over or under compensating your aim since your motion was that much more sensative/responsive.

Apparently web servers also matter

[ForestHill]

she's dead, Jim

Re:Apparently web servers also matter

[Monkeedude1212]

I have a dream of one day hosting a webserver that gets slashdotted.

Just to go through the logs, and see that the referer is some site called slashdot... Just thinking about it gives em shivers.

Grammar??

[jmvbxx]

There ARE many! Not, there's many.

Cached Version

[sabre86]

Looks like it's Slashdotted already. Here's the cached page: http://74.125.47.132/search?hl=en&q=cache%3Awww.significant-bits.com%2Fframerates-do-matter&aq=f&oq=&aqi=

Re:Cached Version

[Zocalo]

Nah, that's not Slashdotted. It's proving the point by showing the importance of a higher framerate.

Than zero.

Re:Cached Version

[Anonymous Coward]

0 fps just doesn't cut it.

Doom 1?

[tepples]

By "Doom" do you mean Doom (1993) or Doom 3? If the former, I never saw this effect while playing the game on MS-DOS (vanilla version), Mac (Ultimate Doom), or GBA.

Re:Doom 1?

[Anonymous Coward]

Maybe Doom II: Hell on Earth (1995) ;)

Re:Doom 1?

[hardburn]

Doom 2 was the same engine, just with new levels. If the engine was changed at all, I doubt it was to put in a poor-mans motion blur.

Re:Doom 1?

[sopssa]

Remember that it wasn't done to get a motion blur effect, but to save in processing and drawing time. Motion blur was just a side effect.

Re:Doom 1?

[JackDW]

I've worked with the Doom source code recently, and can confirm that there was no motion blur at all. In fact, blur of any kind couldn't really be implemented, because Doom's graphics were 8-bit indexed colour.

Also, there were no engine changes at all between Doom 1 and 2.

Perhaps the GP is referring to the bobbing effect that occurs when the Doom guy runs. That moves the camera nearer to and away from the ground, changing the appearance of the texture.

Re:Doom 1?

[interploy]

I'm pretty sure that's because motion blur wasn't much utilized until after Doom. If anything the first games to really use it would have been racing games. And if I recall correctly, the original Ridge Racer even noted it as a feature.

Re:Doom 1?

[operagost]

OT, but Doom 2 also added the double-barreled shotgun, megasphere, and new monsters.

Re:Doom 1?

[Hurricane78]

No. It’s much simpler: CRTs have a reaction time for the phosphor surface. On all screens I ever had (some very expensive professional ones too), I could see the blur when I switched the screen to black, and quickly moved the white mouse cursor around.
I think on older ones that effect was much worse. And the low resolution did not make it any better too.

Re:Doom 1?

[Razalhague]

Creating a blur effect on 8-bit indexed colour is by no means impossible. You just need to be more clever about it.

Re:Doom 1?

[Anonymous Coward]

Oh indeed. Motion blur existed in the Magic Carpet games which were from Doom's era, 1994-95.

They also had SVGA, full deformable terrain, and reflective water too.

Re:Doom 1?

[Anonymous Coward]

Heh.
Multiply the monster move rate by 10 once. They'll move faster than you can see, still without blur.

That's what's great about Doom. Once you kick ass on all the levels, you can increase monster speed or endurance, along with randomize the location of all items except keys.

-mjensen

Re:Doom 1?

[Anonymous Coward]

Also on older LCD monitors, the slow pixel refresh caused smearing. I remember that specifically while playing Doom on a friend's laptop from the mid 90's.

Re:Doom 1?

[Mr. Freeman]

Are you sure that wasn't because you had a professional CRT designed for color accuracy rather than reaction time?

Re:Doom 1?

[sfraggle]

That hasn't stopped source port authors from adding motion blur effects.

Re:Doom 1?

[snemarch]

That's hardly "engine changes", though.

Re:Doom 1?

[snemarch]

...and even though Magic Carpet was one of the most enchanting games ever, the franchise was discontinued. How I pine for a Magic Carpet 3... or *proper* XCOM follow-ups :(

Re:Doom 1?

[snemarch]

That video is so low-quality I can't tell if it has motion blur or not - could just be encoding artifacts ;)

Btw, don't most of the ports actually render to 32bpp, even though the source data is 8bpp indexed? That'd make it a lot easier to add motion blur, transparency, and whatnot.

Re:Doom 1?

[Anonymous Coward]

Also, there were no engine changes at all between Doom 1 and 2.

DOOM underwent quite a few engine changes, actually; the initial release was version 0.99, and it went up to 1.9 in the end, with intermediate releases of 1.0, 1.1, 1.2, 1.666 (1.3 to 1.6 were non-public betas), 1.7/1.7a and 1.8. Both DOOM and DOOM 2 were ultimately brought up to version 1.9 (I'm not sure at which version DOOM 2 started), and of course, the changes weren't huge, mostly consisting of level fixes, engine bug fixes, network and input code improvements and the like, but it's not true that there were "no engine changes at all".

Re:Doom 1?

[JackDW]

I agree but all of those changes are either outside the game engine, or minor changes that simply correct some glitch or other. There was no major engine revision of the sort that would be needed to add a new feature, such as motion blur, true 3D or any other "source port" feature. Therefore I think it is right to say that there were no engine changes at all.

The human eye can dectect 30

[gurps_npc]

The human eye can clearly detect frame rates far greater than 30. So can the human brain.

HOWEVER

The human mind is evolutionary designed to make instant assumptions. Cat in mid air facing us = DANGER. No "Is it dead and being thrown at us?" No "Is it a picture?" As such, video games can quite easily take advantage of this evolutionary assumptions and trick the MIND, if not the brain. into thinking something is real.

So while a higher frame rate will increase the quality of the game, it is not essential. It's like getting gold plated controls on your car's dashboard. Yes it is a real increase in quality, but most people would rather spend the money on a GPS device, real leather, plug-in-hybrid engines before you get around to putting gold in the car.

Re:The human eye can dectect 30

[royallthefourth]

Clearly you have never had to work on software that was designed by someone with no intelligence!

Re:The human eye can dectect 30

[wgaryhas]

In what way does gold plating increase the quality of the dashboard controls?

Re:The human eye can dectect 30

[Anonymous Coward]

What the article and most people fail to note is that your TV is only displaying 30 fps. So any game running at more than 30 fps is dropping frames.

This has nothing to do with what the eye can detect, it is just the TV standard. You are either running 60 half frames (interlaced) or 30 full frames in progressive.

Only PCs actually allow you to have real 60fps. So anytime you hear a console gamer talking about playing a game at a silky smooth 60fps, you have to realize that they have no idea what they are talking about, just like the article writer.

Re:The human eye can dectect 30

[TheCarp]

> The human mind is evolutionary designed to make instant assumptions. Cat in mid air facing us = DANGER. No "Is it dead
> and being thrown at us?" No "Is it a picture?" As such, video games can quite easily take advantage of this evolutionary
> assumptions and trick the MIND, if not the brain. into thinking something is real.

Sort of. Its actually less "Cat in mid air" and more "This sets off a trigger based on something that happened before and hurt me".

Most adults, if you chuck a rock at their face, will toss up their arms to block, or move their head/body to dodge. This is completely learned. Do the same trick with a young child who has never played "catch" before, and your rock is going to bean him right off his skull.

From my own experience, my first motorcycle accident, I was on the ground so fast, I had to think afterwards about what happened. First two spills actually.

The one after those.... whole different story. The adrenalin hit as soon as I felt the bike start to turn sideways, by the time the bike was fully 90 degrees to my momentum vector, and the wheels were sliding out from under me, I was already calmly kicking my legs backwards and positioning myself for the impact. I hit the ground and slid 150 feet while watching my bike spark and slide away. I thought "shit I am in traffic" jumped to my feet and ran to the bike, picked it up and pushed it into a parking lot.

All I am saying is, its more complicated than that. The memory of such things and whole "flight or fight" response is an evolving and learning response. Its more than just visual, it encompasses all the senses. I doubt "cat facing us in mid air" is going to trigger much beyond anything in mid air moving towards us.

Re:The human eye can dectect 30

[GeoSanDiego]

"Designed by process of evolution"

Re:The human eye can dectect 30

[skylerweaver]

How about "optimize" instead of fashion.
After doing some programming work with genetic algorithm optimization, I realized that although evolution may choose one gene path or another, the end result is always the same: the most optimized solution for survival, tool making, etc.
So I agree, not designed; absolutely optimized.

Re:The human eye can dectect 30

[EvilBudMan]

I would say it stops a 120fps but I haven't been able to run tests past that point so far.

Re:The human eye can dectect 30

[Anonymous Coward]

It's like getting gold plated controls on your car's dashboard. Yes it is a real increase in quality,

Not the best analogy. A gold dashboard might increase the value but the quality would stay the same. I'd say it's more like having a car with better accelleration/speed.

Re:The human eye can dectect 30

[Anonymous Coward]

Actually, I read an article a while ago that explained exactly what framerate you can see.

13 FPS.

Per distinguishable object.

Asynchronously.

There's also a secondary brainwave running at 7 Hz that determines light sensitivity.

Re:The human eye can dectect 30

[Aladrin]

Congratulations. That -is- incredibly nitpicky. I'm amazed.

He is not a scientist and this is not a paper he is writing for publication. He is using the word 'designed' as the unwashed masses do all the time, and as such, he is not incorrect in his statement. Everyone knew exactly what he meant and nobody had to stop and trying to figure it out. He accomplished his task without getting excessively wordy or having to explain himself 3 times. As far as communication goes, he scored perfectly.

Re:The human eye can dectect 30

[Anonymous Coward]

"but most people would rather spend the money on a GPS device, real leather, plug-in-hybrid engines before you get around to putting gold in the car."

Most People.....? I've seen plenty of Cadillacs and Buicks running around with lots of gold!!!

SO many places to take this one. Letting it go.

Re:The human eye can dectect 30

[Changa_MC]

Design always implies a designer, so evolutionary design is always a religious reference. He communicated perfectly, but I'm not sure if he communicated what he intended.

Re:The human eye can dectect 30

[SharpFang]

Human eye and brain can tell 30, 60 and 70Hz apart.

But they don't make a difference. The cat will be processed at the natural ~15FPS, and the "feel" associated with 60Hz will be completely detached from information about incoming feline. You don't see 60 blinks per second. You just realize the display is somehow different than unchanging solid light, and your knowledge may make you recognize this kind of visual anomaly as 60Hz display. But there is no way you could tell it's 60hz blinking without being taught it. Just like you couldn't tell it's 500nm wave while seeing blue light and not knowing it corresponds to 500nm.

This isn't like gold-plated. It's like UV filter on the windscreen. You see the difference but you won't know what it means if you're not told.

Re:The human eye can dectect 30

[Troed]

the most optimized solution

Nitpick: No, think local vs global maxima.

Re:The human eye can dectect 30

[DavidTC]

You can think that, but there are a lot of things actually hardwired into people.

And you're wrong about rocks...throw things at people, and they _will_ dodge, or attempt to block, or at least wince, which is just really bad dodging. It is not learned. People can learn _not_ to do it, but people automatically pay attention to and attempt to avoid objects coming towards them.

Very very young children, as in, babies, can't actually recognize moving objects yet, and don't have the ability to do anything about them even if they did. But toddlers, for example, will certainly attempt to do something about objects coming towards them, even if it's completely ineffectual thanks to their poor hand/eye coordination and even if they've never seen it before.

Of course, with no prior training, people will just do random things. Whereas, for example, if you throw knifes at them their response will start being 'dodge', where as if you throw baseballs at them it might be 'catch', or if they commonly have a pillow fight it might be 'close eyes and take it'.

But entirely untrained people who don't normally have things thrown at their head, or even have never had anything impact their head, won't just blithely stand there and let a rapidly moving object impact it. They'll try to do _something_ instinctively, although there's no promise what they try will actually work. Training just controls _which_ thing they do.

There's all sorts of instincts we have. For a better example of _non_-learned instincts, people will try to catch things when dropped, even though this almost _never_ works. I've managed to train myself to not do that, so I can actually watch the damn thing fall and find it afterwards. People are always amazed, saying 'How'd you find that?', and I always think 'Well, if you hadn't been waving your hand around like a loon ten inches above it as it fell, perhaps you could have watched it fall also.'

There's only about a foot leeway there to catch something that fell out of your hand and people simply cannot react fast enough. 95% of the time they utterly miss, and the other 5% of the time they end up giving it a good hard wack across the room. So, considering it doesn't ever work, it is utter impossible that people have been 'taught' to do this. No, this is some instinct left over when we dropped large things, like a spear, that a) we might have caught, and b) we couldn't lose anyway. For the past 1000 years, this instinct has been rather stupid, but there it is.

Re:The human eye can dectect 30

[Troed]

Sorry, but no.

Old consoles/computers could do 240p at 60fps, even on TV sets that were meant to display 480i at 30fps. It's not that hard to figure out how it was done, you should try it. Remember, there's a lot of old C64/Amiga/Atari hackers on Slashdot and we programmed these things to do wonders ;)

With HDTV, we got 720p/60Hz etc.

Re:The human eye can dectect 30

[DoofusOfDeath]

The human mind is evolutionary designed to make instant assumptions. Cat in mid air facing us = DANGER. No "Is it dead and being thrown at us?" No "Is it a picture?" As such, video games can quite easily take advantage of this evolutionary assumptions and trick the MIND, if not the brain. into thinking something is real.

It really bothers me when people reason the way you have regarding evolution. I think you're making a huge assumption about what particular causes have lead the human brain to evolve to its present form. Unless what you listed is the only plausible evolutionary pressure that could lead to the brain's current form, I think you're unjustified in asserting that your creation myth of the human brain's present form is in fact true.

Re:The human eye can dectect 30

[skylerweaver]

Are you implying that the genetic algorithm doesn't find the global maximum?

optimize - to make as perfect, effective, or functional as possible

"As a general rule of thumb genetic algorithms might be useful in problem domains that have a complex fitness landscape as recombination is designed to move the population away from local optima that a traditional hill climbing algorithm might get stuck in." http://en.wikipedia.org/wiki/Genetic_algorithm

While it is true that sexual crossover of genes will only optimize to the best solution available in the gene pool, mutation allows a global solution to be found. When I tested this algorithm myself, I specifically chose an over-constrained problem that had many local maxima and no clear "best-solution." The genetic algorithm then found the multiple "best" solutions and had 3 distinct gene groups, with the largest being the most fit according to the fitting function. So while the entire gene pool did not converge to a single maximum, it did find the global maxima.

Re:The human eye can dectect 30

[suso]

I think these types of myths come about when marketing people have lunch with people who write articles in science magazines or whatever. I remember back in the 90s reading some article that was not talking about computer graphics, but stated that the human eye can perceive only a little over 16 million colors. Gee, isn't that conveniently close to 24-bit colorspace. The whole 30fps thing probably comes from NTSC specs.

Recently, I did see a car wheel do the wagon wheel effect with my own eyes and thought for a second that I was seeing things, but it was at night and I think this has something to do with the 60hz light sources or maybe the vibration of my own car. And no it didn't have spinners on it. The wagon wheel effect article on wikipedia says that indeed this can be seen in real life under certain conditions. But of course it doesn't mean that your eyes can only see at a limited frame rate.

Re:The human eye can dectect 30

[Troed]

How do you know the maximum you've found is truly global - outside of a (very) constrained test environment?

Re:The human eye can dectect 30

[SanityInAnarchy]

the end result is always the same: the most optimized solution for survival, tool making, etc.

I'd strongly disagree here. It will be "optimal" in some sense, but there's still a lot of baggage (vestigal stuff) from evolving, rather than being "designed". The human eye is a great example of this, interestingly.

Re:The human eye can dectect 30

[YojimboJango]

False on the account that the mind doesn't process movement at ~15 fps.

There are an average of 120 million rods (light dark receptors) in a human eye, each individual rod can only send a signal ~15 times per second. Each rod fires data about it's changed state at a different psudo-random moment, so technically the eye could track a floor of 15 on up to a maximum of 1.8 billion changes per second. The actual answer to this question is 15 * rand(0 to 1) * 120,000,000. There aren't as many cones in your eyes (color receptors), only around 6 million, but even then that leads to being able to track color movement at 15 * rand(0 to 1) * 6,000,000.

Obviously the actual frame rate that our eyes can send to our brain vastly outstrips our ability to mentally process. Determining perceptible frame rate is more a function of how fast our brain can process. Trained fighter pilots have been recorded as being able to perceive images at over 1000 fps, while I'm lucky to get 10 before morning coffee.

It's all about your target audience. For competitive FPS players with years of gaming under their belt, the investment into a monitor that can display 300fps could see advantage (yes there are CRT monitors that do this). For your first standard deviation gamer though 60 might be a bit overkill.

I find it's second or third standard deviation gamers that actually sit down to write articles or do research on the subject though. Grain of salt indeed.

Re:The human eye can dectect 30

[CannonballHead]

Hehe, touche. With hte accent that I'm too lazy to put on.

On the other hand, I don't think he was trying to say that the human mind was a buggy piece of software. ;)

Re:The human eye can dectect 30

[HaZardman27]

What if they're playing on a monitor? I connect my 360 to a PC monitor via VGA.

Re:The human eye can dectect 30

[HaZardman27]

I think if got in 3 motorcycle accidents, my brain would recognize my bike as dangerous, and I wouldn't get on it anymore. ^_^

Re:The human eye can dectect 30

[Anonymous Coward]

my rat once slipped and fell of the edge of the fridge... good thing my reflexes were there, cause i caught her just in time ;-)

Re:The human eye can dectect 30

[melikamp]

I can totally catch things I drop, like, half the time. Well, may be not that often, but totally enough to make it worthwhile (because a lot of things can break). When successful, it's mostly because the other hand was nearby already.

Re:The human eye can dectect 30

[Hurricane78]

That’s a very bad comparison, since the effects of 60 fps vs 30 fps for me literally means the difference between winning matches and losing matches. Or a crazy increase in kill/death ratio.

Even worse with the 18 fps I have in GTA IV. It’s virtually unplayable, because if there is action, it gets even worse.

Above 60 fps there no point in increasing it anymore. I tested this 120 fps (with 120 Hz): No changes in feeling and game stats. So I spend that on higher resolution, better enemy movement prediction, useful sound effects (hearing what type of room someone is in), and what’s left on graphics.

Re:The human eye can dectect 30

[skylerweaver]

By iteratively solving across the entire solution space and comparing this to the algorithm output.

For a class project, this is possible; however, I understand that you would truly apply optimization algorithms to problems in order to NOT iteratively solve them.

Re:The human eye can dectect 30

[skylerweaver]

With this I absolutely agree and I concede.

There will definately be vestigal stuff left over unless there is evolutionary pressure to remove it, and usually there isn't.

Re:The human eye can dectect 30

[Troed]

Right, and since we were talking about the human mind and other forms of biological evolution, elaborate on this "entire solution space" you mention ;)

(But yes, you're correct in your comment on an iterative approach failing real world applications)

Re:The human eye can dectect 30

[rdebath]

As usual it's a little more complicated.

For the colourspace limitations the human eye can detect a lot fewer that 16M colours it's more like 64k, or perhaps 100k. A good (YUV based) choice of 65536 colours from a 16M colorspace is usually enough for the human eye (witness the DVD colour encoding practices). BUT if you take pure primary green or a greyscale you can see steps in a 128 level (7bit) ladder. OTOH red and blue often look smooth at 7bpp.

So if you have an RGB display using bits you need 8 bits of green for a smooth scale, for simplicity red and blue are made the same which gives you 24bits. A rather inconvenient number as it happens, three bytes per pixel is slow and inefficient, either 16bpp or 32bpp is much faster.

It's similar with frame rate as someone has mentioned cones have a 15th second 'recharge' time, however, rods (B&W mostly at the edge of the eye) have a much faster response such that you can see a 60Hz flicker 'out of the corner of your eye'.

This means that with something like film, where the whole image is on the screen most of the time with only a tiny bit of blackness between frames, the frame rate can reasonably be 15 or 24fps. But a CRT has a much longer period of blackness between the frames, this leads to more of the cones 'seeing' the black between the frames and so you see the flicker.

But this is for recording of the real world where the film is sensitive to light falling on it for a large fraction of the frame time so you get motion blur. Without motion blur the brain will see motion as fast steps (more flickering) so to combat this you need to up the frame rate to generate the motion blur actually in the rods and cones themselves, sometimes as high as 100fps.

So, yet again, the answer is 'it depends' ...

Re:The human eye can dectect 30

[dreemernj]

Optimize makes it sound as if the whole process was moving it in a positive direction. But natural evolution is not always a positive process. I think the best word is simply evolved. The human brain has evolved and has not yet brought about its own extinction.

Re:The human eye can dectect 30

[gurps_npc]

Reduces corrosion. Gold does not oxidate.

Re:The human eye can dectect 30

[Anonymous Coward]

get a car, Knievel

Re:The human eye can dectect 30

[TheCarp]

Meh, dumping it a couple of times while learning isn't too uncommon. The first time was my first time comming off an incline from a stop onto level ground. More tricky than you might expect while your still learning to use a clutch.

The second was a stupid one, I was going a bit fast around a turn and the person in front of me hit his brakes. Also, being my first couple of weeks on the bike, I did the same. Boy do those wheels come out from under you fast.

The third was also inexperience. I crossed an intersection as someone turned onto my road into a speed up lane. She had a fair distance before her lane ended, and I was going much faster than her (she just turned 90 degrees). I didn't realize she wasn't going to see me from that angle and would probably assume my lane was open and she could move right into it, even before being up to speed.

The road was wet of course. Thats no excuse though, I didn't hit her and a bike on its wheels has more stopping power than one on its side. So I should have been able to stop safely. I also should have known that she can't see into her blind spot and avoided coming into it the way I did (even if I had right of way).

Re:The human eye can dectect 30

[dominious]

jesus christ...

here I start reading about frame rates in video games and I end up reading a philosophical discussion about evolution and design and then some nitpicking about nitpicking! wow!

can we get back to frame rates please?

OTOH this is /. and maybe deep inside that's why I read it :)

Re:The human eye can dectect 30

[Hyppy]

Yeah, if it was at night then it was definitely due to the 60Hz light sources. Strangely enough, you can tighten shiny lug nuts on your car in such a way that it produces the wagon-wheel effect during continuous (natural) light. It all has to do with your eyes focusing on the reflections. Humming at just the right pitch works, too.

Re:The human eye can dectect 30

[2obvious4u]

Hybrid engines are such a waist!! why do so many people promote the BS that they are in some way better. They are a piss poor electric engine and an even crappier combustion engine. They are even a bad financial decision since buying a civic or corolla which both get 35mpg or better each for about half the cost of a comparable hybrid. Go diesel, pure electric, or economy car; but please, please don't buy a hybrid, they are such a waist and a horrible fad.

Re:The human eye can dectect 30

[Idiomatick]

A child's brain isn't fully developed and can't be properly compared as you have. And visual stimulus can effect your emotion centers and cause automatic reactions without even reaching your consciousness. An example of this is surprise, it sets fear into motion before your forebrain determines that you aren't in danger.

And for the final comment. I'm sure your subconscious knows it is a cat... The eyes send a message to the occipital lobe (visual processing), which sends it to the right temporal lobe (image recognition), at which point it goes to the amygdala/hypothalamus (emotion center) which triggers the 'oh fuck I'm going to get hit in the face with a cat' response.

Disclaimer: I'm learning neurology out of curiosity, nearly everything I've said could be wrong.

Re:The human eye can dectect 30

[Ma8thew]

That's way too nitpicky. I just took a course partially about evolution, and I would guess all of my six lecturers referred to evolution 'designing' something at least once. It's just shorthand for an adaptation by natural selection. But you knew that, you're just a pedant.

Re:The human eye can dectect 30

[aaaantoine]

"Evolutionary designed" = it was "designed" via evolution. This does not have to be a reference to religion.

But to further the nitpick, he could have just said "evolved" and saved six syllables (seven if he had used the slightly more correct "evolutionarily designed", or four if he had said "evolution-designed", etc...).

Re:The human eye can dectect 30

[geekoid]

incorrect.

it improves quality, and your eye isn't one then off. cones and rods far 15 times a seconde, but not akll at the same times, so your brain is always getting information. Information you can react to without thinking. Like catching a ball the you didn't realize was thrown at you.

In fact, what you rationally think about may ahve already happened and only AFTER you think about it does the brain tell you what order it happened. You might ahve never ever realized that ball was thrown, had you not reflexively caught it. Many decisions you think you made consiencely have already been made byu your subbconsious. Many tests have shown this, and it'd pretty well known.

So the more infor4mation streamed at you the better. Just watch a 60Hz lcd playing next to a 240 Hz durinf a sports event.

Maybe it would be better to say the more information come at you, the more information you have for your brain to make a decision.

You should try to find some experimental stuff where they are showing movies at 90fps, or tv at over 500fps. Pretty amazing stuff.

Re:The human eye can dectect 30

[Anonymous Coward]

I think you need a larger keyboard =/

Re:The human eye can dectect 30

[oji-sama]

If the thing I've dropped is fragile and I can't catch it, I usually put my foot under it to soften the landing a bit. Once backfired and I kicked my phone to a wall... ^.^

Re:The human eye can dectect 30

[AmberBlackCat]

I think the article is crap on an even more basic level. Maybe a higher framerate game can beat Starfox but I bet Avatar will look a lot better than those 60fps games even shown at 24fps.

Re:The human eye can dectect 30

[brkello]

No, it doesn't. I can go to a yoga class and can jokingly say that my body was not designed to do full splits and not be talking about religion. Maybe to nitpicking idiots it is always a religious reference, but to non-screwed up people, we can use it just fine.

Re:The human eye can dectect 30

[TapeCutter]

"Design always implies a designer, so evolutionary design is always a religious reference."

To you maybe, but to everyone else not so much.

Re:The human eye can dectect 30

[DavidTC]

Yes, it's possible if you're using your other hand and your other hand happens to be there. That actually works okay.

But if you can't, if that hand is too far away, a lot of people try to use the same hand, which is just outright silly. If you're holding something at waist height, you can't catch it with the hand you dropped it from...you only have about a foot you can move your hand downward to catch it in, and you can't react fast enough to get around and have it land in your hand.

The other option to having it land in your hand is to grab it sideways...which is why people often end up batting it things across the room as they mistime the closing of their hand. Or, even worse, trying to go straight down and hammering it straight into the floor.

Hell, that's hard even on purpose. Try it. Be sure not to move your hand to catch it until you feel the object leave it. The easiest way to do that is to let it flip off the back of your hand. And be sure you've got your elbows against your body and arm straight out, like you actually carry things. Try downward grabs, sideways ones, and landing in your hand. You can do them more often then not, but I bet you'll screw up at least 25% of the time, despite knowing exactly what's going to happen in advance. (I just tried that, with a hairbrush, and ended up batting the damn thing into my can of Mountain Dew, luckily before I opened it.) It's near impossible when you don't expect it.

With the other hand...yeah, if you happen to be hold something up high, and if your other hand is free, you can catch that thing if you react quick enough.

But that's an awful lot of ifs, and in most cases it's a good deal more useful to just watch the thing fall, especially when it's something like change or a battery or various other things that are going to bounce or roll.

Anyway, like I said, even if you can catch falling things, you had the reflex anyway in the first place. Attempting to grab falling stuff you want to hold (And, clearly, if you were just holding it, you want to hold it.) is a hardwired brain thing...whether you can actually do that is a matter of training, just like deciding to do something else is a matter of training.

Specifically, I suspect that the 'tracking and grabbing out of the air' part is hardwired into the human brain, as part of necessary hunting skills...and what we're screwing up is the hand-eye (Or, rather, hand-brain) coordination. People's brains know where the thing should be, but can't really move their hand fast enough and exact enough to be exactly where it needs to be in the short amount of time given. (Unlike, say, catching a frisbee, which is more complicated than something falling, but we have time to get our hands into position.) But the brain still tries, without being 'trained' to do this at all.

Re:The human eye can dectect 30

[DavidTC]

Heh, I do that too if I don't want it to hit the floor.

I even do it for heavy things like books and boxes and stuff, which gets some rather shocked looks until I reveal my shoes are steel-toed.

I've ended up kicking some stuff, but luckily nothing fragile. Although I did once kick another _person_, although not very hard. Considering I stopped a book from landing on his foot, he was fine with it.

Re:The human eye can dectect 30

[Anonymous Coward]

A TV has 60 samples per second. Especially in music videos and sports-broadcast. They use interlacing to send 60 fields per second, but where each field(half-frame) is from a different point in time. This is a trade-off between better motion and higher resolution.

Re:The human eye can dectect 30

[Anonymous Coward]

> There are an average of 120 million rods (light dark receptors) in a human eye, each individual rod can only send a signal ~15 times per second. Each rod fires data about it's changed state at a different psudo-random moment, so technically the eye could track a floor of 15 on up to a maximum of 1.8 billion changes per second. The actual answer to this question is 15 * rand(0 to 1) * 120,000,000. There aren't as many cones in your eyes (color receptors), only around 6 million, but even then that leads to being able to track color movement at 15 * rand(0 to 1) * 6,000,000.

The eye might be able to track those changes (in theory), but your brain can't, so it is completely irrelevant.

> Trained fighter pilots have been recorded as being able to perceive images at over 1000 fps

Maybe, but the real question is, do they notice the difference between 200 fps and 1000 fps and at how many fps does increasing fps further no longer significantly increase performance?

Re:The human eye can dectect 30

[Nevyn]

Yeh, I do that with my foot too. But I'm pretty sure a lot of that is instinctual, because although I've done it to great success with class cups etc. ... I've also done it with knives and heavy metal objects (while not wearing shoes :).

Re:The human eye can dectect 30

[Anonymous Coward]

Do the same trick with a young child who has never played "catch" before, and your rock is going to bean him right off his skull.

So you've done this experiment?

Re:The human eye can dectect 30

[oji-sama]

True, I've also moved a foot under a falling knife. Luckily I had time to realize the problem with the procedure and managed to move it back ^.^

Re:The human eye can dectect 30

[Changa_MC]

You're claiming that "a general plan or intention," does not imply a planner?

Words often have specific meanings, you should use those.

Re:The human eye can dectect 30

[Changa_MC]

And when the guy next to you in yoga says, "yeah, God must really hate us," does that make him a professor in biology?

People joke about religion all the time. It's still religion.

Important in movies as well

[Anonymous Coward]

Avatar had a lot of flickering because of the frame rate. The flicker gets more obvious with 3D and Imax. Apparently there is talk of going to 60 frames for projected movies but I wouldn't hold my breath since theaters are already squealing about switching to digital projection and 3D. The technology is becoming available but I'll be surprised if they try to deploy it before the 2020s. Too bad because it would make a massive difference for action films especially 3D. With talking head pictures you'd never notice the difference.

Re:Important in movies as well

[Tacvek]

Interesting. Film has traditionally been 24 fps, and avatar was no exception to that. Some people are sensitive enough to notice flickering from such material.

However the additional flicker for Imax 3D is easy to explain.

I saw a RealD presentation which uses one at 144 fps, which means that each of the two 24 Hz frames are shown 3 times before advancing. This technology uses circularly polarized light. I did not notice any significant flicker, although I'm not particularly sensitive to flicker except in a few small frequency bands.

IMAX 3D can use either circularly polarized light, or can use active shutter glasses. It has a rate of 96 fps, meaning each frame is only shown twice, which results in a very substantial increase in flicker vs a RealD presentation.

Lastly the movie was also shown using Dolby 3D, although I don't know enough about that technology to comment on it.

Re:Important in movies as well

[Chas]

Actually what you were seeing wasn't flickering.

It was stuttering during horizontal movement. It's a byproduct of the Real3D process.

What he (she?) said

[SendBot]

It bugs me that 10 years ago I could play serious fps's and hit 100fps and actually see that on my monitor. It made a huge difference for the kind of competitive precision I was hitting.

My fancy new(ish) wuxga monitor has plenty of pixels, but 60fps feels real choppy to me.

I see some of these future 3d lcd's claiming 480Hz... is there a good inexpensive desktop monitor that can do 120Hz?

Re:What he (she?) said

[TheKidWho]

There are only 2 as of now that can do true 120hz...

http://www.newegg.com/Product/Product.aspx?Item=N82E16824116402&cm_re=120hz-_-24-116-402-_-Product

http://www.newegg.com/Product/Product.aspx?Item=N82E16824001311&cm_re=Samsung_2233RZ-_-24-001-311-_-Product

Re:What he (she?) said

[jgtg32a]

Most 120Hz monitors only do 60Hz and then use image processing to fill in the gaps.

Re:What he (she?) said

[SendBot]

thanks!

Re:What he (she?) said

[Anonymous Coward]

10 years ago I played games on a 19" NEC MultiSync monitor.

Today I play games on the same 19" tube.

I am fearful of the day this tube actually dies, because I am sure that there are no LCD's out there which can do 1600x1200 at 75hz and look this good, and I know they don't make any good tubes anymore that are affordable...

Re:What he (she?) said

[TheKidWho]

There are only 2 120Hz LCDs out right now that do true 120Hz without image processing. Both of them are limited to a 22" size however with a resolution of 1680x1050.

Absolutely

[occamsarmyknife]

I couldn't agree more. That Internal Server Error looks way better at 120 Hz on my 45" HD display.

Re:Absolutely

[Anonymous Coward]

I couldn't agree more. That Internal Server Error looks way better at 120 Hz on my 45" HD display.

And those animated gifs of the flashing siren, flaming torch, and spinning globe look so much better too!

Re:Absolutely

[Anonymous Coward]

For the record, if you have an HD display that claims it is running at 120 Hz or more, that's a post-processing trick. There are exactly two models of LCD displays out there that can do true 120 Hz. See: Samsung 2233RZ and Viewsonic VX2265WM. I own the Viewsonic and am very pleased with it as a CRT replacement :).

Headroom...

[Anonymous Coward]

the biggest reason to go for the highest frame rate possible is headroom. If your framerate is 30 at best, it'll dip down to 10 sometimes. If it's at 120 optimal it can dip down to 30, and still be playable.

Re:Headroom...

[gatkinso]

Wouldnt that be a limitation of your render, rather than that of the rendering?

Any animator knows...

[Monkeedude1212]

You can tell the difference between 30 FPS and 60 FPS.

The way I tested this was I made a 2 second video in flash, a circle moving from the left side of the screen to the right side. 60 frames. Run it at 30 FPS.

Then I made a second 2 second video, same exact positions. 12 Frames. Ran it at 60 FPS. Asked me, and all of my surrounding classmates, which was about 24 students IIRC.

100% of us noticed a visible difference in the smoothness. Whether our eyes were making out each individual frame perfectly or blurring some together to create a smoother effect, it was irrelevant since there WAS a noticable difference. I was going to slowly bump the 30 and 60 FPS up higher and higher to see at what point the difference is not distinguishable, but I got lazy (High school student at the time.)

The point I think most gamers would agree on is that more frames per second are nice - but that 30 frames per second are Necessary. You can occaisonally dip down to 24 and be alright (24 is supposedly the speed that most Movie theatres play at) - but when you get around 20 or so its really does take away from the experience.

Re:Any animator knows...

[Monkeedude1212]

120 Frames* I mean. Sheesh. Not proof reading even though theres a preview button.

Re:Any animator knows...

[Sockatume]

Time Splitters was the first game I played which was locked at 60fps: it was quite a remarkable transition, even from games which were locked at 30fps, never mind games that fluctuated (I'll take 30fps and locked over 30-60fps any day). Gran Turismo had a "Hi-Spec" mode which doubled the resolution and framerate too, albeit at an obvious graphical cost, and it looked like The Future.

On the subject of movie theatres, 24fps was chosen because it's pretty much as low as you can go before people notice problems. Roger Ebert and various others have been arguing for a doubling of movie framerates for years to no avail. Studios don't want to pay the film and (these days) CGI cost of double the frames (which is why they went with the lowest possible framerate to begin with).

Re:Any animator knows...

[StripedCow]

If you use, instead of a circle, a turning wheel with spokes, you can "see" the wheel suddenly going backward as you drop the framerate.

Re:Any animator knows...

[jeffmeden]

You can occaisonally dip down to 24 and be alright (24 is supposedly the speed that most Movie theatres play at) - but when you get around 20 or so its really does take away from the experience.

If by 'supposedly' you mean 'definitely' and if by 'most movie theaters' you mean 'all theaters and even all motion picture production processes in recent years', then yes. The difference is lost on most people, but the reason 24fps is acceptable in movies is that the frame you see isn't what happened at that instant in time when it's displayed, it's everything that happened in the last 1/24th of a second, since it's recorded on film that exposed for that 24th of a second to derive the image. When a computer does it, it only cares about what is happening at that exact 24th of a second; so the difference between a series of exact frames of motion and a series of frames that include the blur of what happens between frames is HUGE.

However, this nuance is lost on pretty much everyone who fires up a computer game, notes the FPS indicator, and goes "OMG I CAN TOTALLY TELL ITS ONLY 30FPSZZZZ!!!! HOW INFERIOR!!!". Whine about framerates all you want, but they are only a small part of the experience.

Re:Any animator knows...

[Anonymous Coward]

You can tell the difference between a rock-steady 60 fps and a rock-steady 30 fps. It feels better. But it is also true that what matters most is that the frame rate doesn't vary.
For some reason the brain doesn't like frame drops. You will get a better "feeling" of motion with a movie at 24 fps than with a game that usually goes at 60fps but has some frame
drops in complex areas that make it go to 30 fps. You feel it awkward, even if the framerate doesn't drop below the 24 fps of the movie.

I've noticed that quite a lot while playing games... I won't be able to tell you if a given game is running at 30fps or 20fps if the frame rate doesn't change. However, I would notice a
framerate drop from 60 to 30 then back to 60. Of course, if you put the game running at 30fps besides the same one running at 20fps the former will "feel" better to me.

Re:Any animator knows...

[skylerweaver]

It would be an interesting experiment to make a webpage that presents you two equivalent flash movies with varying frame rates (as you described).

Have the two movies in a random order with a random frame rate (say, 12, 15, 24, 30, 45, 60, etc)

Then have the user merely select which looks "better/smoother."

It would be interesting (as you point out) to see at what frame rate there is no longer a clear winner.

Re:Any animator knows...

[Aladrin]

Actually, the lower limit matters based on the person. My eyes are apparently a bit wonky and my lower limit is 15 fps, which would drive most people insane in a video game. Below that and it drives me insane. As for telling the different between 30 and 60... I can do it... Barely. Compare 60fps and anything higher and it's absolutely pointless for me to try. However, I've met people who can definitely tell the different between 100fps and 60 fps.

Re:Any animator knows...

[MaZeR73]

it's everything that happened in the last 1/24th of a second, since it's recorded on film that exposed for that 24th of a second to derive the image.

Sounds like you are confusing frame rate with exposure time.

Re:Any animator knows...

[sootman]

Thanks for saving me from having to write the same post. :-) And this is why ILM invented "go-motion" in the early 80s which made the animation in 1981's Dragonslayer look much better than Clash of the Titans from the same year. Video games, producing instantaneous, perfectly-sharp images, look like Ray Harryhausen films unless the frame rates get really high.

Re:Any animator knows...

[Ksevio]

Even in movie theaters, the projectors show each frame twice to give a framerate of 48 fps. Makes it look smoother that way.

Re:Any animator knows...

[Anonymous Coward]

Movies are completely different from computer games because... (drumroll)...

THE NEXT SEVERAL FRAMES ARE ALWAYS KNOWN WELL IN ADVANCE.

24 FPS where you blend between each frame and the next because there's no interactivity? OK
24 FPS where you can't blend because you don't look ahead? Ugh.
24 FPS where you blend by imposing an input delay of 1/24th of a second? People *feel* that delay.

Re:Any animator knows...

[7-Vodka]

I've always noticed I've been quite picky about FPS in games. I once recruited a friend to help me to a blind test on myself with the help of quake3. He set the max framerate in game with com_maxfps and I had to say whether it was 90 or 120fps after playing for 30seconds or less.. I got it right 100% of the time.

Try it for yourself, you can do it with quakelive as well. I may have been taking in extra queues, the game mechanics are sort of tied to the frame rate, in fact on certain maps you could only make certain jumps within a certain FPS range. It also may have helped that I have a preference for 120hz CRT screens. I haven't repeated the test on my current 60FPS LCD.

Re:Any animator knows...

[Odin_Tiger]

I think it also depends quite a lot on the type of game or what's going on in the game at the time the frame rate dips*. If quick, precise reactions / controls are important, a dip in fps is just horrible. For e.g. adventure games (even with free movement instead of in 'steps' like in Myst), or RTS's where you may need to act quickly, but not terribly precisely due to the map being made up of a grid, or an MMO like WoW or Eve where again speed is necessary, but not really precision because you lock onto and switch targets by clicking anywhere on them, and then all actions thereafter are automatically applied to the target, dipping down to 24 - 30 won't hurt gameplay. It might be annoying to me, but it's not going to really hurt me.
OTOH, with games like First Person Shooters or Platformers with difficult 'jumping / timing puzzles' or enemies that have to be attacked in a special way to be defeated, a sudden dip in frame rate will make the game noticeably more difficult.

*This is another important thing that others have alluded to - suddenly dropping from 60 to 30 fps in the middle of a game is BAD, especially since it's likely that the reason the frame rate dipped is because the game just got a lot more difficult / complicated. However, running consistently at 30 fps on the exact same game might be playable (if still somewhat annoying). Noticeably bad fps is not nearly as frustrating as generally good fps with a few points of noticeably bad fps mixed in.

Re:Any animator knows...

[Anonymous Coward]

If by 'supposedly' you mean 'definitely' and if by 'most movie theaters' you mean 'all theaters and even all motion picture production processes in recent years', then yes.

This is not true. Two theaters in my area (at least) have the ability to display other frame rates. I can think of at least one incident where they did display at a much higher frame rate (the movie was filmed at a high FPS to capture the real world effect and it was not reduced in the theater). One example would be IMAX HD which displays at an effective 48 FPS.

Re:Any animator knows...

[TheEvilOverlord]

If by 'supposedly' you mean 'definitely' and if by 'most movie theaters' you mean 'all theaters and even all motion picture production processes in recent years', then yes.

I'm sorry but that's not quite correct. I worked as a movie projectionist for several years, so I know this from experience. While 24fps works, and is what used to be used in cinemas, it is noticeably flickery. As with most advancements in cinema technology, they came up with a bit of a hack. While there are still only 24 frames of film per second, the projector shows each frame twice, giving an effective frame rate of 48fps.

Re:Any animator knows...

[cnaumann]

Avatar in 3D looked really jumpy to me, kind of like watching movement under a strobe light. I wish movies would move beyond 24FPS. The technology is there, but I am not holding my breath.

Re:Any animator knows...

[Hurricane78]

How weird... You contradict your whole argument the last paragraph.

Anyhow: That is an interesting idea: Why not make an engine that does render all changes since the last frame in the same manner? Not with the primitive motion blur though. Something more intelligent, that helps you predict things too.

Re:Any animator knows...

[Anonymous Coward]

yes!

the basic reason you can see the difference is because you are tracking something with your eye (following the circle over the screen).

If you keep your eyes focused at a fixed position on the screen, 24fps is more than enough.
As soon as you start tracking things, the difference between 30fps and 60fps is enormous.

Re:Any animator knows...

[MobyDisk]

I don't think the film motion blur entirely compensates for the low frame rate. IMHO, high-speed action sequences often look stuttery. there must be some unsampled time-slot while the shutter is closed and the film is advancing (or in digital, when the CCD is being read).

Also, CG films must artificially create the motion blur. I am not convinced that they actually bother to do this.

Re:Any animator knows...

[master_p]

I was trained to detect 60 vs 30 FPS on the Amiga: the games that took advantage of the Amiga's hardware were silky smooth, meaning 60 FPS, and the games that were simple ST ports were jerky, meaning 30 FPS or lower.

I loved the Amiga game Drivin' Force (http://www.youtube.com/watch?v=RrJDgYxgyNM) simply for this: absolutely stunningly smooth 60 FPS visuals. It looked a little bit like Sega's Powerdrift.

Re:Any animator knows...

[tlhIngan]

I've always noticed I've been quite picky about FPS in games. I once recruited a friend to help me to a blind test on myself with the help of quake3. He set the max framerate in game with com_maxfps and I had to say whether it was 90 or 120fps after playing for 30seconds or less.. I got it right 100% of the time.

Try it for yourself, you can do it with quakelive as well. I may have been taking in extra queues, the game mechanics are sort of tied to the frame rate, in fact on certain maps you could only make certain jumps within a certain FPS range. It also may have helped that I have a preference for 120hz CRT screens. I haven't repeated the test on my current 60FPS LCD.

I think it also depends quite a lot on the type of game or what's going on in the game at the time the frame rate dips*. If quick, precise reactions / controls are important, a dip in fps is just horrible

What's probably happening is that you don't notice the framerate, but you notice the change in input rate. In games, you typically have a game loop consisting of get_user_input, process_input, update_state, render_frame, update_screen, and repeat (waiting if framerate locked). Thus, the game's input "lag" varies with the framerate because the game only samples input once per frame.

Let's say the average framerate dips from 60fps to 30fps, that also means the input is sampled half as fast. For "twitch" games (like an FPS), sampling at half the frequency can make the whole thing feel a lot more sluggish, and human perception can note this.

Let's say our hypothetical game takes roughly 2 frames from the user hitting left, and the character moving left (because of double-buffering and other such things, including OS sampling of the input device). This means roughly that when the user hits a button and the user sees the reaction is around 1/30th of a second. Now the framerate drops by half to 30fps, which still has a 2 frame lag, but now the reaction is slowed to 1/15th of a second. Which is enough of a delay that users can notice.

A useful experiment to try is to create a program that reacts in an event-driven manner - the framerate is locked at 30fps or 60fps or other setting, but the input is sampled at a different, but constant rate. Then vary the display framerate and see if the user can tell any difference between 120, 90, 60 and 30fps, but keeping the input sampling rate at a constant. Then crank the framerate to 120fps, but halve the input sampling rate.

I'm just saying those claiming they need 120fps may not need 120Hz, just that games tend to sample at the display framerate because of the game loop, and framerate drops impact more than the displayed framerate - i.e., they impact input sample rate.

Re:Any animator knows...

[pjt33]

If by 'supposedly' you mean 'definitely' and if by 'most movie theaters' you mean 'all theaters and even all motion picture production processes in recent years', then yes.

If by "all" you mean "all US". In Europe it's 25Hz.

Re:Any animator knows...

[Hyppy]

I think that might be possible, but hardware is so far behind that it's unreal. Let's assume the most basic implementation, with minimal advances in hardware technology needed. The engine would have to take sample snapshots during the inter-frame period. 3 snapshots would provide a decent sample, especially if it was weighted against the previously viewed frame. The engine is now rendering 3 frames per visible frame. Add on top of that the processing time to merge the frames and create a plausible blur. With some ingenuity, I could see that taking about the same power as rendering 1.5 frames. So, now for every visible frame, 4.5 frames worth of rendering power is being used.

Your 90 FPS game just dropped to 20 FPS. Somehow I see that going over like a lead balloon. Crysis anyone?

Re:Any animator knows...

[geekoid]

That's cool to know.

So cool in fact I looked it up*.

http://en.wikipedia.org/wiki/Movie_projector

*code for "it sounded cool but I didn't believe it".

Sometime the do it three times.

Re:Any animator knows...

[ucblockhead]

I've seen 120 fps video running on a 60 fps TV and a 120 fps next to it and the difference was very noticeable.

Re:Any animator knows...

[alexhs]

The only think you proved is that your test is wrong.

Render the animation once in (say) 240 fps.

Then blend the frames to drop the framerate (so, blend 10 frames for your 24 fps rendering)

You should get different results.

However, as another poster mention, there is tracking, which will still make some difference.

I've read a long time ago (no source, sorry) that some neuroscientists observed that the brain processes about 20 / 25 images/second.

Now the content of the images is determined partly by the movement of the eyes, so practically, the higher the framerate, the better experience.

If you had high-end equipment, with a laser tracking your eyes movement and calculating the images taking into account those movements, 30 fps would be more than enough though.

Re:Any animator knows...

[FredMenace]

It depends heavily on the type of game and the type of player. For a purely visual experience that looks nice, 30fps is often adequate if the game isn't super fast-paced. However, for a fast action game where you may need to react and aim very quickly (turning the whole player around in an instant), framerate makes a huge difference.

Among other things, if you slew your aim point (point of view) around very fast, even with a fairly high framerate, the view will jump between different points of view, rather than moving smoothly. This makes it difficult for the eye to track, making gameplay more difficult and contributing to headaches and nausea. (Input lag also contributes to these same problems.)

As for latency, from the time you move the mouse to look around or aim at something, it takes at least the length of one frame before you will see it on screen, and up to two frames - if the input is received just as a frame is starting to be rendered based on previous input, it will have to wait for that one to finish before it can even start rendering the next frame, which reflects that mouse movement. It could be even longer if things go through several stages of a "pipeline" of actions before showing up on screen - there could be an additional constant lag on top of the inverse of the framerate. (This is the input lag the article talked about.)

In addition, as pointed out in the article, framerates are often variable, by a factor of 2 or more. So a game that averages 30fps may hit 15-20fps at times (if not lower, though most commercial console games are tuned to avoid this, and many are tested to ensure they rarely drop below 30fps). At 20fps, each frame takes 1/20th of a second to render, or 50ms. So the delay after taking an action before it shows up on screen could be at least 50-100ms (or more, if there's some pipelining), which is starting to be noticeable when you're flicking the mouse around quickly.

If you are playing online against other players, all this is compounded not just by the network, but by the other computers involved. Many actions, including anything that another player did, will be delayed by your control latency (in this case the time from receiving the network information to displaying it on screen), plus the round-trip network latency between you and the server, plus the delay due to the server framerate (how frequently it updates the global state, which might also be only 20-30 times per second - assuming there is a dedicated game server in between, typical for the Quake and Unreal Tournament series of games for instance), plus about the same delay on the other player's end (from the time they give some input to when it is sent to their network connection) plus their network latency to the server.

So even if you only have a 10ms ping time, if you, your opponent, and the server are all running at only 20fps, the total delay before seeing another player's action could be several hundred ms, or over 10x your reported ping time! In this case most of the latency is due to the computers and related to framerate, and not due to the network at all. If all these framerates were 100fps instead of 20fps, then the latency would drop to something closer to 50ms.

So the article that says that most reviewers and casual players (who are generally trying to appreciate what the game has to offer visually, not to compete intensely against other players online) don't notice framerate much is probably true. But the competitive players out there definitely do notice, and will shed any and all graphical amenities if it helps them boost their framerate (= reduce their latency). To them, a slow framerate is like attaching lead weights to a tennis racket.

Re:Any animator knows...

[Yvanhoe]

A big part of my former job was to make our AR tools work at 60 Hz instead of 30 Hz (Well 59.94 Hz and 29.97 Hz, these values have driven me crazy). You can definitely tell the difference at first look, we did it routinely up to the point that it is now blindingly obvious. A correctly displayed HD video at 60 Hz with a monitor synchronized at the same frequency is incredibly smoother.

The most evident effect is the presence of persistence of vision : at 30 fps, objects can almost not move faster than the eye can perceive, therefore you can perceive successive positions of an object in a sharp way. At 60 fps, it can be too fast and successive sharp images of an object will be perceived as having a motion blur. I think that using this effect and with a little training, one could differentiate between even higher framerates. People I know that tried 100 or 120 Hz monitors told me that the comfort was also very appreciable.

The motion blur picture

[Lord Lode]

Whoa, the motion blur image with the birds and the mountain is nice, what game is that screenshot from??!!1

Re:The motion blur picture

[H0p313ss]

Whoa, the motion blur image with the birds and the mountain is nice, what game is that screenshot from??!!1

Reality 1.0, very popular MMO with billions of users. Excellent graphics but the developers have been very unresponsive to bug reports.

Re:The motion blur picture

[doti]

developers have been very unresponsive to bug reports

Like, avocado seeds being too big.

Re:The motion blur picture

[readthemall]

Life is like a game, it can be hard, but what graphics!

I notice a difference from 30fps vs 100?

[Kungpaoshizi]

The difference is very minute, but I've noticed if you play an fps with no vsync, 30fps is ok, but if you're doin around 100fps, then you notice the tearing. I'm sure everyone knows what I'm talking about, but if you enable vsync, and then look at the difference, of course your monitor hz comes into play, but there is very slight fractions of a second you notice, such as dude lookin away from you, then all of a sudden he's looking at you (30 fps) But at max fps your monitor can support + whatever your vid card is kickin out, you don't just notice the guy looking away and then looking at you, you notice the guy looking away, and then the fraction of a second it takes him to TURN and look at you. You can see that turn. No vsync, and it was probably a tear, but it is still there, and having won a fps tournament, I can tell you it does matter.

Friend could see 57 vs 60.

[Maxo-Texas]

I had a friend who was bothered by anything less than 60fps.

The screen looked "stuttery". He would take a lower resolution to maintain 60fps.

We could verify this in one game with a built in frame rate command.

This is like the "myth of the g spot" post a few days ago. sheesh.

Re:Friend could see 57 vs 60.

[Icegryphon]

I would not doubt it for a second.
Many people can see the difference,
maybe not that small of an increment.
There are a few people like your friend who can probably see even more then just 3FPS difference.

Re:Friend could see 57 vs 60.

[Anonymous Coward]

That's not strange, a game running at 57fps on a typical monitor with a 60hz refresh rate will suffer from image shearing, as the frame being displayed changes partway trough the monitor's refresh cycle. This is quite noticable even at high frame rates, and the reason for using vertical sync in gaming.

Now if your friend could tell the difference between 57hz and 60hz, that would be quite special.

Re:Friend could see 57 vs 60.

[Anonymous Coward]

I bet he loved Monster cables and acoustically balanced wooden volume control knobs too...

Re:Friend could see 57 vs 60.

[man machine]

I think it's pretty obvious that you can see a difference between 60 FPS and 57 FPS on a 60 Hz screen.
57 FPS means that you will get a doubled frame 3 times a second, easily detectable by the human eye.
For optimal smoothness your frame rate should be constant and an even multiple of the display rate.

Re:Friend could see 57 vs 60.

[Knara]

I dunno about frame rates, but I can tell you that in the days of CRT dominance, if a monitor had a refresh rate of less than 85Hz it was headache city for me in just a few minutes.

Re:Friend could see 57 vs 60.

[Anonymous Coward]

So it seems likely that your friend was playing on an LCD and 57fps was less than the refresh rate of 60Hz. Some frames would stay up for 1/60 of a sec, some would stay up for 1/30 of a sec. That would look "stuttery".

Right?

Re:Friend could see 57 vs 60.

[Maxo-Texas]

CRT.
We were playing quakepong and some car/robot shapeshifting game.

Likely his CRT was a 60hz since 85 made many crt's whistle back then.

There was screen lock to the refresh signal so there were no partial sheering frame display issues (anyone could see that). What he could see was that three of the frames were on the screen for 1/30th of a second while the rest were on the screen for 1/60th of a second. It bothered him enough to take a lower resolution while to most of us anything over 30fps was "glass smooth".

This is the guy who at a full run, heard two of us shout his name from behind, jumped in the air spinning and caught a disk (instead of being hit in the back), continued to spin, landed facing forward, and threw the disk on to a target without even stopping (or violating the 3 step rule).

The one time someone shouted my name, I looked up and got hit in the face with the disk. His reflexes were very good. He was a tremendous athlete as a result.

Saying "Most people can't see the difference" would be more fair than "No one can see the difference".

Re:Friend could see 57 vs 60.

[Anonymous Coward]

I had a friend who was bothered by anything less than 60fps.

The screen looked "stuttery". He would take a lower resolution to maintain 60fps.

We could verify this in one game with a built in frame rate command.

This is like the "myth of the g spot" post a few days ago. sheesh.

That's because anything less than 60fps (with frame-sync) is 30fps, so if it says 57fps, that means one fifteenth of the time it is displaying the same frame on two consecutive retraces, or to put it another way, it is switching to 30fps 3 times every second.

Anyone that's read the OpenGL redbook (or any other decent text on computer graphics), ought to know framerates can only be integer divisions of the display refresh rate. I consider it better practice to either decrease the complexity of the scene or enforce a frameskip, so that the framerate is consistent, than to be switching framerates continously and arbitrarily.

120 and 240Hz TVs would like to have a word...

[Anonymous Coward]

I mean, all you have to do is convert a regular 24FPS movie to 120Hz and you can see the massive difference that framerate makes in the smoothness of playback. That said, I am generally comfortable with 30FPS when it's consistent. Little in gaming is more annoying than seeing framerates adjust from scene to scene. Try watching a movie like that...

Re:120 and 240Hz TVs would like to have a word...

[Nick Ives]

The reason films look nice on 120 or 240Hz screens is because you can show each frame exactly 5 (or 10) times. You don't have to do a pulldown like on 50 or 60Hz screens.

In addition...

[Anonymous Coward]

Can we edit the post to tell people that no matter what their FPS meter says that they cannot display more frames per second than their monitor can draw? It seems to be a widespread myth that a game showing '500FPS' is actually drawing those to a 60hz monitor.

Also... Isn't the minimum FPS for fluid motion something like 23.4? Not 30.

Movies at only 24/25 FPS are horrible

[nmg196]

Personally I get annoyed by the fact that although they've invented HD (woohoo) they're still shoving it out at only 24 or 25 FPS. To me, this looks really jittery! I wish they'd go up to 50FPS for HD.

Watching Avatar in 3D seemed to accentuate that problem. I'm not sure how they do the whole left/right thing in terms of projection, but it seemed to me that the left/right image was shown alternately and at nothing like a high enough speed for me to perceive it as fluid motion. Did anyone else notice this?

Re:Movies at only 24/25 FPS are horrible

[Anonymous Coward]

So every movie you've ever seen that wasn't shot on video tape is horrible? Yeah right.

Re:Movies at only 24/25 FPS are horrible

[Anonymous Coward]

Unless it is interlaced, then you might get lucky and get 48 to 50 half-frames per second. This is how television works and why sports broadcast are even watchable.

Re:Movies at only 24/25 FPS are horrible

[The_Deacon]

I personally didn't notice it -- I was more bothered by the left/right coloration artifacts from the Dolby3D setup. :) This surprised me, as Dolby3D is generally thought to provide a better picture than its main competitor, RealD, at least from the conversations I've seen.

Most likely you watched Avatar 3D in either Dolby3D or RealD, both of which alternate images between the left and right eyes. This occurs at a rate of 144 FPS (so each eye sees 72 FPS) for both systems. It has been said that the 24- or 48-fps input signal can cause stuttering during fast horizontal pans, which may be what you were seeing.

http://en.wikipedia.org/wiki/RealD_Cinema
http://en.wikipedia.org/wiki/Dolby_3D
http://www.avsforum.com/avs-vb/showthread.php?t=1085037

Re:Movies at only 24/25 FPS are horrible

[joggle]

Yes, I noticed it. I think it's more obvious when you're closer to the screen.

It was most obvious to me during fast motion, with objects near the screen (such as when he was flying on that creature for the first time).

I also wish they'd increase the frame rate substantially at movie theaters, especially for movies that have a lot of fast motion (like action movies).

The strange thing is I'm not a FPS snob with games. While I can see the difference between high and low frame rates, so long as the frame rate is OK (say always above 20 fps and usually above 30 fps) I have no complaints.

Re:Movies at only 24/25 FPS are horrible

[Anonymous Coward]

Watching Avatar in 3D seemed to accentuate that problem. I'm not sure how they do the whole left/right thing in terms of projection, but it seemed to me that the left/right image was shown alternately and at nothing like a high enough speed for me to perceive it as fluid motion.

The RealD movies display as a sequence of flashes:

left frame #1
right frame #1
left frame #1
right frame #1
left frame #1
right frame #1
left frame #2 ...

Basically each eye sees the image at 72 frames per second, but 2/3 of those frames are duplicate images. That's awesome unless your vision is faster than normal, in which case you see motion as alternately completely stopped and then moving.

Incredibly some people actually do have visual slow perception, but I can easily tell the difference up to at least 75hz and 120hz fluorescents have a 'thinner' quality than thousands-of-hz CFLs (60hz fluorescents make me physically sick). So Avatar was my first 3d movie, and it will probably be my last for a while. Once they stop shooting in 24 fps so that each frame is different it will probably look reasonable though.

Age-old confusion.

[SD-Arcadia]

The 30-fps-is-all-you-can-see myth was probably born of the notion that the illusion of continuous movement starts to set in around 25-30fps (in film for example). Therefore actually 30fps is the minimum you need rather than the maximum you can perceive.
I could tell in a glance the difference between 72fps and 100fps (both common refresh rates that translate to the max fps when v-sync is on) in Counter-Strike just by briefly moving the mouse to pan the scene.
This site has had the definitive explanation on this issue for a long time, along with many other useful faqs: http://www.100fps.com/how_many_frames_can_humans_see.htm

Re:Age-old confusion.

[Chris Burke]

The 30-fps-is-all-you-can-see myth was probably born of the notion that the illusion of continuous movement starts to set in around 25-30fps (in film for example). Therefore actually 30fps is the minimum you need rather than the maximum you can perceive.

I think it's more likely born of the notion that film gives a completely convincing illusion of motion that is not greatly improved by higher frame rates, because the process by which it is created automatically includes motion blur because it's recording continuous data, just broken up into 24 fps. Computer games display discreet moments in time, not many moments blurred together into one picture. That's why film looks smoother than computer games with 3 times the framerate.

Nevertheless, the illusion of continuous movement is apparent at much lower framerates than even film, even in a computer game. Quake's models were animated at 10 fps, and they gave a convincing illusion of movement, and you can probably make due with a lot less since the brain fills in so much. But it's not a completely convincing illusion, and neither is 30, 60, or even 100 when using static instants in time.

But the basic myth comes from the fact that film is so convincing and thus you don't "need" more... as long as each frame is a blurred representation of the full period of time it is displayed for.

Re:Age-old confusion.

[Psyborgue]

But the basic myth comes from the fact that film is so convincing and thus you don't "need" more... as long as each frame is a blurred representation of the full period of time it is displayed for.

Not quite. Film cameras, because of they way they work, max out about half of the time they are exposed for (180 degree shutter). 24fps is usually shot at 1/48 second exposure time per frame. The full time (a 360 degree shutter) would be far too blurry.

Re:Age-old confusion.

[bughunter]

You're getting close to the real reason that some (most? all?) people can detect a difference between 30 and 60 fps -- it's the artifacts of motion.

These folks are not claiming that their eyes can discern the difference in angular rates across the field of view in that regime of framerates. But they are picking up on visual cues that are a combined effect of object motion and eye tracking.

To understand this, first consider an immobile eye: as an object moves rapidly across the field of view, there is a perceived blur due to a finite integration time at the retinal cells. In low framerate systems this "exposure time blur" can be simply modeled by producing blur for objects in each frame in proportion to their rate of movement across the field of view. And it creates a reasonably accurate illusion as long as the eye is stationary.

However, as soon as the viewer's eye moves to try to track the object, the illusion is broken. The mind expects to register the object visually with more clarity, but there is none -- detail has been removed by the motion blur effect. Higher framerate systems are more robust in this regard, but no system moves fast enough to account for rapid eye tracking.

This is why some people are adamant that 30 fps (or even 60 fps) is not good enough, and vociferously deny "scientific" claims that they cannot discern framerate improvements beyond 30 fps. Because they can; they are detecting the tracking blur errors.

It's not that they can detect the actual frame rate, but their eye/brain coordination is perhaps better, or at least their brains expect more clarity when tracking moving objects - even if its just scenery motion. I would expect that for trained eyes (e.g., experienced videogamers) this effect is noticeable right up to the performance limit of the LCD/phosphor/plasma display in use.

Re:Age-old confusion.

[Chris Burke]

Ah, thanks, I didn't know that.

Re:Age-old confusion.

[malp]

But the basic myth comes from the fact that film is so convincing and thus you don't "need" more... as long as each frame is a blurred representation of the full period of time it is displayed for.

You can really see this behavior in '28 days later'. The screen goes dark and stuttery whenever a monster appears. All the cameraman did to create this effect is reduce the exposure time of each frame, so you see only a tiny fraction of the 1/24th second. This created an instantaneous snap-shot of the scene, instead of the motion blurred image we're used to. The reduced exposure time also had an added benefit of darkening the scenes.

Re:Age-old confusion.

[Anonymous Coward]

Quake's models were animated at 10 fps, and they gave a convincing illusion of movement, and you can probably make due with a lot less since the brain fills in so much.

Quake and Quake 2 used linear interpolation of the verticies to make the animations look smoother at higher frame rates.

Re:Age-old confusion.

[Chris Burke]

Quake 2 did, but Quake did not. This was quite apparent when making custom models that made too large a sweep of motion between frames -- the engine did nothing to interpolate, and your brain couldn't do the job either. Part of the reason for this is that the quake model format/object did not have a concept of animations. Animation was coded by manually changing frames in the quake-c game logic. Each 100 ms tick the game entity's think() function would be called, and it could set the current frame for the model as it chose, including randomly. So the engine didn't know what the 'next' frame was supposed to be until it was already supposed to be displaying it.

Quake 2 added in the animation concept. I think they also upped the basic logic/animation framerate to 20fps.

Re:Age-old confusion.

[Hyppy]

Well, the exposure time may have been off by a factor of two, but the explanation still holds water. 1/48 second exposure time multiplied by 24 frames is still a half second of actual motion conveyed per second. There is an effectively (Planck, blah blah) infinite amount of points in time that each 1/48 second image captures. Compare that to zero seconds of exposure time, with explicit points in time captured measured by the number of frames.

Re:Age-old confusion.

[calderra]

I thought the Slashdot community was smarter than this! Yes, you can tell the difference on a PC screen up to 300+ fps, if we're talking about a very quickly updating display where there's no motion blur, interlacing, etc. But the human eye can barely see anything past television's 24 fps because that 24 fps is motion-blurred and interlaced. Since modern video games can replicate at least the motion blur side, most people can't see much more than 30 fps when a developer bothers to include these effects. But all of this is a side argument. The REAL problem that virtually nobody tackles here is that games are now developed in 30fps because they must render in high-def resolutions. When you're trying to crank out 1080p with high dynamic range and AA and etc, you've got to trim the fat somewhere, and that somewhere is pretty universally framerate.

Re:Age-old confusion.

[Chris Burke]

The REAL problem that virtually nobody tackles here is that games are now developed in 30fps because they must render in high-def resolutions. When you're trying to crank out 1080p with high dynamic range and AA and etc, you've got to trim the fat somewhere, and that somewhere is pretty universally framerate.

You must be talking about consoles. PC games have been rendering in "high def" resolutions for years without being limited to 30fps. If current gen consoles can't handle more, then the solution is simply more horsepower and it will likely be solved in the next generation.

Any problem whose solution is "wait one Moore's Law generation" doesn't seem like a particularly tricky one.

Plenty numerous all right

[1u3hr]

the numerous advantages of a high framerate, and there's plenty of those.

Brought to you by the Department of Redundancy Department.

Psychic pain.

[hyperion2010]

From my own experience its not so much a specific constant frame rate as it is the fact that the frame rate fluctuates. If all game developers made sure that their games NEVER dipped below 30fps during major action (fps obvs) that would be fine, but when 30fps is the average between 45 and 15 fps we have a big problem. Heck, even a drop form 60fps to 30fps is noticable and disconcerting. I bet if you were to hook players up to a machine that could measure emotional responses you would find that sudden drops in framerate elicit strong negative responses which negatively effect performance.

Thanks...

[rickb928]

Now I can justify another $1000 worth of hardware to my wife, to play the same game I can get on a $300 console.

She gets it. I'm the Computer Guy. I know how it works. I know what is needed. I know how to keep her from not being able to play Farmtown. Or is it Fishville? hard to keep up with the Facebook privacy violations/games.

Ya gotta have priorities.

Re:Thanks...

[Nicolay77]

I have just bought about 40 games in Steam for just $100, in the Christmas shopping season.

That would mean about two games for a console.

However, I'm a bargain guy. My 21" CRT was about $25.

(All prices in USD)

You can see the difference between 72 and 100

[Anonymous Coward]

LCD framerates being capped have confused the issue, but on a CRT there's was more than one reason the minimum should be 72. Then one day a friend suggested bumping it to 100 for better gameplay (Quake). I thought he was nuts, but I tried it and dammed if I didn't start making shots that I had been missing previously. Or at least thought I was missing! The difference was quite notable.

While this is probably on the upper end, it does happen. Also this is after many years of intensely competitive play, which will develop specific skills. [No, I don't strafe around corners in real life.]

That said, games have been getting slower and slower over the years. Modern titles just aren't at that warp 12 pace now, which is probably for the best given the limiting factor of LCD panels.

3d

[StripedCow]

if you use one half of the 60fps for each eye, you have 30fps in 3d... that's probably one reason for the drift back to 30fps.

The 30fps myth

[asdf7890]

The 30fps myth is simply an over simplification. The eye+brain starts to naturally perceive movement at around 10fps, usually a little lower. Motion usually starts to appear smooth somewhere between 15 and 25fps though it depends on many factors other than just the framerate (smoothness of the frame rate, relative change velocities of objects (or parts thereof) in the image, absolute colour and tone, colour and tone contrasts within the image, the existence or not of dropped frames and other inconsistencies, ...).

People often take this (the "15 to 25fps" bit, ignoring the "depending on..." complications) as meaning there is no need to go above 30fps, and in many cases there probably is no need, but in a number of conditions a higher framerate can affect the perception of movement quite significantly especially for fast moving objects/scenes.

Re:The 30fps myth

[asdf7890]

Also, replying to my own post with bits I forgot before, things seeming smooth somewhere up to 25fps doesn't mean they don't appear smooth*er* at higher rates. And to complicate things further apparently our peripheral vision, which comes into play when sat close to a large monitor as many gamers do, is more sensitive to higher frequency changes than the rest.

I never understood why...

[JustNiz]

Nearly everyone these days uses LCD monitors that have a pathetic maximum of 60hz display at HD resolutions (I think because of DVI spec/bandwidth limitations, Whatever moron invented DVI needs to be shot because of that).
I still have an analog CRT monitor that supports much higher frame rates at HD resolutions which gives a very noticeable edge when playing twitch-games like Unreal Tournament.
I never understood why people claim framerates above 60hz are better when their monitor is only capable of displaying 60hz at the resolution they play at. The only difference that framerates above 60hz (i.e. vsync turned off) is obvious tearing. You're still getting an actual 60hz framerate because the monitor, regardless of what the PC is doing.

Re:I never understood why...

[kalirion]

I still have an analog CRT monitor that supports much higher frame rates at HD resolutions which gives a very noticeable edge when playing twitch-games like Unreal Tournament.

I really doubt the "noticeable edge" part.

However CRT images just look painful at 60Hz refresh rate, regardless of the frame rate. LCDs are just fine.

Re:I never understood why...

[Blakey Rat]

I still have an analog CRT monitor that supports much higher frame rates at HD resolutions which gives a very noticeable edge when playing twitch-games like Unreal Tournament.

Yeah right.

You have to prove statements like that, you can't just spit them out and expect us all to believe it. Let's see a double-blind test, you playing 10 games limited to 60 FPS and then playing 10 games at your usual FPS and average the scores.

Re:I never understood why...

[Jeremy Erwin]

( Whatever moron invented DVI needs to be shot because of that.)

Perhaps you could promulgate a different standard?

Re:I never understood why...

[stewbacca]

Pardon my ignorance, but is 60hz really the same thing as 60 frames-per-second?

Re:I never understood why...

[JustNiz]

yes

Re:I never understood why...

[stewbacca]

I thought fps was the rate the game could render the screens, and the hz rating was how fast the monitor could refresh...maybe I'm being pedantic, but it seems to me these are two different things.

For example, if a CPU/GPU/Game can only generate the game at 30 frames per second, but my LCD monitor has a 60hz refresh rate, those are two different things, right?

Re:I never understood why...

[JustNiz]

My understanding is you're basically right but that the differentiation between Hz and fps is based on colloquial usage rather than technical correctness. For example, you could correctly argue that a monitor has a certain number of frames per sec too.
Its probably my bad for my original post not differentiating the two more clearly.

Brightness

[ChrisMaple]

The brightness of the image and ambient lighting makes a difference. The more light that goes into your eye, the faster it responds. I run 1600x1200 @ 62 Hz interlaced, and sometimes I notice flicker. When that happens I close the shades, and the flicker goes away.

30 Fps myth

[ggendel]

There were a lot of studies done a long time ago, and there are some very accurate psycho-visual computer models of the human visual system. I had the pleasure of working with the Jeff Lubin model when I worked at Sarnoff Corp, which won an Emmy Award back in 2000.

The 30 fps requirement is not a fixed point, but depends upon a lot of other factors, including viewing distance, field of view, and lighting conditions. The reason that film operates at 24 fps is because it is expected to be viewed in a darkened room. When film is trans-coded for TVs, they have to modify the gamma for a normally lighted viewing area or it will look bad. NTSC TVs are interlaced, displaying 60 fields per second, even though the frame rate is 30 frames per second.

Bottom line is that this article should include the environmental factors under which this point was made.

The difference in framerate

[DeskLazer]

15 FPS vs 30 FPS vs 60 FPS. This is a visual representation. There are points made, however, that when you watch a movie, the image is "softened" and runs at a lower framerate [something like 24 or 25 FPS?] because your brain helps "fill in the gaps" or something of that sort. Pretty interesting stuff.

Re:The difference in framerate

[RedTeflon]

You mean to tell me Im subconsciously thinking when I watch TV filling in these gaps? No wonder my brain hurts so much all the time

Sorry, you lost me

[nobodyman]

As more and more games move away from 60fps *snip*

Hmm... I don't accept that premise, either on the PC (where midrange graphics cards can easily pull 60fps with any game on the market now) or on the consoles (where framerates are only going up as PS3 and 360 development matures).

I think that this article (or at least the summary) is a bit of a strawman. Most of the gamers I know recognize that good framerates are important.

Re:Sorry, you lost me

[pjt33]

I understood the summary as meaning "As more and more games have framerates better than 60fps I'm noticing lots of people saying that it's not really necessary to do better" (and presumably that the extra processing power should be spent on AI or what-have-you).

Re:Sorry, you lost me

[TheThiefMaster]

UE3 uses a 30fps cap on consoles. Quite a lot of games use UE3.

Re:Sorry, you lost me

[stewbacca]

"Good framerates" aren't important. "Good enough" frame rates are important. Anything below good enough is obviously bad, and anything above is not noticeable.

Same with audio...

[QuietLagoon]

Everyone says a "framerate" (i.e., sample frequency) of 44.1kHz is all that is needed. Yet many people hear better imaging, depth and transparency at higher sample rates.

Re:Same with audio...

[anup_at_mac]

Yet many people hear better imaging, depth and transparency at higher sample rates

Now, I may not have gold-plated Monster cables, but how do you hear better imaging?

Re:Same with audio...

[compro01]

That's a common misunderstanding of the Nyquist-Shannon sampling theorem. A sample rate of 44.1kHz gives you a Nyquist frequency of 22.05Khz. Any signal below that (given ideal filtering) will not fold and alias to a different signal. It gives nothing on the quality of the sampling and capability for that to be reconstructed to the original signal beyond that.

Re:Same with audio...

[LionMage]

He's talking about stereo imaging for an audio signal. I'm going to assume you figured out that he was making an analogy between sample rates for digital audio and frame rates for video (recorded or computer generated). Since his ears are the engaged organs in this example, "hear" is the appropriate verb. And the creation of a 3D soundstage is the whole point of using stereo, as opposed to having a single channel for sound.

As for what it means to "hear" imaging, it's really obvious to anyone who's ever listened to a high-end system with a musical sample they're intimately familiar with. Better imaging means you can more precisely locate instruments in the stereo soundstage -- at the very least, better localization in the horizontal plane, and if you're lucky, the "height" of the soundstage may seem to expand. Instead of getting a smear of sound across some general area, you can point your finger and say, "I hear the oboe coming from right there."

Obviously, higher frequencies are more directional than lower frequencies, so some musical instruments localize better than others.

And no, you don't need gold-plated anything to get this kind of performance, but it doesn't hurt to have gold plating on terminals to prevent corrosion. I've had cables with RCA connectors on them give a poor, noisy connection until I've cleaned off the oxidation.

I will say, though, that after listening to a $500,000+ system at a local HiFi boutique and comparing to, say, my current approx. $5000 system at home -- a difference of 2 orders of magnitude -- I know that I'm missing out on some of what the top-shelf system can reveal to me, but not enough to justify the massive difference in price. (The high end system was fed by Mark Levinson equipment, with two massive monoblock amps plugged into 220V mains feeding the speakers. Pretty outrageous.)

On the other hand, I can say that SACD definitely provides a superior experience to CD, as the GP indicated -- the effective sample rate for SACD is ridiculously high, even though the technology isn't directly comparable since CD uses 16-bit PCM and SACD uses 1-bit DSD. DVD Audio, not quite as much, but I have heard some impressive DVD-A demos at 192 kHz sample rate, 24 bit/ch sample size, which were almost holographic.

The bottom line is, whether you're talking about audio or video, more samples/sec or frames/sec is a good thing, especially with (in the case of video) a rapidly moving subject. This truth is why, for instance, "Doc" Edgerton at MIT pioneered high speed cameras and strobe technology to photograph bullets ripping through playing cards. OK, in the still photography case, you care about the inverse of frame rate, so to speak, but the principle applies; I do know "Doc" also did film work, and even won an Oscar.

Re:Same with audio...

[Quiet_Desperation]

"Imaging" I get. How does he hear transparency?! Must be synesthetic.

Re:Same with audio...

[SD-Arcadia]

..and yet no one can ABX it. unless you have Monster Cables (TM) ofcourse ;)

Re:Same with audio...

[HunterZ]

What does "signal below" mean? My understanding (which may have eroded over time) was that a digital sample of an analog signal can only accurately encode signal frequencies up to half of the sampling rate. Thus, if you can hear frequencies above 22050Hz then you should be able to hear the difference between a 44.1kHz sampling and, say, a 48kHz sampling of a source sound with frequencies above 22.05kHz.

Re:Same with audio...

[Anonymous Coward]

The problem with this is that 20kHz realy IS a limit of hearing.
Many studies have been conducted and the ones that are properly double blinded invariably say humans can't hear beyond about 20kHz.

There is a different issue tho, mainly plauging low and mid range products.
The thing is that for a certain sample frequency you need a filter (a socalled reconstruction filter) with correct behaviour for that particular samplerate.
Most products that can switch samplerates have only one filter and it will be optimal for only one samplerate.
Specificly, when the filter is designed for a higher samplerate it will sound worse at a lower samplerate.
Many people doing comparisons are unaware of this fact.

Another important factor is that filter design for a 44.1kHz (or thereabout) is nessesarily more constrained.
If one chooses a higher samplerate then it becomes easier to make a filter that will cleanly let through sound up to 20kHz.
The original idea of using higher samplerates is to ease the strain on filter design, and NOT to reproduce sound above 20kHz.

So it's all about the quality of the filter and not particulary about humans hearing above 20kHz!

However, things are completely different with vision and i can testify that my old CRT only gives a completely steady picture at or above 100Hz.
My guess is that this is close to the perceptual limit of human vision.

Re:Same with audio...

[Anonymous Coward]

But those many people apparently need to be hit with the clue stick called the Nyquist-Shannon sampling theorem.

Re:Same with audio...

[olau]

The bottom line is, whether you're talking about audio or video, more samples/sec or frames/sec is a good thing

No, it's not. You should never use more samples than actually needed. More samples means more data, which means slower processing which means you're wasting your cycles on something silly when you could use them for something better (like improving your other processing, or using cheaper/more low power equipment, or including more material, etc.).

The trick is to know how much is enough, of course.

Re:Same with audio...

[Anonymous Coward]

Except that in this case Nyquest theorm states that the original waveform is prefectly reproducable in the audible 20hz-20khz range. The difference you hear is an engineering problem, of how do you filter out the digital crap that results from DAC without affecting the audible range if there is very little room to work with.

Put a DVD-A on a DAC with an output stage designed for 44.1khz playback and you'll hear no difference. Run a 44.1khz signal through a DAC with little filtering on the output (say designed for 192khz) and what you hear is the result of artifacts of the DAC process.

Re:Same with audio...

[Idiomatick]

erm you are talking about something 1470x faster so the comparison is a bit unfair.

Studies show that everyone can see way above 60fps. Sample rates above 44.1kHz can only slightly change spatial awareness in highly trained ears, on good equipment, in a sound proofed room, sometimes.

Re:Same with audio...

[smaddox]

Actually, it is completely different. The mechanism for hearing is not like the mechanism for seeing. Hearing uses a fluid filled sac that oscillates when tapped by your eardrum. The cilia inside the sack then stimulate neurons. So, there is a step in between the absorption of the stimulus, and the firing of the neurons. This step has the effect of damping (or filtering) both very low, and very high frequency sound. It is true that some people can hear higher pitches than other people, but 20 KHz is generally considered the upper limit of human hearing.

The Nyquist–Shannon sampling theorem basically states that a signal with frequency no higher than 20 KHz is uniquely defined by a sampling at 40 KHz. Of course, this theorem only really holds when the signal extends infinitely long in time, but in practice - a sampling rate twice the highest frequency provides for very accurate signal recreation.

This is in contrast to human vision, in which neurons are directly stimulated by light. In vision, there is no stimulus frequency filtering (note that the stimulus frequency is seperate from the frequency of light, which IS filtered by pigments. This is what allows us to see colors).

Re:Same with audio...

[tangent3]

..and strangely enough, their capability of hearing "better imaging, depth and transparency at higher sample rates" disappear the moment they attempt an ABX Text.

Re:Same with audio...

[Anonymous Coward]

Only those who do not really understand the Nyquist Theorum say it is always enough.

Of course sampling above 44.1 kHz records more nuances; what matters is whether those nuances improve the listening experience or not. And that depends on many factors: the sound being recorded, the recording equipment, the playback equipment, and the ear and attention of the listener.

In practice, though, 44.1 kHz is usually "good enough."

Re:Same with audio...

[sahonen]

Even if you can hear pure sine waves above 22khz, most things that people actually listen to are a complex mess of frequencies. In content that people actually listen to it is difficult to pick up frequencies above even 16 khz or so, even if you can normally hear a dog whistle.

Re:Same with audio...

[Anonymous Coward]

I'm sorry to say, but all aural spatial cues perceived by humans are firmly within the hearable range, so 20kHz and lower.
The fact that you heared some cool demo's doesn't say anything about the samplerate.
All correctly conducted double blind tests show that people can't tell the difference between good CD's and SACD's, even when the test subjects think of themselfs as audiophiles.
The whole thing with higher samplerate devices is that they are non-optimal for reproducing lower samplerate signals.
When 44.1 was ruling there were a lot of cheap devices on the market that sounded better than many 96 or 192kHz devices these days.
But these new high samplerate devices are even worse at reproducing 44.1kHz signals so if you do a simple test on the same device with different samplerates the higher will sound better.
This is because of the reconstruction filter, not because of the samplerate!

About DSD, you can directly convert DSD to PCM and vice versa.
If you downsample the 1bit DSD stream to about 117khz you are left with 24 bit words, which is slightly more information than 96khz 24bit PCM.

No listening test ever showed that any content above about 20kHz can be perceived by humans.
Tests have been done with extremely boosted supersonics and people could not hear a difference.
It is a myth.

You could argue that there are tiny interaural timing differences that suggest that there are higher frequencies involved, but that is a misunderstanding of the nyquist-shannon theorem.
Yes there are tiny interaural timing differences that suggest shorter times than half the samplingrate, but they are in fact covered by the theorem.
The temporal resolution of a frequency is unaffected by the samplerate.

About sound localization, i have made binaural recordings at 44.1kHz 16bit and holographic precision is perfectly possible at these rates.
What is way more important for localisation is the reproduction of the material.

If you want to get your spatial cues delivered correctly then:
You need a controlled listening environment (that's why they have special listening rooms for these demo's)
You need a good pair of speakers.
You need to be seated in the right position.

These 3 factors actually contribute to your perception of the location of sound, in contrast to higher samplerates.

This is also why binaural recordings work so well, you listen to them on headphones and that makes for a nicely controlled listening environment (no reflections, no standing waves, each ear only hears what it is supposed to hear).

About the significance of a $500,000+ system, noone creating music uses them.
Therefore you are further away from what the artists intention was.
If you like the sound, great, but remember this is not what was intended.

Re:Same with audio...

[noidentity]

Entirely different, or rather, you're comparing something that's over 700 times the rate. Give me a video game and I'll tell you in two seconds with 99% reliability whether it's 30 or 60 FPS, and after a few seconds whether it's 30 or less than 30 (Zelda Ocarina of Time, I'm looking at you!). The real question is what sort of mood the rate sets; I think having the Zelda games at 60 might not look so great, but having something like Metroid Prime or God of War at 30 would be noticeably inferior.

Re:Same with audio...

[compro01]

By "signal below" I meant a signal without any frequencies above that line.

You understanding is mostly correct, but the definition of "accurately" for this purpose means only "will not alias to a different frequency". If you tried to record a 30kHz tone using a 44.1kHz sample rate, sans filtering, and then tried to play it back, that tone could reconstruct as a 14.1Khz tone. The prevention of that is as far as it goes.

As far as the accuracy of the waveform capture, it gets worse as you approach the Nyquist frequency. If you recorded a 22kHz sine wave tone at that sample rate, and then looked at it as a waveform, it would actually look like a somewhat screwy triangle wave instead of a sine wave, which sounds different. You can test and examine this for yourself quite easily in a program like Audacity by playing with the sample rate and generating various tones.

Re:Same with audio...

[Backward Z]

I was going to post this, but to expand on it, one should mention the roll off filter.

The idea behind the 44.1kHz number is the highest recorded frequency we're going to want to 20kHz, given that's typically considered the highest pitch (young) people can hear. So you end up with an extra 2.05kHz over that which is used as a roll off for the lowpass filter that's applied before encoding.

The argument for higher sample rates is that you can give the roll off filter a more gentle slope (or eliminate it completely if you're working at >~60kHz or whatever double the highest recorded frequency your physical gear/transducers can handle is). The gentler slope creates much less harmonic distortion than the steep 2.05kHz slope. Consider a 48Khz signal. Now I have double the rolloff space--4kHz. 48kHz sounds better than 44.1kHz not because it's recording more high pitched frequencies and that's what we're hearing, but because the low pass filter that's applied to prevent Nyquist foldover distortion has a more gentle slope which in turn creates less harmonic distortion in the passband.

This is the kind of shit the fascinates me. Get me started about noise-shaped dithering sometime.

120fps vs 60fps is like night and day

[Iyonesco]

For a high speed game like Quake even 60fps is totally unplayable and there's a massive difference between 90fps and 120fps. I consider 120fps the minimum for Quake and for that reason I continue to use a CRT. If you put my CRT at 120Hz+120fps next to a 60Hz+60fps LCD the difference is night and day and the LCD looks extremely choppy. You don't even have to do a side by side comparison and if you're used to playing at 120fps on a daily basis then you'll instantly see the difference when you see the game running at 60fps.

People who think you can't tell above 60fps have obviously never done any sort of valid comparisons because the difference is extremely pronounced. Research done by Sony found that "240Hz is the perception limit for the degradation of motion image quality for the human eye in following natural images" (Journal of the Society for Information Display Vol 15.1). I suspect there would be a noticeable difference between 240fps and 120fps but I've never had the opportunity to compare.

These comments are all in the context of playing Quake which is a very fast moving game so there is a large difference between each frame. If you play a much slower game then the difference between each frame will be significantly less, in which case 30fps might look absolutely fine. However, just because some games look fine at 30fps doesn't justify the whole "the human eye can't perceive above 30fps" idiocy.

Re:120fps vs 60fps is like night and day

[JSBiff]

"For a high speed game like Quake even 60fps is totally unplayable and there's a massive difference between 90fps and 120fps."

Which raises an interesting point - maybe the games shouldn't be designed to be quite that high-speed to begin with, or the games should be designed to allow them to run at the necessary higher speed? Certainly with a console game, the game should be designed for the limits of the platform, yes? The problem seems to be that game designers are designing games that *should* be played at 60fps (because they are designed to be high speed play), but then they are designing the graphics of the game such that the console can only run it at 20-30fps.

Take as a counter-example, the game Oblivion. That was a game which focused on having pretty killer graphics, but the design of the game is such that you don't really need high frame rates - reaction times do matter somewhat in Oblivion (it's sort of a cross between an FPS and an RPG), but the nature of the game is that reaction times (and thus framerates) do not need to be nearly as high as Quake, MW2, or any other first-person shooter. But, if you are designing a first person shooter, framerates should be a top priority, trumping graphics quality where necessary. A lot of people passed over Crysis when it came out a couple years back, because even though it has superb graphics, it just wasn't playable on their computers.

Re:120fps vs 60fps is like night and day

[fprintf]

To support your point about acceptable frame rates varying by game, there are those that think an acceptable frame rate in flight simulator is 18 FPS, and that it can dip as low as 10 FPS in highly "populated" areas like cities and areas with lots of scenery. The difference is that there is no need to react instantly in flight simulator.

I have a crappy computer so I have learned to adjust. I can only dream of 60 - 240 FPS as I play the "buy the 2nd or 3rd generation behind" hardware.

Re:120fps vs 60fps is like night and day

[chocapix]

For a high speed game like Quake even 60fps is totally unplayable and there's a massive difference between 90fps and 120fps. I consider 120fps the minimum for Quake and for that reason I continue to use a CRT. If you put my CRT at 120Hz+120fps next to a 60Hz+60fps LCD the difference is night and day and the LCD looks extremely choppy. You don't even have to do a side by side comparison and if you're used to playing at 120fps on a daily basis then you'll instantly see the difference when you see the game running at 60fps.

You may want to try one those new 120Hz LCD monitor when your CRT dies, or when you need the space on your desk. I really like mine (and like you, I can't stand playing Quake at 60fps, it actually hurts my eyes.)

I'd really like to try a double-blind test to see if I can tell 120Hz vs moreHz.

Re:120fps vs 60fps is like night and day

[jgtg32a]

Actually building off of your point, I adapted my play style in Crysis to take into consideration the bad frame rate, extensive use of cloaking sniping and suppressors. That made the game enjoyable and very play then again God help me when I was spotted, I would usually die rather quickly.

Re:120fps vs 60fps is like night and day

[DarrylM]

I'd really like to try a double-blind test to see if I can tell 120Hz vs moreHz.

I'd be interested in finding out how to do a double-blind, visual test. ;-)

FWIW, I fully agree with you and the GP about higher frame rates. I remember back in the day when I had the latest Need for Speed game (I think it was NFS III, which ran at 640x480x256 colours), and the original, EGA graphics IndyCar game from Papyrus (320x200x16 colours). The graphics quality was much lower in IndyCar, but my computer ran it at a much higher frame rate than NFS III. As a result I found IndyCar much more playable.

Re:120fps vs 60fps is like night and day

[Nick Ives]

Stroke your epeen much? People used to play Quake just fine at 30fps, and still do. The main reason why Quake feels snappier at higher framerates is because a higher framerate allows you to run a bit quicker and jump a bit higher. That's why speedruns are done looking at the ground!

Re:120fps vs 60fps is like night and day

[chocapix]

I'd really like to try a double-blind test to see if I can tell 120Hz vs moreHz.

I'd be interested in finding out how to do a double-blind, visual test. ;-)

Heh. It know it was probably a joke but, it's pretty easy actually. You can even do it by yourself. I did it to convince myself I could tell 60 versus 120Hz.

I wrote a shell script that generated a config file for quake that contained lines like "bind x set com_maxfps y" where x is 0 through 9, and y is randomly chosen between 60 and 120.
Then I loaded the config file in-game, and wrote down what I thought the framerate was for 0, 1, etc. and compared it to the generated file. I never got it wrong.
It's double blind because I never had to look at the config file before I completed the test. Oh, and don't forget to set cg_drawFPS to 0. ;-)

Re:120fps vs 60fps is like night and day

[logicassasin]

"For a high speed game like Quake even 60fps is totally unplayable"

that's crazy talk.

The original Quake was VERY playable at less than 30fps. I know I was a menace in Quake and ESPECIALLY Quakeworld while running a Cyrix 200MX w/32MB of RAM with Win95b.

Quake 2 was good at 30 and typically good under 30. At it's release, I didn't have hardware that would run it at 30 (intel i740 and Permedia2, AMD K6-2 300MHz), but had no problem playing it well.

Quake 3 was the first time I could actually say that, while 30 was playable, 60 or more would have been better.

Quake 4 is also good at 60.

The only reason I see to specifically target a sustained 60fps is while some areas of the game may be less graphically intense (resulting in higher frame rates), others, especially in the heat of battle, can tax the hell out of a video card and processor. If a machine running a graphically intense game can maintain 60fps even under the most extreme conditions, the game will be quite playable when being played "normally".

Re:120fps vs 60fps is like night and day

[metrix007]

What rubbish. I regularly play Quake on my laptop, which obviously isn't a CRT, and it looks anything but choppy, and just as good as a high priced CRT.

You people are as bad as the folk who buy gold plated audio cables.

120fps ftw

[Twinbee]

Really high frame rates (say 100fps+) from the SOURCE and display are the only way that we're going to get rid of flicker and all type of blurring for good. Even a perfect display technology which has incredibly fast pixel refreshing will still either flicker or blur (depending on how much black is inserted between each frame), *unless* the source is higher than around 60fps. Future OLED tv will be in this position soon.

Disregarding flicker and blur, 60fps and especially 120fps rates provide silky smooth graphics which really improves over the horrid 30fps rates everyone has had to endure for so long. It bugs the heck out of me that films still only use 24fps. Yes I know each frame is a motion blur of the past 24th of a second, but it would be much nicer to just have a higher frame rate to begin with.

2 is completly untrue

[Anonymous Coward]

it does not matter, if the game reacts in the first frame of 30fps or 60fps

The conclusion may be right, but...

[Improv]

It may be true that high framerates are a good thing, but the linked article is rubbish - the author's arguments are really very stupid.

Any engineer knows...

[gillbates]

That it isn't framerate which matters, but the color/luminosity delta.

Light impinging on the eye produces a chemical reaction; naturally, it takes some time for equilibrium to occur in the rods and cones. The greater the light change, the faster the reaction occurs. This means:

1. Large changes in luminosity or color (the eye is actually more sensitive to luminosity) produce a sensory result faster than small changes.
2. Conversely, small changes in luminosity take longer for the eye to recognize.
3. The end result is that the effective "framerate" at which motion transitions from jittery to smooth is a function of the luminosity difference between the moving object and it's background. For example, an Air Force study once found that pilots could identify an enemy plane flashed for a mere 1/220th of a second. It's rather easy to do, when the background is a bright blue sky and a plane is a dark black object; the difference in luminosity is considerable.
4. From my own personal experience, I know that a mere 10 milliseconds is enough for my eyes to detect an LED flash.
5. While 30 fps might have appeared smooth on older CRT monitors with low contrast ratios, today's high contrast LCD/LED displays are much better at producing luminosity deltas than the models of 2 or 3 decades ago. Thus, you actually need a faster framerate to produce smooth motion than you did 20 years ago.
6. While I cannot pick out individual frames at 60 Hz, I do notice the flicker and it is *REALLY* annoying when playing flight simulator games. The irony is that while the terrain and plane are modeled to a photo-realistic resolution, the flickering framerate won't let me forget I'm sitting in front of a computer, rather than in the cockpit of a WWII era plane.

Smooth gameplay is not just about framerate, but a decent framerate (i.e. > 60 Hz) is the essential foundation on which FPS and flight simulator games depend. Though it's only my personal opinion, a game with a low, or jittery frame rate is much less playable than a visibly pixelated one. The eye is very capable of approximating missing lines, completing shapes, etc... but much poorer at interpolating missing frames.

Controller lag is the biggest problem

[Artifex33]

The real problem with low framerate is controller lag. I had a copy of Unreal Tournament 3 for my PS3, which had the amazing distinction of allowing you to use a compatible keyboard and mouse combo instead of the regular sixaxis controller. As a die-hard FPS gamer who had been resisting an expensive PC upgrade, this was welcome.

Unreal Tournament 3 for the PS3 is pegged at 30 FPS. The result when used with a kb+mouse was horrible controller lag. It was as if the view angle attached to the mouse was on rubber band that would stretch during a quick mouse move and then snap back into position.

When I tried the sixaxis, the controller lag wasn't noticable at all. My best guess at this was because the joystick-controlled view had a finite acceleration, rather than from any hardware lag. The keyboard, mouse and the sixaxis were all bluetooth connected. Using the same mouse on a PC game playing Quakelive showed no signs of lag. The sixaxis just isn't capable of the whiplash movements that a mouse is, so it couldn't show the same responsiveness issue.

The kb+mouse combo was still an advantage, but for a PC gamer, it was crippling to adjust to the laggy feel.

I'll have to try out some of the PC games that end up in the sub-30 FPS range to see if I can reproduce the same feel.

Re:Controller lag is the biggest problem

[msu320]

...Unreal Tournament 3 for the PS3 is pegged at 30 FPS. The result when used with a kb+mouse was horrible controller lag. It was as if the view angle attached to the mouse was on rubber band that would stretch during a quick mouse move and then snap back into position.

Ruling out the bluetooth hardware has to be considered before making such a statement. I've experienced controller lag on my laptop+BT mouse setup in a multitude of situations.

Sometimes i was able to link it to the bluetooth driver, or when a game that had locked cpu affinity to the same core as what the driver was locked on. Other times I was able to link it to a battery that needed to be recharged.

Lots of evidence for higher frame rates

[pz]

I am a visual neuroscientist (IAAVNS). The standard idea of refresh rate comes from CRT based monitors where the image is drawn by a scanning electron beam. If you use an instrument to measure the instantaneous brightness at a given point on the screen it will rapidly peak as the beam swings by, and then decay as the phosphor continues to release absorbed energy in the form of photons. Different monitors have different decay rates, and, typically, CRTs that were designed for television use have pretty slow decay rates. CRTs that were designed for computer monitors typically have faster decay rates. If the decay rate were very very fast, then the hypothetical point on the screen would be dark most of the time and only occasionally very bright as the beam sweeps by on each frame.

As you can imagine this highly impulsive temporal profile is hard to smooth out into something closer to the constant brightness of the world around us. The human retina has an inherent dynamic response rate to it, but it's actually quite fast, and there have been studies showing clear responses in higher order visual areas of the brain up to 135 Hz. But standard phosphors used in CRTs have a little smoother response, and so at more-or-less 80 Hz, the brain stops seeing the flicker (at 60 Hz most people see flicker on a computer monitor). The exact refresh rate where perceptual blurring happens (so the flickering goes away) varies widely between individual, and with the exact details of the environment and what is being shown on the screen. More-or-less at 100 Hz refresh, no one sees the flicker anymore (although the brain can be shown to be still responding).

Contemporary screens, however, are LCD based (I'm going to ignore plasma screens since the field is still working out how they interact with the visual system). Making the same experiment as above, the temporal profile of brightness at a given spot on the screen will look more like a staircase, holding a value until the next frame gets drawn. This is a far, far smoother stimulus for the visual system, so a 60 Hz frame rate produces a perceptually far more flicker-free experience. That's why most CRTs at 60 Hz make your eyes bleed, while LCDs at 60 Hz are just fine.

Except that newer LCDs have LED backlighting which is no longer constant, but flashed (WHY? WHY? WHY? Just to save some power? Please, computer manufacturers, let *me* make that decision!), so the experience is somewhat more like a CRT.

So that's one part of the equation: flicker.

The other part of the equation is update rate, which still applies even there might be no flicker at all. Here, we have the evidence that the brain is responding at up to 135 Hz. In measurements made in my lab, I've found some responses up to 160 Hz. But the brain is super good at interpolating static images and deducing the motion. This is called "apparent motion" and is why strings of lights illuminated in sequence seem to move around a theater marquis. The brain is really good at that. Which is why even a 24 Hz movie (with 48 Hz frame doubling) in a movie theater is perceptually acceptable, but a 200 Hz movie would look much more like a window into reality. On TV you can see the difference between shows that have been shot on film (at 24 Hz) versus on video (at 30 or 60 Hz). Video seems clearer, less movie like.

For games, 60 Hz means 16 ms between frame updates -- and that can be a significant delay for twitch response. Further, modern LCD monitors have an inherent two or three frame processing delay, adding to the latency. As we know, long latency leads to poor gameplay. Faster updates means, potentially shorter latency, since it is a frame-by-frame issue.

So, just as with audio equipment where inexpensive low-fidelity equipment can produce an acceptable experience, while a more expensive setup can create the illusion of being at a concert, so too inexpensive video equipment (from camera to video board to monitor) can produce an acceptable experience, while a more expensive setup can create the illusion of visual reality.

Re:Lots of evidence for higher frame rates

[WaXHeLL]

Except that newer LCDs have LED backlighting which is no longer constant, but flashed (WHY? WHY? WHY? Just to save some power? Please, computer manufacturers, let *me* make that decision!), so the experience is somewhat more like a CRT.

The reasoning between flashed LEDs is so that manafacturers can advertise a greater dynamic range and color gamut. It's kinda like a pissing contest between manufacturers. However, many of those features can be at least partially disabled by disabling dynamic contrast on your set (which is highly recommended on the CFL backlit sets, as an overall dark scene with a very bright spot will cause some nasty image quality issues).

Also do note that the flashed LEDs do switch significantly faster than the LCD pixels update, so although there is flicker, it "should" be past the perceivable threshold.

Re:Lots of evidence for higher frame rates

[smellsofbikes]

For the record (as an ex-LED-backlight hardware designer) the LED's are waaay too bright to run full-out, both visually and from a power usage and heat generation standpoint, and the only good way to dim an LED is by cycling it on and off rapidly to approximate the desired brightness. The reason I say 'the only good way' is because LED's are constant-current devices and all the drivers I'm familiar with are all designed around that, so you can't just go varying the voltage to try and dim them: the drivers aren't really voltage devices.

With THAT said, I have absolutely zero idea why any sane LED driver dimmer would be anywhere near frequencies that any human could see. LED's can turn on and off in nanoseconds, so a reasonable dim signal should be in the kilohertz range, at least, not the 100hz range. It's *possible* to put a 100hz dim signal on an LED driver, but it seems really dumb to me.

Re:Lots of evidence for higher frame rates

[Anonymous Coward]

Thank you, thank you, thank you!

I've been wondering why soap operas looked 'more real' than prime time dramas in the 80s and 90s for like, well... since the 80s and 90s.

Re:Lots of evidence for higher frame rates

[jabuzz]

The light out put of an LED is mostly directly proportional to the current. There tends to be a small about of non linearity around the point when the LED first starts conduction, but only an idiot would operate an LED anywhere near this region. This is all relatively simple solid state physics.

You want the LED to be dimmer reduce the current. Now of course you have to keep the voltage above the forward voltage of the LED. However constant current circuits are child's play even for a physicist like me. If an electronics engineer can't do one then they deserve to be sacked.

Re:Lots of evidence for higher frame rates

[pz]

Right --- LEDs are controlled by current, not voltage (unless you really are a glutton for punishment and want to design the hard way). I haven't looked at backlight driver circuits that are available, and while PWM is an easy way out, there's no inherent reason designs must be that way, or at least couldn't include some filtering. And, as you point out, there's really no reason at all to not be driving the PWM at 4x the frame rate for a 60 Hz update (or 8x for that matter). There is good reason, however, to synchronize the frame updates and the backlighting, if the backlighting must be a strobe design.

Same goes for the idiot designers who make LED brake light systems for modern cars refresh at 60 Hz (or lower!). No reason that can't be cycling at 1000 Hz to avoid flashing trails.

Re:Lots of evidence for higher frame rates

[AK Marc]

Also do note that the flashed LEDs do switch significantly faster than the LCD pixels update, so although there is flicker, it "should" be past the perceivable threshold.

I haven't tried with the LED LCDs, but the car lights want to make me throw up. Most LED cars seem to have the rear red marker lights also as the brake lights, and they run at reduced intensity when they aren't doing brake duty. They choose to flicker them, and they have to be flickering them at a very low rate, because they make me want to throw up. It's like someone's driving around brake light strobes. Can someone ask Cadillac (one of the worst, since they have the largest real estate at the back being the LEDs) why they hate humanity?

Re:Lots of evidence for higher frame rates

[purplepolecat]

On TV you can see the difference between shows that have been shot on film (at 24 Hz) versus on video (at 30 or 60 Hz). Video seems clearer, less movie like.

This is more likely because when you show a 24fps movie on 29.97fps TV, it has to go through a framerate conversion process called Telecine, which (grossly oversimplified explanation) breaks each frame up into 2 fields (odd & even numbered lines), repeats every 1 in 4 fields, and plays the result out at 59.94 fields per second. (See http://en.wikipedia.org/wiki/Telecine#2:3_pulldown for detailed explanation)

The result is that any moving objects look jerky, especially for smooth horizontal movement, because one of the fields will appear to stand still every 1/12 second. It's a product of the framerate conversion process, not because the source material has 20% lower framerate.

Re:Lots of evidence for higher frame rates

[smellsofbikes]

In the case of all the drivers we make -- which, now, specifically include car lighting -- the only brightness control we provide is a dim pin that accepts 0 or 5 volts, telling the driver to turn off or keep running. As such, PWM is the only way to dim them. It's possible to drive an LED using other circuitry but a closed-loop switcher is amazingly efficient compared to most alternatives, and when you consider both cost and efficiency, they're pretty much unbeatable.

With the LED brake systems, again, that's just poor design on the part of the automotive engineers. All the parts I know of can take dim signals in the tens of khz so there is no reason to use low refresh rates -- and in that case, it's even dumber because if you're working with a fixed lighting value, you change the sense resistor on the switcher and run the LED's at a reduced current, without bothering to dim them. Only if you want to vary the brake brightness with ambient lighting would it make sense to play with dimming brake lights, and while that's a really great idea, there aren't any cars that currently offer it (to the best of my knowledge) except possibly high-end Audi's.

Re:Lots of evidence for higher frame rates

[Anonymous Coward]

Tell that to the auto manufacturers. Some crappy cars out there have their tail lights set blink in the tens of hertz. It looks like there are a dozen cars about every time you scan your eyes around, which is no problem for the Chatty Cathies that aren't looking at the road in the first place. For anyone that does actually put effort into driving though, it's a dangerous mirage that should have never been make street legal.

Re:Lots of evidence for higher frame rates

[sowth]

Well, any device I know which converts a voltage into a specified current uses power. Pulse width modulation allows the circuit to deliver nearly all the power to the LED.

In the ideal approximation, the device delivering the current will have either 0 volts across it and full current in one state. (0V times any Amps=0 Watts) The other state will have all the voltage across it, but no current flowing. (any V * 0 A= 0 W) All the power used goes to the LED.

Re:Lots of evidence for higher frame rates

[batura]

I'm surprised that as an ex-LED-backlight designer, you state:

The reason I say 'the only good way' is because LED's are constant-current devices and all the drivers I'm familiar with are all designed around that, so you can't just go varying the voltage to try and dim them: the drivers aren't really voltage devices.

LED are Light Emitting Diodes. Diodes function as a switch- once you reach and surpass the 'turn on' voltage, the brightness of the diode remains constant. The only option is to apply the voltage to the diode as a wave.

Re:Lots of evidence for higher frame rates

[Anonymous Coward]

LED are Light Emitting Diodes. Diodes function as a switch- once you reach and surpass the 'turn on' voltage, the brightness of the diode remains constant. The only option is to apply the voltage to the diode as a wave.

Don't try to play an expert if you aren't one. Diodes don't work the way you think they do. Look up the V-I curve in a diode datasheet; they're not the perfect all-off or all-on voltage switches you seem to think they are. And the light emitting kind are not made of magic fairy dust so they can emit the exact same brightness no matter how much current is flowing through the diode (in reality, brightness is proportional to current, though I'm not sure if it's actually a linear relationship).

When the OP said they're constant current devices, that was engineering shorthand for "LEDs have a maximum continuous current rating based on their ability to dissipate heat and environmental specifications". See, not all of the energy delivered to the LED goes into making light; some of it becomes heat. So it's very possible to overheat the diode and thereby destroy it. The ideal power supply for an LED is a constant current supply, set to deliver the rated continuous current (or less if dimming is desired).

True constant current supplies are (relatively) expensive and typically not very power efficient. There is a very common approximation of a CC supply used in indicator LEDs and the like: just throw a current limiting resistor in series with the LED. This, however, is very inefficient since a lot of power is wasted as heat in the resistor. These are the reasons why waveforms are often used to drive LEDs. It's not the only option, just a good one for a large number of designs.

Re:Lots of evidence for higher frame rates

[noidentity]

Except that newer LCDs have LED backlighting which is no longer constant, but flashed (WHY? WHY? WHY? Just to save some power? Please, computer manufacturers, let *me* make that decision!), so the experience is somewhat more like a CRT.

Even an LCD with zero lag (i.e. 0 msec GTG performance) and a backlight that's constantly on looks blurry when things move and you follow them with your eye. An LCD with 0 msec GTG and an LED backlight that pulses ONCE per frame will not have that blur as you follow the object with your eyes. This is similar to how a CRT has much less blur for the same situation, because as you note, the phosphor is only brightly lit for a very small part of the refresh.

When you follow the moving object with your eyes, your eyes move smoothly, even though the object is jumping to a new position every 1/60 second. If the object only flashes for a short time at each new position, the flash is always at the center of your vision. If the object is shown for each entire frame, it gets smeared across the center of your eyes, since your eyes are constantly moving during the entire frame. Thus, until LCDs flash the backlight for a short duration only ONCE per frame, there will be motion blur for human viewers (unless the viewer fixes his eyes on a point and never moves them).

Re:Lots of evidence for higher frame rates

[noidentity]

For the record (as an ex-LED-backlight hardware designer) the LED's are waaay too bright to run full-out, both visually and from a power usage and heat generation standpoint, and the only good way to dim an LED is by cycling it on and off rapidly to approximate the desired brightness.

I don't follow. Isn't an LED's brightness based on current, and the heat dissipation based on average power? Thus, you could run an LED at 20 mA continuously, or at 40 mA pulsed at 50% duty cycle, and have the same average brightness and power. Of course the latter will probably look brighter to a human, since I understand perception to be based partly on the peak brightness. I think it's why LED tail lights run pulsed instead of continuous.

There IS of course the other issue of the driver being much easier to make using a variable duty cycle rather than adjusting the drive current, I'll grant that. With variably duty cycle, you can use a saturated transistor, so it's either fully on or completely off, thus little dissipated power due to resistance. And generating the pulse drive waveform is trivial, and easy to adjust the duty cycle of.

Re:Lots of evidence for higher frame rates

[Agripa]

When the OP said they're constant current devices, that was engineering shorthand for "LEDs have a maximum continuous current rating based on their ability to dissipate heat and environmental specifications".

LEDs are considered constant current devices because their forward voltage changes significantly with temperature and between units and they present as a rather low impedance load. All of these things make for more convenience if driven with a high impedance source (current) instead of a low impedance source (voltage).

LEDs are rated based on current because that is how they are expected to be used.

True constant current supplies are (relatively) expensive and typically not very power efficient.

What?

Take a standard buck regulator with a voltage divider on the output for feedback sensing and remove the top resistor and ground the load into the remaining resistor. With some minor adjustments to the frequency compensation, you now have a constant current buck regulator of roughly the same efficiency if you lower the sense voltage. The output inductor current becomes the current through the load. If the inductor is sized for the regulator to operate in continuous conduction mode, then the output ripple current can be made arbitrarily small within reason.

shutter speed isn't fixed

[Anonymous Coward]

The shutter speed is independent of the framerate. Film cameras don't have a fixed 1/24th second shutter speed, just like a video camera doesn't have a fixed 1/30th (or 1/60th) second shutter speed.

Double and Triple buffer

[Anonymous Coward]

Sure it is hard to tell the difference between 30 and 60, but in a very high speed game FPS or fighting game, or even your super mario type people feel the difference.

This is compounded by the fact that almost all games are double or triple buffered.
For example double buffering:
Frame 0 = Computer logic is here, think of things like I move the controller or mouse. Also draw stuff to the back buffer.
Frame -1 = This is the frame you actually see

For triple buffering:
Frame 0 = Computer logic is here, think of things like I move the controller or mouse. Also draw stuff to the back buffer.
Frame -1 = This is waiting in the off buffer
Frame -2 = This is the frame you actually see

For double buffering plus deferred draw(many games do this)
Frame 0 = Computer logic is here, think of things like I move the controller or mouse. Also schedule stuff to the to draw.
Frame -1 = This is being drawn to the back buffer
Frame -2 = This is the frame you actually see

For 30 fps the delay from what you do to the game controls to what you see is usually 100 milliseconds.
For 60 fps the delay from what you do to the game controls to what you see is usually 50 milliseconds.

Showscan

[davidjohnburrowes]

fwiw, the reports I read of folks that watched showscan movies ( http://en.wikipedia.org/wiki/Showscan ) 20+ years ago overwhelmingly said that the higher framerate gave the films an level of realism that they'd never seen in films before.

Crysis

[Anonymous Coward]

and here I thought investing in hardware to run Crysis at 60 fps was a waste of money...

No they aren't! Movies at 60+ fps are horrible!

[Anonymous Coward]

Movies would lose their dreamy movie-like quality at higher frame rates and would look like they were filmed on cheap video cameras. Most HD TVs today have hardware to generate more than 60 fps, and it's usually turned on by default. It makes movies look horrible! It's great for sports, but terrible for dramatic content.

Newer TVs

[Enderandrew]

There are plenty of TVs on the market now that take a 30 FPS source, and then create frames in between. Many LCDs are offering 120 HZ and 240 HZ models. I've heard some new Plasmas are bragging about 600 HZ.

Does it matter if a game only provides 30 or 60 FPS, if the TV is actually refreshing at 240 HZ?

Re:Newer TVs

[gringer]

If you have interpolated frames, then you need at least two frames (at the source frame rate, e.g. 30 FPS) in the buffer before it can be put on screen, making input lag a bit of a problem.

If you were older...

[DrYak]

I manually compiled moves learned while playing and discussing Mortal Kombat and sold them for $3-$5 a piece.

If you were an adult at that time, you would start suing everyone putting up Geocities page with the same purpose on the internet for free, on the grounds that they are distroying your business model.
and your name would probably end with "AA".

Re:If you were older...

[asills]

Luckily for me Geocities was about 1.5 years after that. I was able to fund many hours of MK and MKII arcade gaming for about 2 years before most people even really knew what the internet was.

Jade Empire

[kalirion]

I recently bought Jade Empire on Steam, and when I started it I could tell right away how choppy everything was. Seems by default it runs at 30fps. Tinkered with the config file to unclamp it to 60, and everything became much smoother.

Re:Jade Empire

[VoltageX]

Replying so I can remember to do the same when I get around to installing it. What kind of machine are you running it on?

Re:Jade Empire

[kalirion]

Core2Duo 3GHz, 8800GT, WinXP Pro. I saw somewhere that disabling the clampFPS property can lead to some errors though. Haven't noticed anything myself, but only played for about 15 min so far.

Most film cameras don't have a 'shutter speed'.

[Animaether]

more accurately - most film cameras don't have a notion of a shutter 'speed'.

The film roll still goes by at 24fps, but the actual shutter is a wheel. That wheel can have various sizes of gaps (to increase/decrease exposure *time*) and sizes (to produce specific motion blur effects; e.g. an object leading its own motion blur path requires a small shutter opening at first, ending in a large shutter opening). You use fairly sensitive film and a small shutter gap, and you'll get nearly motion blur-less shots like that of Saving Private Ryan (watch explosions in that film and every speck of dirt that gets thrown about appears almost razor-sharp; some find this objectionable). Heck, you can even expose twice per frame if you want to get all experimental and stuff.

That said.. you can't - short of electronic shutters - expose for -more- than the film's fps, though. A bit under 1/24th of a second is the most you'll get (that 'bit' being required to transport the film to the next frame).

Anyway.. wiki: http://en.wikipedia.org/wiki/Rotary_disc_shutter

Re:Most film cameras don't have a 'shutter speed'.

[Anonymous Coward]

A much better example of this is the movie "Gladiator". I fucking hate that movie because of the total lack of motion blurring. If I pause the movie, yeah it looks great, but when I play it all I see is flicker. It's choppy as hell. It makes my eyes hurt and I can't follow the action.

Re:Most film cameras don't have a 'shutter speed'.

[Wayne247]

And this is why Slashdot remains to me the best site to read news and discussions about technology-related subjects. I'm here reading comments about framerate in videogames when all of a sudden, a group of people spontaneously burst into the technicalities of cinema cameras and how the motion is captured on film.

This is FASCINATING stuff, thank you for that. You're already +5 Informative, but I'll pretend-mod you to +6.

'High" Framerates only matter

[Anonymous Coward]

IF the display can actually display them (most display 60fps). I agree with TFA in that it compares 30fps to 60fps, a lot of people don't consider 60fps to be a 'high' framerate.
I constantly hear arguments that some people want 120 fps or higher, on their LCD monitor that can only display 60 fps, which is not reasonable.

Appropriate news icon...

[yakumo.unr]

as Quake players have known this for over a decade..

Good animators know that motion blur matters

[Boss Sauce]

A film camera shooting at 24fps typically has an exposure of 1/48 of a second. The other 1/48 of a second is used to move the exposed bit of film out of the way and to position the next bit of film in the gate for exposure. This is referred to at a 180-degree shutter, since the shutter is open for 1/2 of the 360 degrees of the camera movement's rotation. When the film is shot overcranked-- at 48fps or 72fps or whatever for a slo-mo effect-- the shutter is still usually 180 degrees, so the motion blur looks about the same when played back at 24fps.

Your test doesn't indicate as much about frame rate as it points to flaws in your rendering technique.

With accurate motion blur, your viewers would have a very hard time telling the difference between 24fps, 30fps, and 60fps.

Re:Good animators know that motion blur matters

[Monkeedude1212]

With accurate motion blur

This is where the problem arises, that motion blurring shouldn't be something the renderer has to account for - motion blurring isn't the physical object bluring itself out, its the way our brain reacts to the fast change in light.

When this is attempted through post rendering processes it can often times look too overdone, I am put off by racing games because they always attempt this blur effect when you go fast, and in my mind I can remember times when I reached 150 Clicks and it didn't blur that badly.

So until we derive an algorithm that -accurately- handles motion blurring, framerate matters.

Games does have to be realistic

[Anonymous Coward]

The problem with modern computer games is that they are trying to be too realistic. Being distinct from realtity is the true value of ant game.

Outside Looking In

[DynaSoar]

I'm a neuroscientist that covers sensation and perception and its bidirectional interaction with cognition, particularly attention. I've got comments and questions and very few answers after reading this. I'm seeing a lot of things stated as facts that I've never heard of before. Some of them make sense, and some don't. Some of them are correct, some not, and many more than the others combined I have no experience in and can't say. Those seem to be well supported, or at least well known, particularly among those who've obviously done their homework. I can find references to these among the publications (like ACM) that are most applicable to the field in question, but I can find precious little in my customary pubs and books. That's not to say the stuff in the technically oriented pubs is wrong, just that some may not be covered much (ie. 'not of interest') in my field. My field is very cautious about experimental evidence, but I suspect in gaming's perception area there are common knowledge kids of things that came from hear say (we have many of those in rocketry too). It might do well for both fields to compare works.

What catches my eye at first is this "myth". As stated it's overly simplistic. Which humans' eye? Some have different reaction times. Those who could probably detect 30 fps discontinuity are those who see the TV screen jiggle and waver when they chew something crunchy while watching (you know who you are, here's a place to own up to it). What part of the visual field, central or peripheral? They operate differently. Jittering or blurring of objects attended to or not? Betcha it happens more to those not attended to, but that's not noticed for the same reason (hypnosis can bring that out right nicely). And how is it frame rates matter when the visual system evolved as a constant flow analog system? If a phenomenon that shouldn't make a difference does, and that frame rate is strictly due to technical considerations, how do we know that a variable frame rate might not give even better results? Since the visual system does not have full-field frames that refresh, why should artificial presentations? Why not present faster moving objects at a high change rate, slower moving at a slower rate, more or less a timing equivalent to some video compression techniques? Some of this makes good sense from my perspective, some appears goofy but may not be, and some clearly is whack according to well supported experimental evidence from my side, not sure about yours.

Here's an interesting one, apparent motion from blurring, occurring at the retina, ostensibly due to 'reaction time' of light receptor cells (rods and cones). I can see how this might occur. But if it's a time lag that causes blurring, everything should be blurred, because the layers of cells of different types in the retina between the receptors and those firing down the optic nerve operate strictly by slow potentials -- there's not a 'firing' neuron among them. Or, if their processing, though slow, accounts for motion and compensates, preventing adding to the blurring, how can that be used to increase apparent motion?

A last point which I'm fairly certain isn't covered in gaming and graphics presentation because very few know much about it and we don't understand it well: 10% of the optic nerve is feed-forward, top down control or tuning of the retina and its processing. Motion perception can be primed, can suffer from habituation, and has variance in efficacy according to several factors. What cognitive factors have an influence on this, and how can that be used to improve motion perception and/or produce motion perception that's as adequate as what's being used now but requiring less external computational effort because internal computation is being stimulated.

It's probable that both fields have things of interest and use to the other, including things the other isn't aware of. I've said much the same following another article on a different subject. From this one I can see it's probable there's a few peoples' careers worth o

Re:Outside Looking In

[ratboy666]

No. Because "more fps is better". Helps AMD and nVidia sell new graphics cards, and helps sell new systems to gamers. (similar to "horsepower", and "ping time", "fps" is a single brag-measure).

And, because it is easiest to generate a game using a generate/event loop. The technique was described in

James D. Foley , Andries Van Dam, Fundamentals of interactive computer graphics, Addison-Wesley Longman Publishing Co., Inc., Boston, MA, 1982 (http://portal.acm.org/citation.cfm?id=6684&dl=GUIDE&coll=GUIDE&CFID=69932769&CFTOKEN=26597849)

and is still used to this day.

Personally, I find "twitch games" hopelessly boring, but there is some literature that playing these games improves (in some sense), the human visual system. http://www.rochester.edu/news/show.php?id=3342

I don't have a subscription to "Nature Neuroscience", and I haven't reviewed this myself (not my field). The brief synopsis indicates an improvement in gray-scale resolving, but I don't know of any supporting experiments yet.

Re:Outside Looking In

[DemonThing]

This is the paper. http://www.bcs.rochester.edu/people/Daphne/Li_NN.pdf

Re:Outside Looking In

[geekoid]

No. Too expensive and many issue in neurology can't really be solved with anythint else then faster is better.

As you are well aware, people make decisions before they are cognitively aware of what's happening.

Since you are the expert, maybe you can think of a way they could use your skills. If so, start your own business and sell a product. If you could create a package that plugs into a game dev. environment you will probably make quite a bit of money from licensing. You could probable grab 50K per title.

Quite frankly, softgames could use more experts in the fields they create. Physics, Environment, HID.

Re:Outside Looking In

[geekoid]

Boring? or just not interesting?

Boring indicates that you have mastered them and they no longer have anything to offer you. If you think that's true I'd like to take you on in TF2.

Oh, and the faster is all you need to sell aspect of computers is quickly fading away because the hardware has outpaced the development. And Moore's law is quickly finding it's practical end.

Re:Outside Looking In

[Anonymous Coward]

Those who could probably detect 30 fps discontinuity are those who see the TV screen jiggle and waver when they chew something crunchy while watching (you know who you are, here's a place to own up to it).

I thought that almost everyone saw that effect? To me, it's something very obvious. Old NTSC CRT TVs are flickery because they only refresh at 59.94 fields per second. I bet PAL CRTs are even worse.

Re:Outside Looking In

[DynaSoar]

Those who could probably detect 30 fps discontinuity are those who see the TV screen jiggle and waver when they chew something crunchy while watching (you know who you are, here's a place to own up to it).

I thought that almost everyone saw that effect? To me, it's something very obvious. Old NTSC CRT TVs are flickery because they only refresh at 59.94 fields per second. I bet PAL CRTs are even worse.

I thought the same. Then I mentioned it in one of my lab classes and about a quarter of the students knew what I meant. I tested them, and those saw it, the others didn't. Same result after explaining it to them, trying to prime them to detect it. Good solid results in only some people, very reliable and unable to bias easily. Got to be in the wiring. Maybe a kind of speed of perception with an action similar to frame rate but not operating discretely as frames.

Re:Outside Looking In

[DynaSoar]

I can't say I can understand or even parse much of that, but I can follow one part and have an answer.

I most certainly can think of ways to apply my knowledge, but that does little good if it doesn't solve a problem a developer has. While I could get behind helping someone working on some product like a game or graphics engine by solving some problems to improve perceptual qualities, I have no interest in developing any sort of product for game development any more than most game developers have an interest in pursuing a project in neuroscience.

Re:Outside Looking In

[DynaSoar]

This is the paper.

http://www.bcs.rochester.edu/people/Daphne/Li_NN.pdf

Cool, thanks.

See, this is one that I find to be trivial, in that focused attention improves perceptual acuity, and maintaining attention on that task trains the visual system to maintain the acuity. The same effect was seen in radar operators in the 50s. Using games is fine since they're commonly available, but a simple program to do the same would be easy enough to create and could be made adaptive to the specific result desired.

But that's in no way a criticism. I find it trivial because my field deals with this kind of thing all the time. Gamers are just learning about it. This is why it'd be good to work together -- to prevent reinventing the Lunar Lander.

And just so nobody think I'm hiding behind fake fellowship to fly my own flag, I'll admit to helping on a project to test potential pilots and train those accepted into training, using long term focused attention to the point of exhaustion, combined with manual dexterity, reaction time and problem solving having to do with geometric form and color. The Hungarian air force still uses it. It's Tetris with eye blink and EEG recording indexed to the game speed and scoring.

Re:Outside Looking In

[Ginger Unicorn]

it happens to me very very infrequently - could tiredness or other factors affect your perceptual "frame rate"?

Re:Outside Looking In

[Kattspya]

I remember that my entire field of vision used to shake due to chewing or fast shaking of the head when I was younger. Nowadays the most I can get is a slight jitter in the periphery of my FOV.

I've always interpreted this as resulting from "practice" of my inbuilt "anti-shake system" rather than some sort of degradation of perception. So my question is how do you know if day to day jitter is a result of (too) good perception or a deficiency of the "anti-shake system" or are the two things the same?

It also matters in input

[caywen]

I think it's a foregone conclusion that 60fps looks smoother than 30fps. However, it also matters for game input for fast action games. 30fps means that you have a maximum latency of about 33ms, or an average of about 16ms between when something happens and the time you could possibly observe it. 60fps cuts that in half, and removes about 8ms of latency on average. Two players being equal and having equal network latency and bandwidth, the one running at 60fps will edge the other out by a small but significant margin.

Re:It also matters in input

[stewbacca]

Is it really a foregone conclusion that 60fps looks better than 30fps? Does Super Mario World (to use the author's own example) really look smoother in 60fps versus 30fps?

Size matter !

[meuhlavache]

And I have the biggest one !

Games does *not* have to be realistic

[Anonymous Coward]

The problem with modern computer games is that they are trying to be too realistic. Being distinct from realtity is the true value of *any* game.

shutter rates above 1/24th of a second

[Anonymous Coward]

I do photography, so I'm not a film expert, but I don't think that It's necessarily true, as you claim, that in movies the film is always exposed for the 1/24th of a second at exists between frames. I would imagine that depending on scene lighting, desired exposure, desired depth of field, and other conditions, the exposure time could be anything from the upper limit of the camera's shutter all the way down to 1/24th of a second.

On a related note, if you're shooting at less than 1/60 of a second, you're likely to see blur in action scenes. If you wanted to eliminate blur from actions scenes in your movie, you'd probably need to be shooting at shutter speeds of 1/120 of a second or higher.

Re:shutter rates above 1/24th of a second

[Khyber]

They had to ask Jet Li to slow down in many movies because he was so fast with his movements that most couldn't be caught by camera. He was making 120FPS high speed film look slow.

Re:shutter rates above 1/24th of a second

[2obvious4u]

Citation needed, better yet go to the Mythbusters fan page and see if you can get them to test this. Snopes?

Re:shutter rates above 1/24th of a second

[Khyber]

Most decent martial artists have to force themselves to slow down because film usually isn't fast enough to see something taking only 1/100th of a second to move and impact. That's how much speed can be present in stiff martial arts like gung fu and tae kwon do, in kung fu movement is a bit more fluid so it's easier to capture.

Re:shutter rates above 1/24th of a second

[geekoid]

Bull. This was said about Jackie chan, Bruce Lee, and Shintaro Katsu.

You will need to provide a cite.

I've always been able to see more than 60FPS

[Rooked_One]

saying "no one can see above 60FPS" is bunk... you can't say that everyone is the same intelligence...

I've always been able to tell a difference between 60FPS and around 75FPS is where I can't see a greater difference...

Not sure if that is because i've played video games since I was 10, or what... but i've always been able to tell the difference - except above around 75FPS. I started noticing this over 10 years ago with GLQuake.

Not all notice the difference

[encebollado]

My parents have a 60 Hz LCD TV that has adjustable levels of motion smoothing. I can tell when the motion blurring is turned on because some parts of the movies just don't look "right". But, most of my family can't tell the difference - they just don't see anything change.

No shit?

[Hurricane78]

Back when we played in the Counterstrike league (before and around when 1.0 came out), a simple raise from 30 fps to 60 fps dramatically rose my kill/death ratio. ”nuff said.

The other thing that dramatically made me better (some years later): Going from a cheap wireless mouse to a good wireless mouse, to a great wired mouse (Razer Copperhead right now). (Also, if you still have mouse acceleration enabled, please go and disable it right now. :)

Online I always prefer 1. more frames 2. faster ping and 3. higher resolution, over better graphics. Only over 60 fps it stops making sense, because it does not change anything anymore. Same thing with a DPI above what my eyes can see. (Got 120%, so that’s pretty high.)

smove?

[hduff]

This granularity helped to smove out movement, including Mario's beloved jump.

A new techincamation term? Nice to see former President Bush got a gig as a copywriter.

24 fps

[Sir Holo]

Movies are 24 fps because film is expensive.

Showscan experiments

[Animats]

When Showscan (which, after a 2002 bankruptcy and acquisition, now just does ride films) was developed, they did blind tests in theaters to find out how fast the frame rate had to be before people couldn't detect any improvement. The threshold turned out to be in the 75-100 FPS range. "Peak emotional involvement" was around 72 FPS. Commercially, Showscan used 60FPS film, which was a compromise between indetectability and projector wear.

James Cameron, the director, talks about FPS in Variety. He knows 24FPS is too slow, and wants to get to at least 48. "Maybe on Avatar 2", he says. He points out that the NBA did the All-Star game at 60FPS (transmitted to theaters) and that was very well received. He points out that doubling theater resolution doesn't do much (only the first few rows are close enough to the screen to see the extra pixels) but increasing the frame rate is noticeable to everyone in the theater. With digital 3D, 24FPS strobing is now the weak point for the quality of the experience.

Cameron: "If every single digital theater was perceived by the audience as being equivalent to Imax or Showscan in image quality, which is readily achievable with off-the-shelf technology now, running at higher frame rates, then isn't that the same kind of marketing hook as 3-D itself? Something you can't get at home. An aspect of the film that you can't pirate."

With Cameron pushing for higher frame rates, it's going to happen. He'll figure out a way to use it, too; with high enough frame rates you can have fast, clean pans without strobing or blurring.

Re:Showscan experiments

[geekoid]

The movie cost 500 million dollars, and is mostly animated. Somehow I thing he could of come up with a higher framerate version if he wanted. Digitally it's not a hard issue, and there ar theaters that can show higher frame rate films.

50-60 frames is support by many new tv's, 72 is just right around the corner, and some people are looking at 300.

Motion Sickness?

[Anonymous Coward]

TimeSplitters 2 was the first 60fps game i ever played - it gave me motion sickness, something that had never happened before in over 15 years of playing video games. I'd say that was pretty important.

Disagree with 2 of 4 items

[dcollins]

Actually, items #1 and #2 are pretty far off the mark.

"The framerate of a game is usually directly tied to the processing of its logic." That really shouldn't be the case. Working on top-end race sim games back in the mid-90's, the very first mission statement was to have separate input, logic, and graphics threads so as to decouple these issues.

Not just there: higher FPS=faster response

[Anonymous Coward]

Not just there: higher FPS=faster response.

You see a picture. 1 frame old.

You react. 1 frame passes.

You see the result. 1 frame old.

So you have three frames possibly in your reaction delay. 30fps goes to 90fps to "get" 30fps reaction.

Especially with FPS (shooters) where you may have fine control with the mouse to get a headshot on a distant target while strafing.

Makes a lot of difference.

FWIW, we do high rate visual studies in my

[Anonymous Coward]

FWIW, we do high rate visual studies in my research lab at school. When using isochronic stimuli (in other words, the amount of time between frames is exactly equal), then the 30 Hz rate does seem to be a good approximation of fusion for most individuals. However, if the frames are nonisochronic, and are jittered by even a small amount, then the brain is capable of detecting rates upwards of even 60 Hz. So speed isn't exactly everything, there is an element of quality that should be considered as well.

Obligatory 100fps.com link...

[KonoWatakushi]

They have an interesting page about the question of "How many frames per second can the human eye see?"

120hz displays make a lot of sense in the near term, but even that frame rate may be woefully inadequate for providing a true, indistinguishable from reality VR experience. (Of course, it will be a while before the resolution and color reproduction also catch up, but it is an interesting topic.)

In any case, no one really knows what will be enough, but 24fps is certainly not. 120hz displays will be a great improvement, both for gaming and video. The latter, because both 24fps movies, and 30fps video (or perhaps 60 fields/s interlaced) divide into it nicely. (If you want to be pedantic, it isn't exactly 30fps, but it is close enough, and even PAL would look better.)

NOT the Same with audio...

[Anonymous Coward]

Audio is different from video. People who "hear better imaging, depth and transparency" at higher sampling rates are either imagining it, or using equipment that does not operate properly at 44.1/48.

Blind test, year long, trained audio professionals and musicians. Random chance results at telling the difference between 24-bit/96kHz audio and 16-bit/44.1kHz audio:

http://mixonline.com/recording/mixing/audio_emperors_new_sampling/

LiveTV vs Movies

[Loopy]

How many of you can tell the difference between live TV (e.g.: your local news) and movies? I can. It's rather obvious.

Re:LiveTV vs Movies

[stewbacca]

I can, but that's because projector technology sucks at the local megaplex and has nothing to do with frame rates.

Decoupling physics and rendering

[Dan East]

Most physics engines simulate best when the timestep is the same every update - larger timesteps result in less accuracy of the simulation, to name just one issue. Rendering time varies every frame depending on the number of polys rendered, etc. So it is standard practice to decouple the physics engine from rendering, which allows the physics engine to run at whatever fixed timestep is desired. Multiple physics updates can occur for a single rendered frame and vice versa. Interpolation of position is used so objects still appear to move smoothly even though the rendering update is seldom, if ever, exactly in sync with a physics update.

So while the parent's post is right in theory, in practice rendering and physics update rates typically have nothing to do with one another.

More info here on implementation details:
http://gafferongames.com/game-physics/fix-your-timestep/

Re:Decoupling physics and rendering

[LordKazan]

Must have changed the last five years.. used to be most games had 1 physics tick per frame rendered...

I would be interested in seeing their solution to concurrent data access problems in these new engines.. if they have normal lock that the physics of rendering threads have to acquire, or if they're doing shadow copy.

also, OpenGL is notoriously thread unsafe... you have to do all openGL operations in the same thread.. so i would like to see good solutions to that too.

This is what nitpicking nothing does to brains

[Chris Burke]

Otherwise, he would have simply said "designed." To put "evolutionary" in front of it seems to imply that this thought process, to me:

1. The brain appears to function according to a design. It works well together and appears to show evidence that it was planned/thought out/designed.
2. Having a design implies that something or someone intelligent planned out that design.
3. I don't believe in God, therefore I will attribute this design to evolution.

Actually, the reason he put "evolutionary" in front of "designed" was to make it explicitly clear that he was not using the definition of "design" that necessarily implies an intelligent designer. He was not saying that the functioning of the human brain could have only arisen from a series of deliberate steps.

The definition of "design" he was using is more like when someone says "those rocks on the ground form a neat design."

You see, regardless of what the truth behind the idea of the Creator is, our bodies do have a "design". They have a plan, a blueprint if you will. That's your DNA (and other things since it turns out biology is even more complicated than that). It guides how your body develops and how your brain works. It is a "design". However just because those words can often imply an intellect or a deliberate purpose behind them, they don't have to. For example, if someone said "the grooves in the limestone formed the blueprint for the subsequent quartz crystal growth", you would understand that they aren't saying that some civil engineer came by and carved the grooves in an intentional bid to achieve the result.

Which, I maintain, does not make sense. If the evidence of a design supports the idea of an intelligence behind it,

Design does not imply a creator. At least, not in all definitions or all contexts. "Design" the noun does not always imply "design" the verb on behalf of a "designer".

Thus your basic was make that most common of errors of /. pedants: First, assume that a common English word has one extremely precise definition from which all sorts of necessary implications arise, and then continue on without ever questioning that initial bad assumption. And thus your argument is the one that does not make sense.

It's like this: Your brain has all sorts of natural facilities, including the ability to learn new facilities. It is "designed" this way in the sense that your DNA creates the proteins that through a bunch of biochemistry end up creating your brain so that it has those facilities. Some of them are quite specific and serve obviously useful purposes for survival. When you brain does not work according to this design due to some issue, serious problems can result.

Where did the "design" come from? From natural selection. That's all.

I think it's pretty obvious what he really wanted to say. He wanted to say that the human mind evidences a design and he wanted to argue from that design and yet deny the existence of God.

I think it's pretty obvious that has nothing to do with what he was saying, but that is what your own beliefs and prejudices tell you he must have been saying.

Which is interesting, because that aspect of "design" and the idea that there must be someone or something that designed it or planned it or is sustaining it, therefore we can extrapolate from it and assume that laws of nature/physics will continue to be the way they are is an argument for the existence of a Deity that many scientists in the past used...

As someone who believes in God (and Jesus and all that)... No, that does not follow. The "something" that "designed or planned" the "design" was random chance combined with natural selection, and what sustains it is also natural selection. That is all that is required.

You can see the potency of this kind of "design" in genetic algorithms. There, random processes still play the role

Mame and other emulators

[pancakegeels]

Emulation is a fantastic field, and one thing they are great at is making old games look great, Scale2x and so on. The thing is, I am pretty sure I have seen it crank out some super-massive framerates. for some titles. The video for Starblazer gave no indication of how it was rendered. Which, in addition to the video being encoded for flv, softens the initial point in this article. Good links in it though - I like it when people give interesting references.

Re:60fps is also twice as fast gameplay as 30fps

[Anonymous Coward]

I know I can tell the difference between 60fps and 30fps because at 30fps everything runs at half speed. That includes input and sound as well as video. Why do you think PAL console games suck, they run at 5/6 normal speed! Of course, that doesn't apply to PC games which can somehow run at the same speed regardless of FPS (i.e. increasing the frame rate to 120fps doesn't make the entire game run twice normal speed).

Idiotic slashdot masses

[Anonymous Coward]

As usual the idiotic slashdot masses only care about what the article says, and not the scientific facts.

Sure, you can 'see' differences greater than 30Hz, but the human brain on average takes 100-200ms to actually react to visual stimulus.

There have been many MANY MANY scientific studies on human response times (simple search on any known paper repository will bring up many whitepapers) to visual and auditory stimulus, none of which have recorded response times of anything less than 80ms to my knowledge - which means in terms of humans reacting to games, you really only need 15Hz+.

But as I said, we can still 'see' things faster than that - even if we don't comprehend it - thus 60Hz will look smoother and prettier than 30Hz.

Associating game mechanics and input processing to framerate is a poor argument as well, any decently written engine/game will have an asynchronous rendering pipeline completely independent of any game processing or input processing - and any that don't are simply poorly written games by poor programmers. Similarly many games written on a fixed Hz game-update-timer are going to have flaws as well, and really ought to have their design completely re-thought.

Re:Idiotic slashdot masses

[grumbel]

Games aren't about pure reactions, they are also about terrain traversal and following objects and those get a hell of a lot easier the more FPS you got, as the difference between frames just gets way to large to perform actions or predictions with any amount of accuracy when you are trotting around at something as low as 15fps.

Associating game mechanics and input processing to framerate is a poor argument as well,

Its not the way todays games are written or should be, but it used to be the way to do things on earlier consoles and those consoles managed to give you very fluent and very low latency gameplay that 20 years later todays games have a very hard time to match or even get close to, as everything is delayed and buffered multiple times before it makes it to the screen.

Refresh rate

[Anonymous Coward]

Isn't the actual perceived image bound by the display refresh rate? If your refresh rate if 80 Hz (say) then all the display is capable of is changing the image 80 times per second. So with a 160 Hz frame rate, you are dropping every second frame. (?)

Infinity Ward understands the need for 60fps

[Asterra]

This article is no news at all to the likes of Infinity Ward. The difference here is that Insomnia are now chiefly PS3 developers, and as the PS3 is famously difficult to develop for, even after you have a tried and solid engine going on it (it's often said that it takes twice as long and costs twice as much money to get a game completed on PS3 as it does for 360), the decision to switch to 30fps is almost certainly STRICTLY thanks to the difficulties associated with PS3 development.

Insomnia choose poorly. To make money, they should have gone multiplatform. Instead they decided to cut back on the quality of their products. It's a choice that smacks of motives they must regard as stronger than profit, and I have to trust they're content with the consequences of their decision. It also puts to rest the famous myth, now three years old, that the PS3 will eventually real some sort of unlocked potential.

24 fps movies

[bobbuck]

Movies at 24 frames per second have the advantage of showing an average of that 1/24th of a second. Doom at 24 fps will show consecutive clear frames, which looks jumpy. A movie will show the range of motion during each frame which blends into the next and makes it look much more realistic.

Change in frame rate matters

[indil]

I've been reading IGN.com a long time, and I've consistently heard from the people there who stare at games all day long that 30 vs. 60 frames per second doesn't matter because it's hardly noticeable; it's the change in frame rate as you play that's noticeable and looks bad.

Death of 120Hz gaming

[Anonymous Coward]

Anyone who has been playing quake3 knows the pain of 85Hz. LCD is so common now that the suckers have degenerated to speculate the need of 30Hz! Hell, I fall sleep between frames on 30Hz. There are few 120Hz capable LCD models around now, I can only hope that the need for fast framerates to be used with shutter stereoglasses saves the day.

Not sure that's consistent...

[Estanislao Mart�nez]

More accurately - most film cameras don't have a notion of a shutter 'speed'. The film roll still goes by at 24fps, but the actual shutter is a wheel. That wheel can have various sizes of gaps (to increase/decrease exposure *time*) and sizes (to produce specific motion blur effects; e.g. an object leading its own motion blur path requires a small shutter opening at first, ending in a large shutter opening).

I suspect that by the criteria you're using, most stills cameras don't have "shutter speeds" either.

The way the rotary disc shutter achieves different exposure levels actually sounds very much the same as the way a focal plane shutter in a stills camera implements fast shutter speeds. With a typical interchangeable-lens camera, when you set a very fast shutter speed like 1/1000, the shutter doesn't open completely for 1/1000 of a second and then close.

What happens is that a focal plane shutter has two curtains: the first curtain, which opens to expose the film/sensor, and the second curtain, which then closes to block the light. The curtains always move at the same speed, so the "shutter speed" is really just the timing between when the first curtain starts to open and the second curtain starts to close. So when you set shutter speed faster than what's called the "maximum sync speed," the second curtain starts to close before the first curtain is fully open. In effect, what you get is a narrow slit of light moving from one edge of the frame to the other; the narrower the slit, the "faster" the shutter speed, even though the curtains are not moving any faster than they always do. This slit of light gives you effectively the same effect as the gaps in the rotary wheel shutter; the wheel shutter seems to simply be a continuous implementation of the same idea.

That said.. you can't - short of electronic shutters - expose for -more- than the film's fps, though. A bit under 1/24th of a second is the most you'll get (that 'bit' being required to transport the film to the next frame).

Indeed. In a stills camera you can set shutter speeds slower than 1/24 because there's no roll of film rolling; even if you open the shutter for 30 seconds, the film or sensor will stay put.

Correction

[Estanislao Mart�nez]

I suspect that by the criteria you're using, most stills cameras don't have "shutter speeds" either.

Um, I'm certainly wrong about the "most" part there. Most stills cameras don't have focal plane shutters. Most interchangeable lens still cameras do, though.

He's got it all wrong

[Eraesr]

The debate about 30fps vs 60fps isn't about whether people can actually notice the difference. I don't think I've ever seen a developer say that the difference is not noticeable. The thing is that if they render at 30fps rather than 60fps, they have twice the amount of time to render a single frame, allowing for much more details and effects in each scene. So the question isn't whether people can see the difference in framerate, but it's about what level of detail the developer wants to achieve and whether or not that's possible at 60fps.

People interested in the subject should take a look at Eurogamer's Digital Foundry (http://www.eurogamer.net/digitalfoundry/). They got loads of technical game reviews and articles about this very subject.

a 30p / 60p example

[plonk420]

source: Neuro's demo Masagin (apologies, it's not my best encode, but i had to work and wanted to get all the samples up ASAP)
(for more wheezy computers)
halfres 720p30: http://www.megaupload.com/?d=ZZT4OGLD
halfres 720p60: http://www.megaupload.com/?d=WHWJIXHZ
(for manly computers that eat HD for breakfast (or just a $30 radeon 3000+ or geforce 8000+ and mpc hc))
720p30: http://www.megaupload.com/?d=LSM31W74
720p60: http://www.megaupload.com/?d=9XC9NN7A

parts that especially leap out at me as being noticeable (altho 90% is noticeably not 60fps smooth to me)
0:28-0:59
1:26-1:29
1:58-2:28

It's called ray tracing

[Jeremy Visser]

I personally would prefer it if computers were stong enough to calculate a photon hitting a material, reflecting its non-absorbed light into a "camera" object in game and taking the rendered picture and sending it to the monitor, thus creating a more realistic lighting effect, but we just aren't there yet.

Already been done. It's called ray tracing, and does exactly what you describe (except the other way around -- it traces light rays in reverse from the camera to the light source).

The trouble with ray tracing is that while it looks absolutely beautiful and stunningly realistic it's extremely impractical to do in real time. It can be done, but only with a supercomputer.

I've heard of some game engines being adapted to use ray tracing algorithms (here's a hacked up Quake 3 that apparently does so, but doesn't look any better than the original game). Here's an interesting interview with an Intel dude talking about it. In terms of actual usability though, there's no way you're going to actually play any of those games yet.

Yet.

Artificial blurring is overrated

[TheLink]

But why? Motion blur is overrated. Sure put it in scenes where it is "important to the story/gameplay", but to use it whenever there is fast motion is stupid.

Why? Because people aren't staring at the same spot on the screen all the time. And nowadays screens are getting bigger.

Say in real life, you're in a room where there are two moving objects that are moving around at fast but eye-trackable speeds in different directions.

If you are staring at sommething else, both objects are blurry.

But if you start to look at one, that particular object becomes _sharp_, the other object becomes blurry.

You look at the other, it becomes sharp and the other becomes blurry.

When a game or movie blurs moving stuff, it just makes stuff you are looking at look out of focus even if they are moving at speeds which your eye can track. You can't focus on it even if in real life you could!

With motion blur, I often experience eye strain when I try to track moving objects/backgrounds that have been blurred.

Then there are the artificial "out of focus" shots in static scenes. These effects should also be restricted to scenes where it is important to the story that only a few items are in focus.

In Avatar (2D), my eyes were often trying to focus on blurry images and it wasn't pleasant - initially I was wondering what was wrong with my eyes - felt like I had difficulty focusing on stuff.

When I watched it in 3D, I realized that a lot of stuff was actually blurry and it wasn't my eyes. In some fairly static scenes the focal range was low - only a few objects were in focus. Then in some scenes the moving objects were blurry. Whereas in other scenes most stuff was in focus. In Avatar 3D it was easier to figure out where I "should" be looking and avoid the eyestrain bits :).

If you ask me I prefer as much of each frame to be sharp and in focus as possible, then let the limitations of my eyes blur it.

Artificial blurring (motion or defocus) is like listening to artificially degraded music/audio. While there are some cases that call for it (distance effect) it's just silly if you use it a lot.

Re:Artificial blurring is overrated

[TheLink]

Oops "focal range" should be "depth of field".

Frame rates are not only what the eye sees

[gerryn]

As briefly mentioned in the article linked, it's not only about if the picture is percieved as being smooth or not. When playing FPS games the mouse's responsiveness is directly linked to the FPS. A good example of this is the V-SYNC option available in many games, even though the game runs very smooth with V-SYNC on, the controls are all but smooth, and you end up with a very unresponsive and "rigid" camera. At least that's how it is for me. I have discussed this issue with quite a lot of people and some say they don't notice anything while others say they do. I have always noticed, and it does not matter what kind of hardware I use (it happens with all hardware and all games). I think some people have the ABILITY to notice these kind of things, while others don't (I might be wrong...) The nightmare scenario is 1) 30 FPS and the responsiveness of the mouse that comes with it, secondly the visual experience which is also greatly reduced. With 60 FPS (at least) it feels good, and I have noticed that with even higher framerates the responsiveness is increased, which makes sense since the DPI of the mouse has increased greatly over the years as well. I dont know the ratio of mouse dpi / framerate for a good experience but its obvious that there must be one. /G

Damn, time for a new graphics card

[stewbacca]

...sounds like this entire article was sponsored by nVidia.

What this blog fails to discuss is the concept of "good enough" is good enough. A game running a 60 fps is plenty good enough and hardly anybody would notice an improvement if you bump that number up. That's why it's good enough, because the amount of money it takes to improve the frame rate isn't worth the return (little to no perceived quality improvement).

I know a few of you like your giant, uncompressed audio files as well, but I'm just fine with my 128 mp3s (anything below that is noticeably degraded, anything bigger than that makes no difference to this 40 year old drummer's ears).

Not common experience or knowledge?

[Anonymous Coward]

I'm a bit concerned here - about the wagon wheel effect. Some articles talking about "people also report experiencing it in real life" like this wasn't widely known before as a real-life experience. Really?

I've noticed it since I was a little kid, think 5 and below! I always wondered when someone would give me a diameter-to-rpm ratio for a) always appears moving forward b) stands still c) moves backwards. Figured that since it was obvious everyone was seeing it, someone would've worked out the details. I see mention of frequencies, etc. but didn't realise it quite likely differs from person-to-person.

I don't think my visual acuity is any higher than the next person's.

Are we next going to say not every sees the little circles that keep 'falling down', at least when they're looking at a plain surface? (because of the micro-organisms in our eyes and the eyewash from blinking). Yeah, that I know not everyone notices.

What about the whitenoise (like tv screen on a non-transmitting channel when looking at a night sky? or when closing your eyes?

Why higher framereate _really_ matters in an fps..

[fcrick]

I play FPS games - 30 fps is fine as far as visual quality goes - sure, 60 fps is better, but I don't care - it's not the visual quality that concerns me.

What does concern me is the delay in getting the information I need when I'm playing.

Ideally, I'd like infinite FPS - then, when an opponent appears, I'd see it as soon as possible after the data makes it from the computer to the monitor. At 30 FPS, there is an additional delay, probably up to 33 ms, probably averaging 16 ms. At 60 FPS, that additional delay is cut in half, and at 120 FPS, it's cut in half again. In short, I get relevant information sooner, and that makes me play better.

Often battles in FPS games are literally two people who both shoot each other in the head for a one shot kill as soon as they see each other. Players want to minimize any delay so the game will decide they shot first, and win the encounter - every little ms matters, as any skilled gamer knows all too well.