Sony VR Headset/Project Morpheus/PlayStation VR

Can someone please describe what a 85 degrees FOV looks like compared to, say, how far a 50" TV would need to be?

Assuming you had one eye at the centre of your head, it should be about 59 cm away from a 50" tv. With two eyes increasing FOV it should be closer than 59 cm away. I think.
 
It's from the TV out looking at you, not from you looking at a screen, so I'm not sure it makes that much sense to compare. But using the PS3 Eye which has 75 degrees FOV, I have to be at 1.7m for Move Fitness, which needs to track my movements in a square of about 2m x 2m. So how far away you need to be and how much space you need all depends on what you are expecting to do. For a driving game that's obviously going to be very much smaller, as for any games that track your head and your controller only. Those should be playable from 50cm easily, as you can just aim the camera at your head and controller, which would be in a much smaller square.
 
Ohhhhh this waiting is excruciating.

It's like when I was waiting for thesis score result all over again.

Ba dum, dum dum.

Btw how long PS move can be alive? Worse than Ds4? (around 5 hours I think)
 
Better than DS4, generally. And my Move controllers are ancient. They just have a bigger battery I think.
 
It's from the TV out looking at you, not from you looking at a screen, so I'm not sure it makes that much sense to compare. But using the PS3 Eye which has 75 degrees FOV, I have to be at 1.7m for Move Fitness, which needs to track my movements in a square of about 2m x 2m. So how far away you need to be and how much space you need all depends on what you are expecting to do. For a driving game that's obviously going to be very much smaller, as for any games that track your head and your controller only. Those should be playable from 50cm easily, as you can just aim the camera at your head and controller, which would be in a much smaller square.
Cool but I really just meant how much of your vision it covers compared to a 50" screen at x distance. So if function's calculation is correct (59cm), that's pretty darn close.
 
The 85 degree FOV is for the camera watching the player. For VR, that'd feel quite tunnel vision. PSVR's FOV is 100 degrees. It's not really comparable to watching a flat screen at a short distance because the view needs to extend well beyond the centre of attention.
 
2) Reprojection has limitations which is in your 3rd bullet point , this doesn't sound that great

3) Same as above. To much marketing speak
You are mixing two things...

Reprojection is better left on, in all cases.

There are limitations when rendering at 60 with a scan out at 120. Some games will need to render at 90 or 120. Reprojection is still enabled.
 
Reprojection will still introduce artifacts sony even talked about it during their last talk about ps vr and we all discussed it then.
 
You are mixing two things...

Reprojection is better left on, in all cases.

There are limitations when rendering at 60 with a scan out at 120. Some games will need to render at 90 or 120. Reprojection is still enabled.

Reprojection will still introduce artifacts sony even talked about it during their last talk about ps vr and we all discussed it then.

My understanding is the that Sony's re-projection is basically the same as Oculus's Async Timewarp. Which as I understand it, is also always left on on the PC as it reduces movement to display latency even when running at full frame rate. So no difference here.

60->120 re-projection is obviously an extreme use of the technology though, and as a consequence comes with more extreme side effects.
 
Reprojection will still introduce artifacts sony even talked about it during their last talk about ps vr and we all discussed it then.
Under the same circumstances, it's invariably worse to turn it off. Whether it's rendering at 60, or 90, or 120.

The artifacts are mostly caused by the reuse of the same frame for two consecutive scan out at 60->120, the objects didn't move, causing a double image with fast object movements, close to the camera.
 
Well it's not in 2D games so would be interesting to see what's different in 3D benchmarks.

To take a real world example of this, Digital Foundry's latest performance comparison uses Hitman to compare unlocked PS4 performance to the Radeon R7 360. which is more or less an identical GPU architecture to the PS4, just slower.

In terms of speed the PS4 compares to the R7 360 as follows (assuming the R7 360 is running at it's full boost clock):

Pixel Fill Rate:
152%​
Texel Fill Rate:
114%​
Geometry Rate:
76%​
Memory Bandwidth:
148%​
Shader Flops:
114%​

Performance wise in the Hitman video, it seems to vary from around 16% - 30% faster most of the time (usually at the lower end of that scale).

So claims of 60% faster than a similarly specced PC? I wouldn't be at all surprised if that's at the headsets native resolutions. i.e. PC running at 2160x1400 and PS4 running at 1920x1080. Or possibly assuming a PC running a Jaguar CPU and only 8GB of RAM total. Or just pure make believe.
 
It's unlikely to be make believe. This is a presentation for developers, not fanboys, informing them of what they have to work with. We've heard a little about PSVR being able to do things PC can't (at the moment) maybe via compute and efficient rendering of dual images or something? So it's probably, IMHO, a 'ball-park' figure for devs working on PC with a certain spec wondering what they can achieve on PS4 if they port. Probably includes the different resolutions, so a PS4 running PSVR is probably 60% faster than the same spec PC rendering to OVR, sort of thing (haven't watched presentation to see!). That's reasonably meaningful for developers wondering if PS4 is a viable target. Especially with the shock of the prices for OVR and VIVE and maybe worrying a bit about whether there'll be a decent sized market for their VR project. It'd be important for such devs looking at PS4 specs and thinking it's woefully inadequate to be informed what it's capable of in real-world performance and whether it's powerful enough for their project without too many sacrifices.
 
The headset is not a rectilnear projection, so the raw resolution of the display doesn't tell us the whole story. It's impossible to compare without knowing the center arc-resolution (which nobody leaked yet).

The 60% in the presentation wasn't exactly about PSVR, it was about consoles vs PC in general. Richard Marks also said this was not his own benchmarks, but a number given by middleware developers. It makes sense for multiplatform middlewares to provide an accurate number.

There must be a lot of variability between titles. If we'd use Bethesda ports as a reference, we'd think the PS3 was on par with an Atari 2600.
 
Middleware can be far more optimised for closed platform consoles than PC. Although I think we've seen occasions where they clearly haven't been as well.
 
What kind of middleware though, and under what kind of circumstances? In current (mostly GPU limited) multi platform games we aren't seeing that kind of gap.

The devil is in the detail, as is often the case.
 
Reprojection will still introduce artifacts sony even talked about it during their last talk about ps vr and we all discussed it then.

Yes, there are some [non deal-breaking] artefacts in specific scenarios, but reduced latency seems to be well worth it.
 
Back
Top