Sony VR Headset/Project Morpheus/PlayStation VR

A lot of people doubted Sony. Most, will never be known. Others, like Dan Mattrick, were pretty famous. But now they are fired. My advise, never doubt Sony: they always deliver :) I really believe they will pull off consumer VR. Maybe lionhead can create Fable VR for them, or Milo VR.. no, make that Mila VR(= Milo's older sister)
 
My advise, never doubt Sony: they always deliver :)
No they don't. They're the same as every other company - sometimes they do well, sometimes they don't. I'll just use on example here, of the concept video for what a future EyeToy HD would provide. They had augmented reality with an example of a guy donning a virtual helmet. I bought a PSEye for my PS3 on the promise of this vid, and it did squat after a handful of launch titles. It's absolutely in Sony's capacity to screw up VR based on past corporate form. However, given the current leadership and the way PSVR has been publically handled to date, I think they'll do a reasonable job. The only major concern for me is if the device gets proper software support or will it be another item thats put out there and then left to fend for itself. If the latter, and sales aren't fast enough to support 3rd investment, it'll die like so many other promising bits of kit we watched the past few decades.
 
It will be interesting ot see how sony keeps both sides of the fence happy. Those with vr and those without. They now have to support 2 platforms with enough games and looking at the last two times they attempted that , they failed pretty hard Egmon.
 
They're the same as every other company - sometimes they do well, sometimes they don't.

And especially so with these sort of emerging technologies that are born out of skunkworks side-projects that were never a part of any overarching business plan. There's too many moving parts in the VR equation right now for anyone to know where VR will be in 2-4 years, what the adoption rates will be, usage habits, content, social acceptance, so the idea that any company could be insulated from missteps or outright failure is nonsense.
 
I think YantraVR is Tony Davidson of http://innervision-vr.com/

So he is unlikely to be making stuff up, but on the other hand neither do I think Sony have any "special sauce" to deal with barrel distortion.

My best guess therefore would be:

* I think Sony do have something in their SDK for Hidden Area Mesh Rendering. I think this is YantraVr's special technique.
* Sony optical warp is less extreme than OR and Vive so you can use a 1.0x render target at the developers discretion, but still less than ideal.

Damm, it would be good to know for sure instead of all this second-guessing.

This was posted at NeoGaf, seems less distortion on PSVR to my eyes BUT without them rendering the same scene its easy to fool yourself: http://i.imgur.com/sMXBeWF.jpg

P.S. I'm not disagreeing with Mr Fox's statement above, just putting this out there to see if we can make sense of it
1.0x was a theory early on, but the warp is directly proportional to the improvement of the center's resolution. There's no way around this.

With a 90 degrees FOV and 1.4x resolution, the matching lens puts the same pixel density everywhere in it's field of view. The optical profile would be the reverse of an equidistant fisheye lens. If the rendered could do spherical projection, this would be an ideal system. Clipping that render in a circular FOV of 90 degrees should be great for efficiency. I think that's what sony is doing.

It's all the same maths as a fisheye camera lens, it's just applied in reverse. If we apply these for VR...
http://www.bobatkins.com/photography/technical/field_of_view.html
map3.gif

A: Rectilinear would have more resolution at the edges than the center. 1.0x needs no warping because the rendering is identical to the screen. Very difficult to make a lens rectilinear in VR. A good choice only for small FOV.
B: Stereographic projection would still have a bit less resolution at the center than at the edges. Most interesting for efficiency if we forget about fixed foveation.
C: Equidistant lens (spherical projection), same pixel density everywhere. 1.4x at 90 degrees. This is my guess. Should be the choice in the future once we get foveated rendering with eye tracking.
D: Equisolid lens, better center than the edges, this would be my preference but needs about 1.5x at 90.
E: Orthogonal lens, no idea what this, I guess it's enough warping that it would require a mandatory fixed foveation. Probably difficult optically. Now I want it.
 
Last edited:
No they don't. They're the same as every other company - sometimes they do well, sometimes they don't. I'll just use on example here, of the concept video for what a future EyeToy HD would provide. They had augmented reality with an example of a guy donning a virtual helmet. I bought a PSEye for my PS3 on the promise of this vid, and it did squat after a handful of launch titles. It's absolutely in Sony's capacity to screw up VR based on past corporate form. However, given the current leadership and the way PSVR has been publically handled to date, I think they'll do a reasonable job. The only major concern for me is if the device gets proper software support or will it be another item thats put out there and then left to fend for itself. If the latter, and sales aren't fast enough to support 3rd investment, it'll die like so many other promising bits of kit we watched the past few decades.

Let's also not forget Move controllers and games. Or the Sixaxis controller and support in games. Or their MP3 players. Or their MD players. Or...the list goes on. Good quality products, but inability to either gain mass adoption or willingness to support them properly.

In other words, as you said, the same as any other company in existence. Most of the past decade is also a good example of Sony struggling to find an identity in the changing electronics marketplace as they struggled versus their competition. They are starting to come back into form and figuring out where to spend their resources, however. PS4 was a good starting point. Focusing on their camera components was also a good move. It remains to be seen if they can redefine some of their other departments, but their current management is certainly better than their previous management.

Also it helps significantly when currency rates become increasingly favorable for your home currency which gives you a significant financial boost which can then be translated into more R&D and less cost cutting.

While for much of the mid-2000's to mid-2010's I felt Sony was an average or below average electronics corporation (which I hate to say as I've loved Sony since the late 1970's), I think they are coming back into good form. In some aspects they are back to being a top tier electronics corporation, but it's not universal across the company just yet.

Getting better, more work to do. PS4 is in a good place though.

I don't expect PSVR to succeed. But then I'm pessimistic and am expecting all VR solutions to ultimately not succeed. Hopefully I'll be proven wrong on all of those.

Regards,
SB
 
Depends what you class as deliver. I agree with shifys comment about over-promise on the eye toy however regarding the other things listed because they failed to gain mass market appeal (which Vita can be added to) - I don't think Sony 'failed', Move for example was a great product (which will now get a 2nd life).

They are like a lot of other big companies, some you win some you lose - but I think we're looking at a newer mindset from Sony and I think all they need to do with PSVR right now is get the price right (I personally think that was what killed Vita) - if they get that right and shift lots of units then it will get support.
 
1.0x was a theory early on, but the warp is directly proportional to the improvement of the center's resolution. There's no way around this.

With a 90 degrees FOV and 1.4x resolution, the matching lens puts the same pixel density everywhere in it's field of view. The optical profile would be the reverse of an equidistant fisheye lens. If the rendered could do spherical projection, this would be an ideal system. Clipping that render in a circular FOV of 90 degrees should be great for efficiency. I think that's what sony is doing.

...

A: Rectilinear would have more resolution at the edges than the center. 1.0x needs no warping because the rendering is identical to the screen. Very difficult to make a lens rectilinear in VR. A good choice only for small FOV.
B: Stereographic projection would still have a bit less resolution at the center than at the edges. Most interesting for efficiency if we forget about fixed foveation.
C: Equidistant lens (spherical projection), same pixel density everywhere. 1.4x at 90 degrees. This is my guess. Should be the choice in the future once we get foveated rendering with eye tracking.
D: Equisolid lens, better center than the edges, this would be my preference but needs about 1.5x at 90.
E: Orthogonal lens, no idea what this, I guess it's enough warping that it would require a mandatory fixed foveation. Probably difficult optically. Now I want it.

Thanks, useful information.

Getting a good "handle" on the graphics rendering is difficult enough without having to add in the optics. And with VR we can think of the graphics pipeline extending beyond the outputted frame to also include display characteristics and then the optical paths; so for gaming graphics at least its a new way of trying to think about this.

Hopefully when all the HMD's are out there some enterprising souls will start taking them apart and give us more information.
 
Real challenge for VR will be the software, it'd help if we had some idea of the development cost but the realist in me says to expect lots of relatively short tech demo titles which flood across all three platforms with questionable quality. IMO it's going to take time for truly compelling content to become common however the immersion is going to paper over lots of cracks along the way.
 
Valve just confirmed that SteamVR headsets will support Desktop Theater Mode for playing of traditional flat screen games inside VR space [VR cinema for PC games].

I wonder will PS4 be able to do the same for its TV game catalogue. If there are no resources to do it directly, maybe they can redirect live DVR stream into VR cinema application.
 
The processing box is quite powerful, at least seeking it's actively cooled..

Maybe it can make cardboard quality virtual theater. While the screen itself high quality stream from ps4
 
VR theatre works great and is impressive, but do realize that it will be a bit like watching SD tv at these resolutions.
 
if it's like SD, then every game can 100% certain support it on a technical/performance level: render game at 720*576, render this to a texture which is rendered inside the VR theatre (=simplistic geometry) @120fps
 
On 1080p cardboard it's still looks impressive.

But if its using ps4 itself, that probably will need a few additional work from Sony. Because some console games can use weird cheats. Or at least, it was like that on ps1 and ps2.
 
if it's like SD, then every game can 100% certain support it on a technical/performance level: render game at 720*576, render this to a texture which is rendered inside the VR theatre (=simplistic geometry) @120fps

No I mean that the pixel size being so close to your eyes and simulating a huge screen makes it look like SD in terms of the resolution and detail you can see.

On 1080p cardboard it still looks impressive because of its size, not resolution.
 
Last edited:
Of course I am! That's not my status update, by the way, it's a dear friend who happens to be a director at Sony, and has been directly involved with VR.
 
@Egmon83 class 1 ticket?

@London-boy so you are not flying?

-_____-

i was hoping you can give us feeble mortals more in-depth look of PSVR (take a picture every bit of it, including the view through the lens and the lens itself with macro camera)
 
Back
Top