Fool me, this is real good shit ! [the VR thread]

[Barney]

To be honest, I'd just like to have an immersive (ie, fills my field of view) stereo effect without any head tracking. That would be a great start imo, and head tracking could be worked on while people play with that.

[Dio]

Yeah, there's a big risk if someone does it properly that someone else will do it very badly and cheaply and destroy the market for 10+ years.

[Barney]

I used to play games with LCD shutter glasses back in the day - and the effect was awesome. Quake 2 was especially good. There was no head tracking then and it's only a small step to go up to fully immersive stereo from there - and still no head tracking needed.

If they got that out on the market now, a lot of games would be able to take advantage of it (disregarding things like HUDs that are 3D flat objects that ruin the effect!) without needing API code to enable them. There's plenty of support for 3D drivers out there, and they'd be picked up again pretty quickly.

Adding head tracking would be nice, but it's not essential at the moment. For me, the most important thing is to get something out there before someone else, as Dio days, does it badly.

[gilesgoat]

To be honest, I'd just like to have an immersive (ie, fills my field of view) stereo effect without any head tracking. That would be a great start imo, and head tracking could be worked on while people play with that.

It's NOT enough.

It's like giving you a very nice lollypop, let you have a couple of sucks and then take it away.

This is what makes me "so angry" about the VR I've seen SO far .. "quite but not fully cigar" is "SO" close to be "almost perfect" BUT .. but it could well be "and we need 5 more years to fix what is SO close but not yet" ..

[gilesgoat]

Yeah, there's a big risk if someone does it properly that someone else will do it very badly and cheaply and destroy the market for 10+ years.

I can only say you are NOT the only/first one that said that, actually this *IS* "the biggest fear" or some, if someone like this comes out and "screws up the market" is going to screw up ALL the others that invested so much into it, there's quite a thing in "getting it really right or not getting it at all".

If people are going to loose confidence in the stuff coming out and/or it really does not match well at least "some" of the expectations then true.dat for probably other 10 .. 20 years you won't hear any more about VR again.

It's risky stuff .. it can be done right it MUST be done right.

Otherwise the screw up is big, IMHO "a cheap thing but that does not go very well" VS "something more expensive that does better" the first should have no place to exist.

We'll see if we'll end up in a story like VHS vs BETAMAX vs VIDEO 2000 or stuff like that, where the shittest of all the tech win only because it's cheaper.

It's crazy because "I know many know" that "getting it right" is the key .. at the same time I think "nobody really knows what IS/should be 'the right price' " .. and this is scary part.

I really would pay "something more for something that really does more" than to pay "something less for something that really does less".

At least give me 2 options with 2 prices, I'll wait, I'll work more, I'll save more but I'll get the one that does more.

[pogo]

If they got that out on the market now, a lot of games would be able to take advantage of it (disregarding things like HUDs that are 3D flat objects that ruin the effect!) without needing API code to enable them. There's plenty of support for 3D drivers out there, and they'd be picked up again pretty quickly.

I have fond memories of them too. I think though 3D is way more complicated than it was back then, when it was a matter of generating a textured scene, setting the camera (which the 3D glasses could intercept to render two slightly different scenes with two slightly offset cameras). That still happens but so much more takes place, in particular on the GPU.

So today it probably needs the game to do some work itself, outputting two scenes in a format a VR device can use.

[zip accident]

We'll see if we'll end up in a story like VHS vs BETAMAX vs VIDEO 2000 or stuff like that, where the shittest of all the tech win only because it's cheaper.

I understand what you're saying, but the thing about VHS was that it was (maybe only just) good enough.

[gilesgoat]

We'll see if we'll end up in a story like VHS vs BETAMAX vs VIDEO 2000 or stuff like that, where the shittest of all the tech win only because it's cheaper.

I understand what you're saying, but the thing about VHS was that it was (maybe only just) good enough.

Someone tought the same even with that Drop Vodka or stuff like that ...

[Barney]

To be honest, I'd just like to have an immersive (ie, fills my field of view) stereo effect without any head tracking. That would be a great start imo, and head tracking could be worked on while people play with that.

It's NOT enough.

That's just your opinion Giles, nothing more. The fact that I myself want just the 3D full-field display proves that it is enough for some people, and because it's enough for people like myself then I won't be surprised when some other company comes along and fulfills that desire.

[gumboot]

To be honest, I'd just like to have an immersive (ie, fills my field of view) stereo effect without any head tracking. That would be a great start imo, and head tracking could be worked on while people play with that.

Here's why I don't agree:


I think stereo is more of a gimmick that falls out of being much, much easier to implement with recording devices. For rendered content on devices more expensive than cellophane and cardboard you really need to re-evaluate your priorities.

[Jamesac]

Head tracking is little more than a gimmick in regards to a standard tv screen. With the Rift, it should be amazing. Thing is, there's no reason you can't use the Rift as a static screen for watching movies/tv or whatever. There's no law that says you have to use 100% of its features all the time. Alternately there's Glyph, which just came off a highly successful Kickstarter.

James

[gumboot]

Head tracking is little more than a gimmick in regards to a standard tv screen. With the Rift, it should be amazing.

Certainly that video is made of exaggerated movements -- and gentle, sweeping movements to hide the latency issue -- which makes it look gimmicky... and it's not such a great concept for staring through a little window like that, but it's a demonstration which I think makes a point.

It might be more interesting to see it with a monster-sized screen.


I've been starting to think that going back to raytracing might make sense, here. That way you can raytrace immediately ahead of the raster to get extremely low latency. You just need the means to put an upper bound on the time taken to compute each pixel, but I think you could do that by generating a simple world using a conventional GPU (with all its latency).

[Dio]

I've been starting to think that going back to raytracing might make sense, here. That way you can raytrace immediately ahead of the raster to get extremely low latency. You just need the means to put an upper bound on the time taken to compute each pixel, but I think you could do that by generating a simple world using a conventional GPU (with all its latency).

Read the Abrash article and it will be clear that raster is not a concept compatible with VR.

It's also clear that VR worlds need to be dramatically more complicated than conventional 3D worlds to avoid everything looking cheap and fake; notably, real geometry is required, bump mapping looks cartoonish.

[gumboot]

I've been starting to think that going back to raytracing might make sense, here. That way you can raytrace immediately ahead of the raster to get extremely low latency. You just need the means to put an upper bound on the time taken to compute each pixel, but I think you could do that by generating a simple world using a conventional GPU (with all its latency).

Read the Abrash article and it will be clear that raster is not a concept compatible with VR.

Well it would be an exceptional display device indeed that could do away with conventional raster-ordered scan-out.

It's also clear that VR worlds need to be dramatically more complicated than conventional 3D worlds to avoid everything looking cheap and fake; notably, real geometry is required, bump mapping looks cartoonish.

I thought it always looked cartoonish. You mean to say that it's substantially worse?

[Dio]

Unusable, apparently.

The raster problem is (another reason) why they're going to low persistence using a strobe flash rather than continuous illumination.