168 replies to this topic

[KuruptU4Fun]

Tykelau, on 14 March 2013 - 12:41 PM, said:

Try this then.

1) Be attractive.
2) Don't be unattractive.

To re-iterate:

[Roland]

Shumabot, on 14 March 2013 - 12:22 PM, said:

640X800 per eye isn't actually that low. That's effective 1280x1600, which is roughly 50% better than a standard high def television.

It actually is that low, because your field of view is not simply split into two sections.
It's not like you can display half of the screen for one eye, and half for the other, and end up getting an effective 1280X1600.

Your two eyes are both perceiving a view of the world that hugely overlaps. So the effective resolution is not simply the addition of the pixels for one eye to those for the other. Look up screenshots for the Occulus, and you'll see that the view presented for each eye is virtually the same image. (although the screenshots are generally shown at far higher resolutions than you see on the Occulus screens themselves)

So, effectively, what you have is a 3D view of a scene being rendered at 600x800. Which is a horrifically low resolution to play a game at, expecially given how close the screen is to the viewer. (although they use some optics to effectively increase the density of the pixels near the center of your view) And even the occulus folks themselves acknowledge this... no one is foolhardy enough to think this is even remotely good enough for a consumer version of this. They are going to need to at least double it.

Quote

As for the lag, every interview I've seen takes pains to explain, EXPLICITLY, that the lag is due to the fact that you're bouncing the image between three screens (rift to tv to camera) and that the rift itself has a latency that is in nanoseconds and is comparable with any other wireless control device. It has the same latency as your mouse assuming you don't live in wired ball mouse hell.

Folks I know who have seen it first hand have said that the claims that the latency is merely some artifact of monitors is not actually true, despite what the rift folks say. They've said that it does in fact have latency. While it's definitely better than many predecessors, it's non-zero, and in a fast moving game you may see some serious problems. Now, that was less recent than their most recent showing in Vegas, so perhaps it's better since then. But a few months back, the guy at Arstechnica, despite being hugely supportive of the technology and thinking it was awesome, also pointed out that it did make him feel kind of nauseous.

Even Luckey himself only claims that the Occulus can achieve a latency of 30-40 ms under PERFECTLY OPTIMIZED conditions, and not accounting for the actual display hardware lag. As you can see from Ciller's link to Carmack's article, as well as tons of other literature, you can actually perceive latencies down to around 20ms.

Again, I will be thrilled to be proven wrong. If they produce a consumer version of this, and it works as they claim it will, that will be awesome. But I will believe it when I see it.

In terms of Carmack, the guy is without question a genius... but he's also financially involved in this effort, so I think his opinions regarding this have to be taken with a grain of salt. That's not to say that his thoughts on this are without merit... that's certainly not the case. But he isn't completely without bias in this regard.

Ultimately though, it comes down to this:
We are consumers. This does not exist as a product. Thus, "supporting" it doesn't actually mean anything. MWO could maximally support the OR, and you would not be able to use it, because it doesn't exist.

[Bulu]

guys,guys..im playing this game in 3d (nvidia vision) and you cant even start imagine how awesome is. Im amazed that this days with all this tech at our reach people is still reluctant to play it in 3d or vr mode.If u like gamming,please, dont fool yourself with things like " its a gimmick" or "it will burn my eyes" or "it will give me headache"..********, nothing of it its true this days. ·3D its the biggest thing since the beggining of videogames,by far.by very very very far. you dont know what "immersion" means. u dont have a clue.

[KuruptU4Fun]

Your definition of immersion comes with a pretty hefty price tag for most casual gamers...

[Bulu]

your right.it has a price.as much as a good videocard like the ones the majority of us have.

[ColdCutz]

kuangmk11, on 14 March 2013 - 10:07 AM, said:

??? this is "Carmacks" VR headset.

Chavette, on 14 March 2013 - 10:16 AM, said:

What? Carmack is one of the guys behind Rift.

Oh god.


I'd buy it if MWO supported it. Maybe not the first gen though.

Carmack was fiddling around with VR hardware himself in his spare time much like Palmer Luckey, but when he saw Palmer's working prototype on MTBS3D forums, both he and Palmer brought it to E3 2012 to show off the rough build with Doom 3 BFG.

Although Carmack was the big name presenting it at the expo, he didn't want to confuse people about who created it, which was Palmer.

[KuruptU4Fun]

[xxx WreckinBallRaj xxx]

Hey Virtual Boy. Didn't think I'd see you again.

[Shumabot]

Quote

It actually is that low, because your field of view is not simply split into two sections.
It's not like you can display half of the screen for one eye, and half for the other, and end up getting an effective 1280X1600.

Actually, yes. Yes you can. That's how eyes work. They form a composite image from both eyes and mental processes makes up for the fact that both eyes have a significant number of blind spots, color blind spots, and areas of other failure. It creates an effective 1280x1600, that's the entire basis of glassesless 3d.

Quote

Your two eyes are both perceiving a view of the world that hugely overlaps. So the effective resolution is not simply the addition of the pixels for one eye to those for the other. Look up screenshots for the Occulus, and you'll see that the view presented for each eye is virtually the same image. (although the screenshots are generally shown at far higher resolutions than you see on the Occulus screens themselves)

Yep.

Quote

So, effectively, what you have is a 3D view of a scene being rendered at 600x800. Which is a horrifically low resolution to play a game at, expecially given how close the screen is to the viewer. (although they use some optics to effectively increase the density of the pixels near the center of your view) And even the occulus folks themselves acknowledge this... no one is foolhardy enough to think this is even remotely good enough for a consumer version of this. They are going to need to at least double it.

Nope. You don't know how the eye works. Human eyes create a composite image, this is the basis for most forms of holography. The human eye can and will take two low resolution images and combine them into one high resolution image.

Quote

Folks I know who have seen it first hand have said that the claims that the latency is merely some artifact of monitors is not actually true, despite what the rift folks say.

Yeah, I can lie on the internet too.

Quote

We are consumers. This does not exist as a product. Thus, "supporting" it doesn't actually mean anything. MWO could maximally support the OR, and you would not be able to use it, because it doesn't exist.

Sure. But honestly, it's not like they're actually working on anything else and this could at least get them some nice publicity other than "oh it's that mech game I heard was bad".

[Roland]

Quote

Actually, yes. Yes you can. That's how eyes work. They form a composite image from both eyes and mental processes makes up for the fact that both eyes have a significant number of blind spots, color blind spots, and areas of other failure. It creates an effective 1280x1600, that's the entire basis of glassesless 3d.

No Shuma, it doesn't.
It is true that with most glassless 3D solutions, you're employing something like a parallax filter, and essentially allowing each eye to see half of the pixels. But this has the effect of halving the resolution. Each eye can "fill in" the blank spots, but it doesn't do so in a way that is the same as actually having the same resolution as if both eyes were seeing all the pixels. This is why those solutions always look grainy.

This is actually why glassless 3D hasn't really been effective yet.. because they couldn't produce displays which were high enough resolution that, when coupled with a parallax filter, still achieved an acceptably high resolution. You may see some TV's this year though that pair such technology with the ultra high resolution 4K displays though. Dolby recently showed off one like this.

Honestly, you don't need to believe me. You can read any of the hands on commentary about this. Everyone, including the makers of the hardware itself point out that the current resolution results in a grainy image.

Quote

Human eyes create a composite image, this is the basis for most forms of holography. The human eye can and will take two low resolution images and combine them into one high resolution image.

You are mistaken.
While the human eye (in reality, it's your visual cortex doing this work) is capable of performing the types of image manipulation that you describe, this is not actually something that takes place with an application like the occulus. It doesn't just magically improve the image. Your visual cortex gets two images of a grainy scene.

Again, you do not need to take my word for this. Every single hands on review of the technology mentions this.

Quote

Yeah, I can lie on the internet too.

Shuma, why would I lie about it?
I mean, in all seriousness, could it possibly benefit me in any way?

And even beyond that, I pointed to actual statements by the developers of the technology THEMSELVES who cite the latency limitations of the hardware currently.


All of this aside though, like I said... prove me wrong! I would love it if they made such a technology and it worked as you imagine that it does. It would be great.

[rustyk]

If you take Nvidia 3d vision for an example, you really don't need anti-aliasing as much as you do in 2d.
I'm not sure about the comparison to a 'grainy image' but anecdotally, something is happening which seems to act as a resolution boost.
It might just be the distraction factor of seeing something come to life in front of you, but although we all want a certain amount of detail (resolution) in the image, once you see it stitched together in 3d, resolution becomes a secondary concern.

Edited by rustyk, 14 March 2013 - 05:12 PM.

[xZaOx]

Crazy, i posted this in 27 August 2012, and it died rather quickly, and suddenly its hot!

Would love to hear if PGI has looked into it at all.

[ciller]

I went hunting for Oculus Rift threads rather then post a new one, so here we are, thread resurrected! Like I said before, someone buy PGI a dev-kit model with a nice note with it saying they should have fun playing around with it and consider it for MWO.

Also to Roland, the consumer model will have a much higher resolution and even less latency then the dev-kit model. 30-40ms latency is more then fine, most people get higher latency over their internet connection. From the demos and responses I've seen, it is quite immersive and very very few people are complaining about latency or nausea.

I do not know why you are so adamant and pessimistic about modern VR, perhaps you jaded by the 90's versions but we are at a technological level today that VR can work and work well. Don't be a doomsayer and listen to your inner gamer, he wants this more then anything. Or should. Is your inner gamer dead inside?

[Dirty Old Man]

Any commercial product will be supported if enough demand comes from its sales pipeline... just make sure the price is below 300 USD with top notch stereoscopic flicker cross talk free images at more than 700p.... then its gonna explode....

[Roland]

ciller, on 15 March 2013 - 07:07 AM, said:

Also to Roland, the consumer model will have a much higher resolution and even less latency then the dev-kit model.

Of course. This is one of the requirements for a consumer model, because the current model is far too low resolution for consumers to buy. For a 5 minute tech demo, the novelty of such a cool 3D VR environment is enough to let you look past the deficiencies... but if you extend that to actual use, the novelty will quickly wear off, and the problems will come into the forefront.

The notion of making a higher resolution display is certainly not insurmountable. However, what may be insurmountable at this point in time is making a high resolution, low latency display, that is so small, and is cheap enough to be accessible by the mass market. That combination isn't something that you can just get by waving a magic wand.

Quote

30-40ms latency is more then fine, most people get higher latency over their internet connection. From the demos and responses I've seen, it is quite immersive and very very few people are complaining about latency or nausea.

I think that you were the one who posted Carmack's blog post regarding latency, and he points out that humans can detect latency rates that are over 20 ms.

Sure, your internet connection's latency is higher than that.. but you don't SEE that latency. The games you play smooth that out, so that you are not aware of it.

And realize, that the 30-40ms latency rate is in "perfectly optimized situations". And this statement came from Luckey, the creator of the Occulus himself.

Again, it's not an unsolvable problem... but I am highly skeptical that it'll be solved, on a high resolution device, in a cheap enough package to market to the masses, in the near future.

Quote

I do not know why you are so adamant and pessimistic about modern VR, perhaps you jaded by the 90's versions but we are at a technological level today that VR can work and work well. Don't be a doomsayer and listen to your inner gamer, he wants this more then anything. Or should. Is your inner gamer dead inside?

*chuckles*
Like I said, if such a thing gets made, it'll be wonderful. But simply because I want it doesn't make it exist.

I'll believe it can work when I see it myself. I certainly won't hold my breath for it though, nor will I demand support for hardware that I don't even own.

[BlackWidow]

Yes, Roland. You're very smart. Now shut up. (said with a smile)

It's very easy (on the Internet) to see developers over-hyping a product and rabid fan bois and grrls frothing at the mouth for it and wanting to jump in pendantically and feel all superior with your awesome insight as to what is real and what isnt.

And yes, your correct. It's still in development and isn't a commercially available product, yet.
And yes, this things have been said before concerning VR.

But, I have seen enough movie to look at a trailer for a movie that wont be out for more than a year, and still have a pretty good idea whether it will be crap or not.

Your later posts look like you may have actually read many of the articles concerning the Rift. But I wonder if you aren't reading with a jaded eye.

http://en.wikipedia....iki/Oculus_Rift

Everyone should read that. It's the quickest way to get an overview of what they may have in store for us.

[ciller]

Thanks Widow for the support! I do believe Roland is mostly right but whereas he is pessimistic on the future of the O.R. until he has a viable product in his hands, I am optomistic of what O.R. will be able to deliver.

Let us say me and Roland were investors. He would not be inclined to risk his money with the O.R. at this time for he feels that more then likely it will not overcome the obstacles he sees in the way. I on the other hand would invest a healthy yet not damaging portion of my wealth into O.R. for I feel that they will overcome any obstacles that are in the way and will be a succcesful commercial venture with a viable product for the masses that will have large appeal.
Both people in this situation are perfectly correct in their analysis despite the difference in approach. One takes a conservative approach and the other more liberal.

tl;dr
There is a certain chance of failure, I do not deny this. My instinct backed by knowledge and information plus my inner gamer leads me to believe that O.R. will be AWESOME. Roland feels otherwise.

[Matthew Craig]

Figured I'd share my thoughts quickly, first off to those saying it's a gimmick I don't think that's true saying VR has been tried many times before and failed is accurate, but you have to look at why it failed and if there's been any break-throughs in the areas where it failed.

I've also kept a close eye on VR tech and from my perspective 3 things holding back VR tech have been price, resolution and latency. Looking at the Oculus the price seems very reasonable and in range of what would be expected for a consumer device. The resolution while not perfect seems acceptable and Carmack has helped make significant strides on the latency so I don't think the word gimmick is fair, though whether or not it truly reaches the level required for wide spread consumer adoption remains to be seen.

We're certainly paying attention and I'm looking forward to attending Nate Mitchell's talk at GDC this year regarding the Oculus to learn more about the details. My expectation without having done extensive research but having read Carmacks article on latency, is that getting down to that 20ms latency barrier that really makes the whole experience work is going to be a significant challenge especially for games with a high level of visual fidelity.

What Carmack is talking about is really the beginnings of a new wave of renderers that operate fundamentally differently from what we currently use. Minimal rendering times have been important for many years but when your trying to attain the levels of fidelity of a modern game within 5-10ms you're really talking about a significant re-factoring of how your renderer fundamentally works.

Many modern engines still have renderers that are largely sequential, draw x objects in one pass, draw x objects in the next pass etc. we're starting to see though that to truly leverage the power of modern GPUs this has to shift to massively wide and parallel tasking which is likely the path to attaining these types of rendering speeds for complex scenes. This has already started somewhat with DX11 and is one of the reasons we want to move to DX11 as we already suffer from these sequential bottlenecks that are starting to be addressed and removed in DX11 and features like this one linked below open the door for developers to work around these bottlenecks.

http://msdn.microsof...s/hh994919.aspx

There's also significant support/development required in the display drivers themselves to ensure we don't just move the bottleneck to the driver level, where the game renderer is requesting large parallel workloads but the driver is still processing it sequentially.

Ultimately I don't get the sense that the Oculus is a gimmick but I do get the sense that it will require a lot of hard-work and similar breakthroughs from developers to truly unlock the full potential of the device which won't happen over night. I don't see any reason we shouldn't start though as the potential is very exciting.

[Shumabot]

Roland, on 15 March 2013 - 07:35 AM, said:

Shuma, why would I lie about it?
I mean, in all seriousness, could it possibly benefit me in any way?

The power trip? You're already pretending to know ophthalmology and are making up scientific facts. I have no doubt you'd make up other things too.

Edited by Shumabot, 15 March 2013 - 08:44 AM.

[Shumabot]

Quote

There's also significant support/development required in the display drivers themselves to ensure we don't just move the bottleneck to the driver level, where the game renderer is requesting large parallel workloads but the driver is still processing it sequentially.

Ultimately I don't get the sense that the Oculus is a gimmick but I do get the sense that it will require a lot of hard-work and similar breakthroughs from developers to truly unlock the full potential of the device which won't happen over night. I don't see any reason we shouldn't start though as the potential is very exciting.

That's phrased very well. Thanks for the response.