Slashdot Mirror


Carmack: Next-Gen Console Games Will Still Aim For 30fps

An anonymous reader sends this excerpt from Develop: "Games developed for the next-generation of consoles will still target a performance of 30 frames per second, claims id Software co-founder John Carmack. Taking to Twitter, the industry veteran said he could 'pretty much guarantee' developers would target the standard, rather than aiming for anything as high as 60 fps. id Software games, such as Rage, and the Call of Duty series both hit up to 60 fps, but many titles in the current generation fall short such as the likes of Battlefield 3, which runs at 30 fps on consoles. 'Unfortunately, I can pretty much guarantee that a lot of next gen games will still target 30 fps,' said Carmack."

33 of 230 comments (clear)

  1. Detail by Dan+East · · Score: 4, Insightful

    Would you rather have double the detail at 30 FPS, or half the detail at 60 FPS? Considering most people can't perceive frame rates faster than 30, it makes a bit of sense to push more polygons instead.

    --
    Better known as 318230.
    1. Re:Detail by Radres · · Score: 4, Insightful

      I think Carmack's point is that the other studios will push half the content at 30fps because they're lazy.

    2. Re:Detail by epyT-R · · Score: 5, Insightful

      Not this again.. This assumption is based on perceived motion from frames containing captured motion blur and even in such (24/30hz) frames, motion is NOT transparent to most people. With games there is no temporal data in frames, so it's VERY obvious. Even 60 is to many gamers, and is why they opt for 120hz (real 120hz, not hdtv '120' interpolated which looks terrible) panels and video cards that can push them.

      Then there is input lag. Its perceived turnaround time is very noticeable at 30fps, and if the rendering is not decoupled from the input polling/irq, the latter's latency actually does go up. id had to patch quake 4 to make it acceptable to play because the 60hz was dropping inputs and looked choppy as hell compared to previous releases. Enemy Territory quake wars, which is also idtech4, was locked at 30 and was deemed unplayable by many.. I think it was one of the reasons the game tanked. It was actually painful to look at in motion.

      Console devs always push excessive graphics at the expense of gameplay because the publishers want wow factor over playability. This was true in the 8bit and 16bit days too. Some games suffered so badly they were deemed unplayable. This is why pc gamers value useful graphics configuration capability in their games. Often what the publishers/devs thought as 'playable' was not what the community thought was playable, not that this should shock anyone with today's 'quality' releases.

    3. Re:Detail by Anonymous Coward · · Score: 5, Informative

      ... Considering most people can't perceive frame rates faster than 30 ...

      This myth needs to die.

      Everybody can perceive frame rates faster than 30 fps. In fact, almost everybody can perceive frame rates faster than 100. Check the linked article, this is really a tricky question. Some things to consider:

      - Games have no motion blur, or, as many modern games are implementing now, they use a pathetic, fake imitation that looks worse than no motion blur at all. Hence, they need much higher frame rates to show fluid motion. At 60 fps with current technology (including so-called next-gen), motion will look much better compared to 30.

      - Decades of cinema have been training most people to perceive low-quality, blurred animation as 'film quality', and smooth, crisp animation as 'fake' or 'TV quality'. Many, many people consider a 48fps Hobbit to be worse compared to a 24 fps one. This is a perception problem. Games could have the same issues, except they've evolved much faster and most people didn't have the time to get used to bad quality.

      - Consider the resolution problem. Higher resolution requires higher fidelity. At higher resolution, you'll demand higher quality textures and shading to reach similar levels of immersion, since details are now much more apparent. Same thing happens with animation and higher frame rates. This doesn't meen we should stay at low resolutions, 16 colors, black & white, or 30 fps. This just means we need to do better.

      - And... a game is not film, and latency matters. A lot. At 30 fps, you need to wait twice the time to see any feedback from your input. In most games you will just train yourself to input the commands in anticipation without even knowing a word about latency, but in action games, where your reaction time matters, latency is a problem. And many other sources of latency add to the sum, such as clumsy 'smart' TVs post-processing your images, or badly engineered 'casual' motion wi-fi controllers.

      In other words, yes, I'd rather have half the detail and 60 FPS. Except if your game is no game at all, and just a 6 to 10 hours movie. Since most of the top videogame chart entries fill this description today, I can see why many developers will remain at the 30 fps camp.

    4. Re:Detail by Nyder · · Score: 5, Informative

      Would you rather have double the detail at 30 FPS, or half the detail at 60 FPS? Considering most people can't perceive frame rates faster than 30, it makes a bit of sense to push more polygons instead.

      When it comes to games, you can tell the difference between 30 fps and 60 fps. TV/Movies, No, you can't. Video games, yes you can.

      I should of mentioned the reason why.

      When you shoot video you capture single pictures. When people are moving in these shots, the have motion blur. How much motion blur depends on how fast they are moving and how many shots per sec you take. Our eyes see the motion blur and our mind fills in the rest, which is why we are okay with 24 & 30 fps for movies/videos.

      When you do video games, each frame is smoother, doesn't have the motion blur that real life video would have. Granted, games started adding in motion blur, but it's not the same. This is why the more frames per sec generally make games look better and play better.

      We did cover this in the Hobbit at 48fps submission.

      --
      Be seeing you...
    5. Re:Detail by Anonymous Coward · · Score: 2

      if you have 60 fps instead of 30 fps, you have 50% more frames

      None of your post really made much sense, but this bit isn't even correct math...

    6. Re:Detail by frinsore · · Score: 4, Informative

      For a 60fps game there's about 16ms per frame and with current gen consoles about 8ms is lost to API call overhead on the render thread. Of course current gen consoles are years behind and constrain rendering APIs to be called from a single thread but I'd still be very surprised if there was a console that could support a triple A game above 70fps in the next 10 years (for resolutions 720p and above).

      You've barely scratched the surface of input to perception lag, here's an answer by Carmack to people questioning another one of his tweets:
      http://superuser.com/questions/419070/transatlantic-ping-faster-than-sending-a-pixel-to-the-screen
      Of course most engines come from a single threaded game mentality where they'd poll for input, apply input to game state, do some AI, do some animations, calculate physics, then render everything and repeat. Current gen consoles has freed that up some but most engines didn't go above 2 or 3 major threads because it's a difficult problem to re-architect an entire engine while it's being used to make a game at the same time. Sadly the better games gave user input it's own thread and polled input every 15ms or so, queued it up, and then passed it on to the game thread when the game thread asked for it. Input wasn't lost as often but it didn't get to the game any faster.

    7. Re:Detail by epyT-R · · Score: 5, Insightful

      lets count the fallacies shall we?

      1. argument from antiquity (it's old so it sucks)
      2. argument from inverse popularity (no one does it now so it sucks)
      3. appeal to realism (when did I say quake was realistic? I said higher steady framerate allows for better perception of action)
      4. ad hominem. I'm not butthurt. Perhaps you prefer COD et al because you can't play something requiring more attention and lower reaction time. It's alright, I'm not crazy at quake either.. I was only a bit above average as far as competent players go, but I enjoyed the fluid, fast gameplay much more than the tedious waiting and camping of CS, action quake and its subsequent 'realism' clones. There's no need for insults.

      If anything, it's the dominant playerbase who reason like your post who are to blame for why so many games today lack actual gameplay learning curves. There's nothing to master and it's all about pressing the right button at the right time a la dragon's lair single player, or having a real time rendered backdrop for VOIP 'multiplayer' conversations...all of this while fumbling around with simplified gameplay mechanics despite the fact they were dumbed down specifically to make the pad workable at all. That's not what I got into gaming for, but to each their own.

    8. Re:Detail by Anonymous Coward · · Score: 2, Insightful

      I should of mentioned the reason why.

      I should have mentioned the reason why.

      Just an FYI...

    9. Re:Detail by Anonymous Coward · · Score: 2, Insightful

      I should of mentioned the reason why.

      Repeat after me: "should've" means "should have", not "should of".

    10. Re:Detail by robthebloke · · Score: 2

      That's not really how it works. Every dev I've ever worked with in the games industry, all aim for 60fps. Given enough time and resources, that's what they'd all end up delivering. Since they are never given enough time or resources, by either management or their publishers, there will be a time when the only option is to drop to 30fps. It has nothing to do with laziness, and everything to do with money.

  2. Re:HOBBIT IN 48 FPS - YECHH! by epyT-R · · Score: 5, Insightful

    people who complain about higher framerates never seem to have a justification other than 'it's not what I'm used to'. What about the 48fps made it suck? Please avoid using 'audiophile-like' subjective/emotional terms.

  3. lack of proper triple buffering by stanjo74 · · Score: 4, Informative

    Neither DirectX nor OpenGL support proper triple buffering to avoid tearing at variable frame rates. Because of that, if you want tear-free rendering, but cannot keep up at 60 fps all the time, you must render at 30 fps or 15 fps, but not, say 48 or 56 fps. You can render at any variable frame rate if you allow for tearing (which most games do and avoid the headache of v-sychs altogether).

  4. Fixed Refresh Rates by bazald · · Score: 4, Insightful

    A display (television or monitor) has a fixed refresh rate. Assuming vertical synchronization is turned on to avoid tearing, you're pretty much limited to a framerate which evenly divides into the true refresh rate of the display. If the refresh rate is 60 fps, possible targets include 60 frames per second (providing 16.7 ms of computation time per frame), 30 FPS (providing 33.3 ms of computation time per frame), 15 FPS (providing 66.7 ms of computation time per frame), and so on. Anything below 30 FPS is kind of a joke, so nobody reputable would consider allowing more than 33 ms computation per frame in a shipping game.

    --
    Insert self-referential sig here.
    1. Re:Fixed Refresh Rates by stanjo74 · · Score: 2

      Unless, you use a technique called "triple buffering", in which case you can have tear-free variable frame rate at any rate. Unfortunately, none of the major 3D APIs have provisions for this. I always wondered why such a fundamental omission for a graphics rendering API.

    2. Re:Fixed Refresh Rates by Brulath · · Score: 5, Interesting

      TechReport analysed the nVidia 680 a bit after its release and had a piece on adaptive vsync which should answer your question.

      Quoted from an nVidia software engineer:

      There are two definitions for triple buffering. One applies to OGL and the other to DX. Adaptive v-sync provides benefits in terms of power savings and smoothness relative to both.

      - Triple buffering solutions require more frame-buffer memory than double buffering, which can be a problem at high resolutions.

      - Triple buffering is an application choice (no driver override in DX) and is not frequently supported.

      - OGL triple buffering: The GPU renders frames as fast as it can (equivalent to v-sync off) and the most recently completed frame is display at the next v-sync. This means you get tear-free rendering, but entire frames are affectively dropped (never displayed) so smoothness is severely compromised and the effective time interval between successive displayed frames can vary by a factor of two. Measuring fps in this case will return the v-sync off frame rate which is meaningless when some frames are not displayed (can you be sure they were actually rendered?). To summarize- this implementation combines high power consumption and uneven motion sampling for a poor user experience.

      - DX triple buffering is the same as double buffering but with three back buffers which allows the GPU to render two frames before stalling for display to complete scanout of the oldest frame. The resulting behavior is the same as adaptive vsync (or regular double-buffered v-sync=on) for frame rates above 60Hz, so power and smoothness are ok. It's a different story when the frame rate drops below 60 though. Below 60Hz this solution will run faster than 30Hz (i.e. better than regular double buffered v-sync=on) because successive frames will display after either 1 or 2 v-blank intervals. This results in better average frame rates, but the samples are uneven and smoothness is compromised.

      - Adaptive vsync is smooth below 60Hz (even samples) and uses less power above 60Hz.

      - Triple buffering adds 50% more latency to the rendering pipeline. This is particularly problematic below 60fps. Adaptive vsync adds no latency.

      So triple buffering is bad because it could cause an intermediary frame to be dropped, resulting in a small visual stutter despite being 60fps. There's a video of adaptive vsync on YouTube.

    3. Re:Fixed Refresh Rates by stanjo74 · · Score: 2

      http://www.anandtech.com/show/2794 "So, this article is as much for gamers as it is for developers. If you are implementing render ahead (aka a flip queue), please don't call it "triple buffering," as that should be reserved for the technique we've described here in order to cut down on the confusion. There are games out there that list triple buffering as an option when the technique used is actually a short render queue. We do realize that this can cause confusion, and we very much hope that this article and discussion help to alleviate this problem."

  5. Re:HOBBIT IN 48 FPS - YECHH! by TubeSteak · · Score: 2

    Please avoid using 'audiophile-like' subjective/emotional terms.

    Our expectations & emotional experience colors our subjective experience.
    And it's a scientifically measurable effect.

    That isn't to say objective measures are irrelevant, only that they are not all that is relevant.

    --
    [Fuck Beta]
    o0t!
  6. Much ado about a single tweet by dirtyhippie · · Score: 3, Insightful

    Good lord, this entire article is based on one tweet - 107 characters. Surely we could have waited for Carmack to say something more detailed than this??

  7. Re:HOBBIT IN 48 FPS - YECHH! by epyT-R · · Score: 5, Insightful

    Well, there's a bandwagon of snobbery out there about this issue. Kinda like people who say vinyl or vhs is superior to digital audio and video, I suspect this whole 'butt is it art' routine is more about social exclusivity and differentiation (and unhealthy doses of insecurity) than it is about their actual experience. I could understand if someone got motion sickness from the higher rate and didn't like that, but otherwise I cannot understand why someone would want animations deliberately choppy.

    With today's style all about fast cuts and jerkycam, I think the higher framerate would help the viewer track the action.. It helps in games and I suspect it would help me in such scenes, esp when they pile on the blur and urinal tournamint style colored lighting..

  8. Re:HOBBIT IN 48 FPS - YECHH! by gigaherz · · Score: 2

    It's less blurry and doesn't give you headaches, why would ANYONE want watch a movie that's NOT blurry or -- if seen in 3D -- gives you headaches?

    I do agree that it doesn't have the "cinematic" feel of standard movies, so it feels weird when you watch it -- different. But it's so clear, smooth and headache-free that it's worth losing that. In fact, I'd like to see a movie in 60 or 75fps someday.

  9. Same as PC. But you can still go for 60. by Sarusa · · Score: 3, Insightful

    It's a given that most will target 30fps since more shinies looks better in screenshots and youtube videos than 60fps does. And most consumers can't tell the difference until put a 60 and 30 fps version side by side and let them play.

    The leaked/rumored PS4/XNext specs show them as equivalent or slightly weaker than current mid-high gaming PCs, and those can't do 60 fps locked on all the recent shiny games at 1920x1080 with all effects on (except those like CoD MP that specifically target it), so it's unlikely the consoles would. Cheap components is the driver, especially for PS4.

    But there's no reason a fighting game or fps can't aim for 60fps on the new gen if it wants to. Use your shaders and effects wisely and no problem.

  10. Re:HOBBIT IN 48 FPS - YECHH! by SecurityTheatre · · Score: 5, Interesting

    That's exactly the problem I had.

    The "Jerkycam" works BECAUSE of the 24fps.

    The only time I found the 48fps showing to be uncomfortable and weird was during very fast action, jerky motion sequences. It suddenly feels like high-fidelity jerkyness, which makes it lose its tendency to portray "oh noez, stuff is blurry and out of control, even the camera", and just feels like "why is the dude shaking the camera so much?"

  11. Re:HOBBIT IN 48 FPS - YECHH! by epyT-R · · Score: 5, Interesting

    I guess my interpretation of jerkycam was always "why the hell is he shaking the camera so much?" Its' annoying and distracting, especially when it's every other scene. If the sharpness of movement isn't sufficient it's because the movements aren't sharp enough. The lower framerate just hid that.

  12. Re:HOBBIT IN 48 FPS - YECHH! by hack++slash · · Score: 2

    Stupid AC, movies are not games, in games you want the highest framerate possible because this (usually) means quicker response times from keyboard/mouse/gamepad, increasing the feeling of immersion in the game.

    This is especially so with the Oculus Rift type headgear being developed, the less lag between your input and the computer's visual output the more immersed you feel, with movies you're simply an outside observer.

    --
    To do something right, you often have to roll up your sleeves and get busy.
  13. Re:HOBBIT IN 48 FPS - YECHH! by simoncpu+was+here · · Score: 2

    My subjective reason for hating 48fps: the movie looks like a sitcom.

  14. Re:HOBBIT IN 48 FPS - YECHH! by Anonymous Coward · · Score: 3, Insightful

    That's just conditioning -- you're used to seeing sitcoms in higher framerates than movies. If sitcoms were traditionally filmed in color and movies traditionally filmed in black and white, you'd be ranting about how much color sucks in movies.

  15. Re:HOBBIT IN 48 FPS - YECHH! by AmiMoJo · · Score: 3, Interesting

    It's just a way of doing action on the cheap. The special effects and stunts don't have to be as good because no-one can see them clearly. A bit of low budget CGI looks much better when blurred and our of focus and only on the screen for 1/24th of a second.

    Transformers invented a variation where the CGI has so much detail and is frames so poorly on screen that you can't make out where the character's limbs are or what is actually going on anyway, so again it seems to be better than it actually is. If you step through the action sequences frame by frame there is a very clear disconnect between the CGI and real objects that get thrown around by poorly hidden explosives and hydraulics. Terrible camera work hides a multitude of lameness.

    --
    const int one = 65536; (Silvermoon, Texture.cs)
    SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
  16. KB+M, multiplayer, no lag: pick two by tepples · · Score: 4, Insightful

    using a controller, playing a lot of single player games

    You can have a mouse and keyboard. You can have multiplayer. You can have no lag. But you can't have them all. Mouse and keyboard + multiplayer = online PC game with net lag. Mouse and keyboard + no lag = single-player PC game. Multiplayer + no net lag = same-screen multiplayer game with gamepads.

    1. Re:KB+M, multiplayer, no lag: pick two by Hatta · · Score: 2

      That's what LANs are for.

      --
      Give me Classic Slashdot or give me death!
  17. Delay vsync by half a scanline by tepples · · Score: 3, Informative

    but how do you do progressive when the hardware is built for interlaced?

    The vertical sync pulse is delayed by half a frame before odd fields according to this diagram. Delay it and the analog hardware will begin retrace a half scanline later, which produces an odd field. Don't delay it and the TV interprets it as an even field.

    We're talking analogue TV sets here - they DON'T DO progressive. Period.

    Then how does my analog TV set do progressive when my NES, Genesis, Super NES, original PlayStation, or Nintendo 64 is connected to it? Question mark?

  18. CORRECTION by tepples · · Score: 2

    "The vertical sync pulse is delayed by half a frame before odd fields" should be "half a scanline period".

  19. Re:HOBBIT IN 48 FPS - YECHH! by lister+king+of+smeg · · Score: 4, Funny

    The real question is why are you expecting quality from transforms?

    --
    ---Saying gnome 3 is better than windows 8 not so much a compliment as it is damning with light praise.