Sony VR Headset/Project Morpheus/PlayStation VR

[MrFox]

Can HDMI port in PS4 [it is not hdmi 2.0 as far as we know] support 1440p @ 60hz output?

Sure their splitter will most surely be fully ready to manage 1440p bandwidth for their 120hz interpolation, but if the game is not rendered nativley at 1440p60... that would be bad for IQ.

Yeah, seems irrelevant somehow.

It's different from Occulus because of the external processor.

The PS4 outputs a normal rectilinear 1080p image, which you can see displayed on the TV with the pass-through. This source has the same image quality from center to corners. The external module scales it and deforms it into a fisheye projection, and it corrects the optical aberration. The maths of this projection makes it a 1:1 scaling ONLY in the center. At the edges and corners areas it's something like 2:1 scaling radial (guessing), so half the radial resolution is lost unless this area of the screen have twice the pixel density. And since the panel cannot have a variable pixel density, the only solution is to have the entire panel at a higher resolution.

That's the price to pay for using really cheap optics (pun intended).

 

[SlimJim]

I don't know that a difference in rendering resolution will be that noticeable with pixels so tiny and close to your face. My mobile has a 1080p screen and I can't make out individual pixels at the closest point of focal range. I imagine if anything it may reduce the 'screen door' effect, it that is still even noticeable at 1080p.

The optics in all VR-solutions work by wrapping the screen so that it covers a large part of 1 eye's field of view, as well as allowing your eyes to focus on it.

Also, instead of "1080p", 1 eye sees half of that; 960*1080. if you take an image about that resolution, and display it on your monitor, close 1 eye and move your head forward until that image covers most of your eye.
I am pretty sure you can clearly see individual pixels, as well as the 'screen-door'/ lines between the individual pixels.

 

[Scott_Arm]

I imagine games that are designed specifically for Morpheus would use the extra resolution for optimal image quality. It should be able to render at that resolution, with sacrifices to rendering "effects". Standard games that might be playable with Morpheus (Uncharted for example) would most likely have to reduce graphics quality a bit, and would then upscale.

 

[Scott_Arm]

It's different from Occulus because of the external processor.

The PS4 outputs a normal rectilinear 1080p image, which you can see displayed on the TV with the pass-through. This source has the same image quality from center to corners. The external module scales it and deforms it into a fisheye projection, and it corrects the optical aberration. The maths of this projection makes it a 1:1 scaling ONLY in the center. At the edges and corners areas it's something like 2:1 scaling radial (guessing), so half the radial resolution is lost unless this area of the screen have twice the pixel density. And since the panel cannot have a variable pixel density, the only solution is to have the entire panel at a higher resolution.

That's the price to pay for using really cheap optics (pun intended).

That's interesting.

 

[DieH@rd]

It's different from Occulus because of the external processor.

The PS4 outputs a normal rectilinear 1080p image, which you can see displayed on the TV with the pass-through. This source has the same image quality from center to corners. The external module scales it and deforms it into a fisheye projection, and it corrects the optical aberration. The maths of this projection makes it a 1:1 scaling ONLY in the center. At the edges and corners areas it's something like 2:1 scaling radial (guessing), so half the radial resolution is lost unless this area of the screen have twice the pixel density. And since the panel cannot have a variable pixel density, the only solution is to have the entire panel at a higher resolution.

That's the price to pay for using really cheap optics (pun intended).

Anti-fisheye projection is done in-engine. What external module does is:

A- creates quick 120hz interpolation of the source
B- creates pass-through image to HDTV that removes fisheye processing [end result is laggy image with lower resolution]

"B" is something that Oculus also uses in DK2.

 

[MrFox]

Anti-fisheye projection is done in-engine. What external module does is:

A- creates quick 120hz interpolation of the source
B- creates pass-through image to HDTV that removes fisheye processing [end result is laggy image with lower resolution]

"B" is something that Oculus also uses.

From their GDC presentation, interviews, and Sony's patents about the optical correction, the module does three things: Frame Rate Conversion (120Hz), Audio DSP, and Optical Distortion Correction, which the patents describes as correcting the pincushion caused by the lenses, meaning the PS4 output is uncorrected, therefore logically the PS4 output is rectilinear.

 

[Shifty Geezer]

From their GDC presentation, interviews, and Sony's patents about the optical correction, the module does three things: Frame Rate Conversion (120Hz), Audio DSP, and Optical Distortion Correction, which the patents describes as correcting the pincushion caused by the lenses, meaning the PS4 output is uncorrected, therefore logically the PS4 output is rectilinear.

Correcting lens distortion doesn't necessitate rendering rectilinear images. They can be rendered fish-eye like OVR to be stretched by the lens to fit, then processed to tidy up the result. A patent from Sony fixes CA caused by the cheap lenses. Optical distortion correction could just as easily be to unwrap the warped image for TV viewing as to warp the TV view for headset.

 

[MrFox]

Correcting lens distortion doesn't necessitate rendering rectilinear images. They can be rendered fish-eye like OVR to be stretched by the lens to fit, then processed to tidy up the result. A patent from Sony fixes CA caused by the cheap lenses. Optical distortion correction could just as easily be to unwrap the warped image for TV viewing as to warp the TV view for headset.

I might be relying too much on the patent, but it says they corrects the geometric distortion and CA at the same time, specifically the fisheye warp (which I think is really called a pincushion distortion, and the fisheye is the correction applied?).

The CA correction is done by applying a different geometric correction to each channel individually, because the refraction index of the lens is different between wavelengths. So the magnification, and geometric distortion, are a bit stronger/weaker based on the wavelengths. Because of that, it doesn't make sense to separate these two as different signal processing steps.

http://www.roadtovr.com/ps4-vr-hmd-sony-patent-optical-distortion-correction/

 

[MrFox]

Chances are, it won't be a niche if developers are seriously interested.

http://gamasutra.com/view/news/2200...Box_interest_wanes_says_GDC_Europe_survey.php

VR is a major emerging market, as interest in Steam Boxes and Android consoles falls away

28 percent of respondents expressed interest in developing for VR as the technology gains credibility. 26 percent of respondents are interested in making games for the Steam Box market, down from 37 percent last year. Just 7 percent of respondents are interested in Android consoles like the Ouya and Gamestick, plummeting down from 32 percent last year.

 

[DieH@rd]

Heh, more are interested in VR than in Xbone.

However, 95% of those interested are most probably indie teams. We shall see how will this interest hold in ~2 years when devs get clear overview how profitable VR games are and how large install base is on PC, mobile and PS4. For now, this interest is very much as the one we had for Ouya and Steambox. Lots of wishes...

 

[Deleted member 13524]

Heh, more are interested in VR than in Xbone.

However, 95% of those interested are most probably indie teams. We shall see how will this interest hold in ~2 years when devs get clear overview how profitable VR games are and how large install base is on PC, mobile and PS4. For now, this interest is very much as the one we had for Ouya and Steambox. Lots of wishes...

How are Indie teams interested in VR? Development costs for VR aren't free and most of the successful indie titles are 2D.

 

[3dilettante]

It's a survey of interest, and VR is the new hotness. It can also be that because indie teams are small, there can be more of them to add to the totals in a survey.

The small size of indie teams, the limited rate of big teams being created, the larger percentage of passion projects, and not having as much tied up in big corporate obligations and policies can also make them more prone to change focus with the hype.

 

[Shifty Geezer]

VR is a real gamble at this point. Who's going to target Morpheus with no idea when it'll launch or at what price? You've absolutely zero idea what the launch audience demographic is going to be, nor how many. Without a single basis for estimating returns, how do you determine and finance and product? Basically the most venturey of venture capitalists willing to take a gamble. OVR isn't much better. VR is very interesting, but I'd be very surprised if anything like as many devs are targeting VR as many of the other more predictaable platforms.

 

[3dilettante]

The survey question asks what items the respondents find interesting, not what they've committed to.

 

[SlimJim]

VR is super easy for indie developers if they already have a 3D game; the SDK takes care of almost everything; the only thing the developers need to do is play test their game because VR can change the feel or perception of dimensions for example.

Literally 1 man can port a game to VR in less then a week

 

[Shifty Geezer]

That goes very much against what Marks and Mikhailov have been saying, and why Sony are waiting to release VR with games that really matter to it rather than to release a half-arsed platform that could sour people's perception of VR.

 

[SlimJim]

I fully agree; VR should be it's own thing, and add to the experience, and whatnot.

But even if they wouldn't use the sensor input and just use it as a

Seeing how a driver setting can force 3d on about all 3d pc games ever released; I am pretty sure that even this would require minimal effort.
If I get the Sony Morpheus, I want to play every title with that thing on my head, even if it's just a 2d game. Like with the HMZ-T-series, it's immersive for all content.

 

[DieH@rd]

The survey question asks what items the respondents find interesting, not what they've committed to.

There are two other graph that ask about current and future commitments

VR is a real gamble at this point. Who's going to target Morpheus with no idea when it'll launch or at what price? You've absolutely zero idea what the launch audience demographic is going to be, nor how many. Without a single basis for estimating returns, how do you determine and finance and product? Basically the most venturey of venture capitalists willing to take a gamble. OVR isn't much better. VR is very interesting, but I'd be very surprised if anything like as many devs are targeting VR as many of the other more predictaable platforms.

At launch, both Oculus and Sony will have to court and "force with money" smaller developers into supporting their hardware. Sony will most likely have Witness, No Man's Sky, Until Dawn, Project CARS, War Thunder, Among the Sleep, SOMA, Eve Valkyrie, H1Z1... which is enough for a nice start.

 

[MrFox]

That goes very much against what Marks and Mikhailov have been saying, and why Sony are waiting to release VR with games that really matter to it rather than to release a half-arsed platform that could sour people's perception of VR.

I think 3D (in cinema) was pretty much the same problem, directors just put two cameras thinking that was all there was to it... and we had the most horrible vomit inducing 3D films in the early years. Now a lot of people are burned about 3D and it's too late. The 3D films are amazing when they understand what can and cannot work in 3D, but few directors care about this even today. (I think only Phil McNally knows what he's doing).

My hope is that early experimentation with VR is happening without pressure, and they will solve all the problems before it's on the market and there's no turning back. Supposedly there are many developers interested enough that they are working on VR in their spare time (spare time? where to they live?), I guess they will figure out a lot of do's and dont's, specially once Occulus is officially out.

The worst thing that can happen is a bad game that makes people think it's the hardware that causes the headaches.

 

[tuna]

I think 3D (in cinema) was pretty much the same problem, directors just put two cameras thinking that was all there was to it... and we had the most horrible vomit inducing 3D films in the early years. Now a lot of people are burned about 3D and it's too late. The 3D films are amazing when they understand what can and cannot work in 3D, but few directors care about this even today. (I think only Phil McNally knows what he's doing).

They made awesome 3d films in the fifties (M for murder for example). But I don't really think there is more to it than putting two camera on a scenes.