Sony VR Headset/Project Morpheus/PlayStation VR

Does it require more GPU power than regular 1080p?

 

No. 1080xrgb is a display type. The game outputs 1920x1080 pixels regardless of display. The display then turns that into little dots of light to create a visible image. 1080p Pentile and 1080p rgb are taking the same resolution from the game and drawing it in different ways, RGB having more fidelity for the same display resolution.

 

But what decides which colours need to be displayed within the pixel? Not the gpu

 

The game provides the value or color number for the pixel, the display then decides how the pixel is to create/approximate that value/number.

 

On-board hardware on the display. That hardware can sometimes be overclocked so that display can work at higher refresh rates.

 

It's part of the display circuitry. The gpu still needs to render full rgb pixels.
The display controller spreads the missing colors to adjacent pixels.

 

No & that's probably the purpose of Sony using RGB 1920 x 1080P instead of RGBG 2160 x 1200 like Vive.

Less pixels to push but more sub-pixels on the screen. Morpheus will have 6,220,800 sub-pixels & Vive will have 5,184,000 sub-pixels.

 

That's an interesting theory. I'd like to see both screens side-by-side, I can't imagine the Morpheus looks noticeably better.

Maybe it's also worth considering the number of pixels/second of these devices.

 

The optics quality and design choice should make a good difference too. There's a balance between the density of pixels in the center versus the corners (the fisheye projection function that they decide when designing the lenses), and sony changed it between the first and second prototype. There's also some kind of diffuser that can help reduce the visibility of pixels. It's all compromises.

 

Pentile vs RGB is very subjective. You'll get Samsung fanboys claiming it's better than RGB and vice versa. Indisputably, not having all three colours for rendering a pixel value means Pentile is compromised. Whether it's an inferior appearance or not is subjective.

The Pentile screens I've seen/owned haven't impressed in pixel arrangement. At ultra high densities they work, but expanded to full screen, I imagine it's a compromised choice.

 

I tried everything on my unit, never got them to work more than a few minutes without having to point almost 70 degrees to a different direction to actually point to the center of my screen.
The Killzone thread on the official forums is a good indication of the problem, and none of the suggested solutions made any difference.

 

commenters on ars said that morpheus may be using OLED with tighter "gap" between pixel and bigger pixel size. Thus decreasing the screen door effect although only runs 1080p.

 

Sound like it might be Stacked OLED but I would guess that Sony would have made that clear. Then again maybe that's what they meant by 1920 x RGB x 1080.

 

The most important thing, all journalists who tried morpheus were satisfied with both resolution and performance of PS4. Those two points are very important for Sony if they want to promote this new platform to the customers.

 

Here are the pixels per second for each of the VR headsets as we currently know them to be:

Oculus Rift CV 1:
2560x1440 @90hz = 331,776,000 (pixels/second) - pentile

Morpheus:
1920x1080 @120hz = 248,832,000 (pixels/second) - RGB

Vive:
2400x1080 @90hz = 233,280,000 (pixels/second) – pentile(?)

This doesn’t take into account of the sub-pixel arrangement, especially considering that I’m not so sure of the difference it will make. Maybe it’s worth doing a similar calculation to see whether the difference between Oculus Rift and Morpheus normalise to some extent.

It actually shows the Vive to be of the lowest spec if we consider hz as an especially important factor with these VR headsets.

 

From what I understand, every third sub-pixel on a pentile display is shared. So wouldn't you take the total pixels and multiply by 2.5 rather then by 2? The Morpheus device would simply be multiplied by 3, once for each sub-pixel.

6,480,000 for Vive instead of 5,184,000?

Edit: clarity.

 

onQ was counting subpixels, not pixels. Doesn't matter how many subpixels make a pixel if you're comparing subpixel-counts.

 

I love how Sony managed to start this counting subpixels thing years ago (when they started saying their TVs had 6million subpixels and therefore better than other 1080p TVs, something that's completely false), and that now we're still counting subpixels.

Gotta hand it to them, they're pretty genius.

 

So you're saying it's irrelevant or simply not true in the context of VR glasses/display technology?

 

I think Sony could actually be onto something, I've been looking through images of closeups of a pentile display and it seems that the 'screendoor' effect could actually be due to this screen type and not necessarily a pure resolution issue.

These displays weren't designed to be a couple of inches away from one's face, so it's less of an issue for mobile phones. Now we've got these VR displays that are very close and the effect is probably exacerbated with the lenses. Oculus have been pushing for higher and higher resolution and the highest when first available is usually pentile, then later switched to standard RGB (as is the case for mobiles).

I can't help but think it makes sense to rather use a standard RGB screen for VR.