4 ms·

One thing the article doesn't get quite right is you can't compare ppd of a virtual display directly to an IRL screen. The virtual pixels are not 1:1 to device

One thing the article doesn't get quite right is you can't compare ppd of a virtual display directly to an IRL screen. The virtual pixels are not 1:1 to device pixels. Your perspective is constantly shifting slightly and the virtual pixels being sampled from is constantly changing. This is providing your eyes and brain a lot more data to sample from and means the perceived resolution can be higher than the actual resolution.

garbageman 3y ago
This is an interesting perspective. Genuine question: would it not be the other way - seeming lower than actual resolution when slightly moving? Why? Could this be used to simulate a higher pixel density? Maybe doing something crazy vibrating the display in some kind of circular motion or something silly?
- axoltl 3y ago
  There are projectors that do this trick! The technology is called XPR: https://www.projectorjunkies.com/4k-dlp-projectors-is-it-real-4k-xpr-technology-explanation/ https://www.projectorjunkies.com/4k-dlp-projectors-is-it-rea...
- layer8 3y ago
  The issue is that the original Mac pixels get 3D-projected and thus interpolated onto the VP panels pixel grid. This invariably loses some sharpness and detail. Even just 2D-rotating a Mac screenshot by some non-orthogonal angle will make it more blurry.
monkeynotes 3y ago
I am not convinced. Your eyes are literally magnifying the pixels compared to being at a distance from a modern 2k display. You may have dense PPD in AVP but being that close to them doesn't do it any favours. I mean, if you can see a pixel you can see a pixel, there is no getting around that. It's like once you notice a blemish you can't not notice it anymore. From what I hear from users is you can absolutely see the pixels.
- ninkendo 3y ago
  My 2¢, not an expert, etc etc... I think the key detail here is that each eye is getting its own set of pixels for looking at the same virtual content. This can lead to more detail, because the pixels aren't 100% redundant (they're both looking at the content from a slightly different angle.) If I'm looking through the AVP at a "square inch" of my virtual Mac display, it may be that each eye is getting say, a 100x100 grid of pixels in this angular area. But since each eyepiece is giving me a slightly different projection of that same inch of space, the pixels themselves are going to be of subtly different values, essentially contributing more information to what I'm looking at. This is a lot different from a "real" display, when both of my eyes are staring at the same physical pixels. I think the idea is that my brain will be combining this information into a perceptual image in a way that will appear slightly more detailed than the equivalent-sized pixels in meatspace. > It's like once you notice a blemish you can't not notice it anymore Interesting that you say this... because when you move your head towards a virtual object to inspect any pixels you saw, the image literally gets clearer (because you're getting "closer" to it) and you don't see the pixels any more. IME this goes a long way towards tricking my brain into thinking the pixels aren't real and therefore aren't there. (I've been using my AVP for real work all day today and I've been mostly very happy with it. The resolution is absolutely the least of my concerns, and it looks subjectively phenomenal to me.)
ShamelessC 3y ago
I feel like the article covered this.
crazygringo 3y ago
I've noticed this as well. I kind of think of it like when you look through a chain-link fence, there's a lot of visual data that gets blocked. But if you're in a car driving past, the chain link fence gets blurred out and you can see everything again. Something similar happens with the constant resampling happening in a VR headset. The hardware pixels are constantly shifting over the image pixels, as your head constantly moves just from breathing and whatnot. So not only does the signal-over-time counteract any blurriness from resampling in a single frame, but I find myself wondering if it enhances the resolution slightly. (After all, you can buy 1080p projectors that "jiggle" the image by half a pixel diagonally, several times a frame, to effectively double the resolution and claim it's halfway to 4K. And what is the constant movement of pixel resampling if not a similar jiggling?) I'm curious if there's an image processing term for this effect? The effective signal resolution from a constantly shifting pixel grid over time, and how to calculate it mathematically.
- CarVac 3y ago
  Multi-frame super-resolution. What matters is illuminated pixel aperture size, in the end. The smaller the pixel aperture, the less smearing you get from overlapping.
dwallin 3y ago
I put together a simple gif to roughly demonstrate the effect: https://giphy.com/gifs/JImONk8V5YK6iGBFkH/tile https://giphy.com/gifs/JImONk8V5YK6iGBFkH/tile While it still can't fully resolve shapes smaller than a pixel, it definitely helps with identifying forms.

garbageman 3y ago

This is an interesting perspective. Genuine question: would it not be the other way - seeming lower than actual resolution when slightly moving? Why? Could this be used to simulate a higher pixel density? Maybe doing something crazy vibrating the display in some kind of circular motion or something silly?

axoltl 3y ago
There are projectors that do this trick! The technology is called XPR: https://www.projectorjunkies.com/4k-dlp-projectors-is-it-real-4k-xpr-technology-explanation/ https://www.projectorjunkies.com/4k-dlp-projectors-is-it-rea...
layer8 3y ago
The issue is that the original Mac pixels get 3D-projected and thus interpolated onto the VP panels pixel grid. This invariably loses some sharpness and detail. Even just 2D-rotating a Mac screenshot by some non-orthogonal angle will make it more blurry.

monkeynotes 3y ago

I am not convinced. Your eyes are literally magnifying the pixels compared to being at a distance from a modern 2k display. You may have dense PPD in AVP but being that close to them doesn't do it any favours. I mean, if you can see a pixel you can see a pixel, there is no getting around that. It's like once you notice a blemish you can't not notice it anymore. From what I hear from users is you can absolutely see the pixels.

ninkendo 3y ago
My 2¢, not an expert, etc etc... I think the key detail here is that each eye is getting its own set of pixels for looking at the same virtual content. This can lead to more detail, because the pixels aren't 100% redundant (they're both looking at the content from a slightly different angle.) If I'm looking through the AVP at a "square inch" of my virtual Mac display, it may be that each eye is getting say, a 100x100 grid of pixels in this angular area. But since each eyepiece is giving me a slightly different projection of that same inch of space, the pixels themselves are going to be of subtly different values, essentially contributing more information to what I'm looking at. This is a lot different from a "real" display, when both of my eyes are staring at the same physical pixels. I think the idea is that my brain will be combining this information into a perceptual image in a way that will appear slightly more detailed than the equivalent-sized pixels in meatspace. > It's like once you notice a blemish you can't not notice it anymore Interesting that you say this... because when you move your head towards a virtual object to inspect any pixels you saw, the image literally gets clearer (because you're getting "closer" to it) and you don't see the pixels any more. IME this goes a long way towards tricking my brain into thinking the pixels aren't real and therefore aren't there. (I've been using my AVP for real work all day today and I've been mostly very happy with it. The resolution is absolutely the least of my concerns, and it looks subjectively phenomenal to me.)

ShamelessC 3y ago

I feel like the article covered this.

crazygringo 3y ago

I've noticed this as well. I kind of think of it like when you look through a chain-link fence, there's a lot of visual data that gets blocked. But if you're in a car driving past, the chain link fence gets blurred out and you can see everything again. Something similar happens with the constant resampling happening in a VR headset. The hardware pixels are constantly shifting over the image pixels, as your head constantly moves just from breathing and whatnot. So not only does the signal-over-time counteract any blurriness from resampling in a single frame, but I find myself wondering if it enhances the resolution slightly. (After all, you can buy 1080p projectors that "jiggle" the image by half a pixel diagonally, several times a frame, to effectively double the resolution and claim it's halfway to 4K. And what is the constant movement of pixel resampling if not a similar jiggling?) I'm curious if there's an image processing term for this effect? The effective signal resolution from a constantly shifting pixel grid over time, and how to calculate it mathematically.

CarVac 3y ago
Multi-frame super-resolution. What matters is illuminated pixel aperture size, in the end. The smaller the pixel aperture, the less smearing you get from overlapping.

dwallin 3y ago

I put together a simple gif to roughly demonstrate the effect: https://giphy.com/gifs/JImONk8V5YK6iGBFkH/tile https://giphy.com/gifs/JImONk8V5YK6iGBFkH/tile While it still can't fully resolve shapes smaller than a pixel, it definitely helps with identifying forms.