They advertise the "4K per eye" as some kind of an advantage. They need to do this to have 3D vision, but it doesn't improve the resolution, it's still just 4K.
And the 4K is across your whole FOV, I would like to see a comparison with some standard monitors (4K at 27" and 32") at particular distances from your eyes.
I suppose because they didn't publish anything it is still significantly worse than hi-res monitors (just better than all previous headsets).
At 90-100 degrees, a 4K/eye display will likely end up being around 35-40 PPD (pixels per degree). This is about equivalent to what a FHD display would like like (27" monitor at 60cm: https://qasimk.io/screen-ppd/).
There are a lot of other factors like fill factor (screen door effect) and there are definitely some optical designs that can increase PPD as well. Apple bought Limbak recently, btw, it's possible they use a similar design as what is described here: https://simulavr.com/blog/ppd-optics/
Assuming 4k resolution per eye and 120 degree FoV would give 3840*2/120 = 64 ppd.
Using the link that you gave a 4k screen at 60cm is 73 ppd, so it is reasonably close.
Apple haven't said what the FoV is yet, so I'm guessing 120 based on the pi-max and an early impressions video from mkbd. The horizontal range isn't clear either: 23 megapixels would be about three 4k screens so it depends on the overlap region between the eyes.
It's not that far off 4k, and when sitting that close to a 28" 4k screen I find the pixels barely noticeable, i.e. I can see them if they are aliased but would not notice them when the colour gradient is smooth.
Based on your followup post, I think you're confused. Binocular overlap doesn't magically give you double the perceived resolution. If it worked that way, everyone would have a way higher PPD, but it doesn't and they don't. For a naive calculation on resolution you should just use the per eye resolution.
As mentioned resolution on the centroid may be higher depending on the lens design, but there's so much dependent on the rendering pipeline and the optics that I don't think it's worth doing more than ballparking. If you're interested in HMD displays, Oliver Kreylos has posted a lot on resolution measurements over the past few years:
The correct number is somewhere between a multiplier of 1 and 2 depending on the setup of the display. But 2x 4k screens is only 16 megapixels, and so it also depends on where the other 7 megapixels are.
I can't edit this comment after 5 days but anybody coming along and find this - the multiplier is very close to 1. Go up a bit and find the sibling comment with the four links. Very good explanation of the difference between the monocular binocular FoV is buried within.
Apple’s keynote showed an image similar to the 3-lens optical train at the bottom of https://simulavr.com/blog/ppd-optics/ and they have a screenshot of HN.
I think 4k might really be a bare minimum for this kind of thing. For the text heavy use cases that apple are pushing things like pixel density are going to be much more important than the gaming usecases VR has been used for thus far.
Once those virtual displays show text at any sort of angle the aliasing effects will be much higher too.
I think the tech might need to get to 8k+ before it starts feeling as good as "retina display" did when Apple launched that.
Why the wondering, there are already devices on the market with higher resolutions than what Apple will bring in near future. It may not be comparable for some high FPS use case, but as virtual desktop, especially crispness of image is very much comparable.
Different market, I think. The PSVR2 is aimed at gaming where resolution matters less but the demands on graphics hardware to push that resolution are much higher because they're rendering complex scenes with fancy visual effects and fast motion, all on a console that's limited in graphics power compared to higher-end gaming PCs. Projecting mostly flat windows with little overdraw is much less graphically demanding; the SimulaVR folks and others experimenting with spatial computing have even managed it using the integrated graphics on low-power Intel CPUs, though admittedly with not quite as high a resolution as Apple.
Yeah, I actually wonder if 4k ends up being as good as it sounds in practice for the use-cases Apple are pushing.
A 27" 4k monitor in real life should look better than a virtual 27" window pushing the same resolution within the 4k display of the headset. For stuff like text that's going to have a corresponding loss in sharpness, and because you could be seeing something at an angle there's going to be lots of weird aliasing effects. There's a reason why Apple starting pushing 5k monitors.
I'd guess the tech would need to get to 8k+ for each eye before it starts feeling "really good" for these use cases.
I feel like they have their "game porting tool" out now, because they want to drum up games for visionOS, so it's likely that they'll also do gaming if they can.
The PSVR2 also does foveated rendering, which the Vision Pro does not seem to do, which is about the biggest disappointment about this. I don't care about millions of pixels in the corner of my eye eating up my battery life when I can't even see them.
It's not 4K per eye. They say 23M pixels across two eyes, which is 11.5M for each eye. 4K is 8.2M. It's 40% more pixels than 4K, with the intent presumably being to be able to display a 4K-resolution display within a larger space.
And the 4K is across your whole FOV, I would like to see a comparison with some standard monitors (4K at 27" and 32") at particular distances from your eyes.
I suppose because they didn't publish anything it is still significantly worse than hi-res monitors (just better than all previous headsets).