Slashdot Mirror
Valve's SteamVR: Solves Big Problems, Raises Bigger Questions
An anonymous reader writes: When Valve debuted its SteamVR headset recently, it came as somewhat of a surprise — it certainly hasn't gotten the same level of hype as the Oculus Rift. But people who got to try out the new headset almost universally impressed with the quality of the hardware and software. Eurogamer has an article about the device expressing both astonishment at how far the technology has come in three short years, as well as skepticism that we'll find anything revolutionary to do with it. Quoting: "R demands a paradigm shift in the thinking of game designers and artists about how they build virtual space and how players should interact with it. We're only at the very beginning of this journey now. ... but this process will likely take years, and at the end of it the games won't resemble those we're currently used to. In short, they won't be Half-Life 3."
The author thinks simulation games — driving, piloting, and space combat — will be the core of the first wave, and other genres will probably have to wait for the lessons learned making sims good. He adds, "...the practical challenges are great, too — not least in persuading players to clear enough space in their homes to use this device properly, and the potential for social stigma to attach to the goofy-looking headsets and the players' withdrawal into entirely private experiences. I still think that these present major obstacles to the widespread adoption of VR, which even more practical and commercially realistic offerings like Morpheus will struggle against."
The author thinks simulation games — driving, piloting, and space combat — will be the core of the first wave, and other genres will probably have to wait for the lessons learned making sims good. He adds, "...the practical challenges are great, too — not least in persuading players to clear enough space in their homes to use this device properly, and the potential for social stigma to attach to the goofy-looking headsets and the players' withdrawal into entirely private experiences. I still think that these present major obstacles to the widespread adoption of VR, which even more practical and commercially realistic offerings like Morpheus will struggle against."
Have you changed hemispheres semi-recently?
I had the same issue when I relocated from Australia to the UK, I had gone from have a good sense of direction to always being about 180 deg out. It wasn't till my old man said it is because the sun is on the wrong side that it clicked. In the southern hemisphere the sun is always to the North, in the Northern it is to the south. Sub-consciously I must have been drawing on that.
I shouldn't get into the details of my project, but I've been thinking this over. We're going to need lots of testing for a paradigm shift in interfaces.
Okay, so your basic game camera modes are top-down, isometric, third person, and first person. Simulations build on that with chase cams, orbital cams, fixed cams, and mobile cams that rotate around the subject. In VR, each one of these will have consequences, and those consequences need to be known.
Let's start with the obvious. First person will work, right off, no changes. It's the most suited to VR. Third person will work, but where we feel like we're watching our characters when we're in third person mode on a screen, in VR we it will feel more like an out of body experience. That is, if we continue to identify with the character, otherwise we'll feel like some kind of disembodied entity following a protagonist. We have to keep the player from feeling compelled to look "around" things, so keeping from obstructing the player's view will take on a new importance. If a tree gets in the way of the shot, instead of feeling like a tree is in the way, we'll feel like we've run into a tree.
But speaking of disembodied entities, that's exactly what top-down and isometric views will feel like. So, let's hone in on that. Will virtual worlds feel like shoebox dioramas or will we feel like birds, aircraft, or perhaps deific figures peering down upon the world? These analogies can be expressed literally in virtual spaces, so playing with them in interfaces can potentially do amazing things for the experience. Imagine a city-builder game, top-down, in VR, where the occasional cloud or birds below are timed and positioned just right to reinforce that feeling. Now imagine that the borders of the window and map make us feel like we're looking down on a model. Tilt-shift post processing can become very important, very soon!
Now we come to sims. Making these the first wave of VR games is a gimmick. It's like the gratuitous addition of objects protruding from the screen in 3D movies; done just to let us get the full experience. What do you imagine in VR? Feeling like you're flying, roller coasters, feeling like you're going very fast. But look at 3D. Having arrows or monster claws or whatever come out of the screen is neat the first few times, and then it takes more finesse. Simulations will probably be just like that. But there's a much bigger issue to think about here. It becomes apparent with simulations, but applies backward through this post all the way to first person.
Hone in on that rotating cam. Can you see the potential for motion sickness and dizziness? Uh oh. That same potential applies everywhere. The awesome thing about VR is that you can feel like you're there. The tricky thing about VR will be that you feel like you're there. I foresee posts about people throwing up while playing flight sims; not even trying to be funny. So, there's some balance between free movement, the rush of certain kinds of motion, gameplay, and the not-so-nice things our brains will do to us under certain conditions.
And we have absolutely no idea how to quantify or even accurately describe the balances involved in this. VR is going to rock when it fully takes off! I can foresee even an entirely new cinematic experience where we watch movies shot such that we can feel like the director or cameraman as we go. Imagine The Matrix with a character selector and cam changer similar to video games. Yes, please! Right? But VR is also going to involve some pain. We need labs quantifying these boundaries and building limits into engines, and we need that starting two years ago!
Well, for "faux-VR" interfaces like the Wii you need to consider that it has to do a *lot* of procedural faking of response (gesture recognition), which adds a lot of extra lag since it typically can't recognize the motion and initiate it in-game until after you've already completed it in real life.
For more "hardcore" VR (well, at least not for most of the consumer-oriented stuff being done since the Oculus was announced, professional stuff has had different priorities) the problem is generally not that the lag is any worse than on a traditional PC - but that the same amount of lag is *much* more obvious. Your brain has a lifetime of experience correlating head motion and visual response to build a consistent map of the world, with continuous response times that make 60FPS look glacially slow in comparison. Insert more than the slightest amount of lag and it thinks something has gone very wrong and acts to correct it: the nausea/vomitting response - which helps correct the vast majority of lag-inducing effects in the natural world.
As for interaction: Consider - when you move a finger there's a roughly 1/8 second lag between when your brain sends the signal and when your finger actually moves - that's a hardwired signal propagation delay, but under most circumstances you'll never notice it because that lag is built into your brain's understanding of the interface and has been continuously updated as you've grown and the lag increased. Suddenly add another few milliseconds of lag though, and suddenly your brain is constantly saying "Hey, WTF!?! I just moved my hand, why isn't the damn thing respond... okay, NOW it's moving".
It's not that the interface is substantially more laggy, it's just that you're going from a completely artificial interface (button-presses, etc) to one that closely mimics what your brain already knows how to operate. Any discrepancy is going to throw it off-kilter. In fact I would suspect that hard-core VR enthusiasts are going to have some issues with real-world coordination. Maybe not to much of an issue if you're already a klutz anyway, but you may have to make a choice between being good at virtual or the real thing.
--- Most topics have many sides worth arguing, allow me to take one opposite you.
What if I told you that Valve's VR has been here all along and it was hiding in plain sight -- inside the Oculus Rift itself? Early on, Valve worked with Oculus to improve their head mounted display. The Oculus Development Kit 1 added Valve technology and became the Crystal Cove prototype. The prototype was re-released as the Oculus Development Kit 2. Oculus continued to receive assistance from Valve, but then in March of 2004, Oculus was purchased by Facebook.
For reasons not published, the cooperation between the two companies ended. Oculus went ahead and developed a Crescent Bay prototype, which was very similar to the best of Valve's prototypes at the time, but with the alternate camera arrangement that was used on Crystal Cove. The Crescent Bay prototype was not sold to the public, perhaps related to issues involved with the split between the two companies.
What you've seen with the Valve/HTC Vive is actually the culmination of Valve's ongoing research which Oculus has benefited from. After the split (and losing Abrash to Oculus), Valve continued to work on the hard problems and developed a new tracking technology based on lasers and inexpensive photodiodes, and controller input. The Valve/HTC Vive prototype is the latest public revelation of their ongoing work. It isn't any wonder that Valve's "new headset" has gotten high praise -- they've been breaking ground for some time, you just never knew.
We can expect both HTC/Valve and Oculus to evolve between now and the release of their first consumer product.
I just want to have an infinitely large computer screen with no additional cost beyond buying the VR glasses. I don't give even one tiny shit about the "virtual reality" part. I basically want to be able to set my virtual desktop to some ridiculously large size, then look around at it using the VR glasses as a viewport.
Then I can drag windows off to the side, look at something on the other side of my virtual desktop, and basically never have to maximize a window ever again. Bonus points if I can have a few layers of z-axis ordering to play with, where I can shove a low-priority window (like IM) and have it emit a glow or flash or whatever if it needs to get my attention. Then just use a fullscreen menuing system (start screen, anyone?) to launch things on the desktop. When launched, they should appear at center-of-viewport, then stay at that position on the virtual desktop until otherwise moved. A HUD would be useful for things that go in the task-tray (battery life, notifications, time, etc.).
For these tasks, a mouse and keyboard are sufficient. For games, well, existing mouse, keyboard, and/or controllers are just fine. Now, if you have an easy-chair and VR glasses, it might be nice to have a fully handheld or lap-based input setup, too, but that's really just a small matter once you have the desktop environment working.