Optical Smoothest yet not specular metal or plastic sheets?

[nwytg]

Where did this requirement come from?

Because the screen viewed through a stereoscope or head mounted display is viewed through a magnifying lens, that's quite different than viewing something from up close because basically the image becomes a virtual image the size of a large wall appearing meters away from you and any texture on it will get scaled up accordingly. 50mm is stretched over 90 degrees of your eye's field of view.

This I don't understand.
Is it something to do with the setup of the optics?

If one projects an image onto a screen or wall, what one is seeing is the reflection of light from the screen

Screens and most walls cause diffuse reeflection, not specular. With specular reflection you either will see the projection lens only or a bit of diffuse reflection mixed with that. "Artifact of the lens" is you seeing the lens with light scattering inside the lens barrel and illuminating it, projecting slightly on the lens itself because no AR coating is perfect. This appears to our eyes as glare as is blurry and slightly bright. If we had perfect optics with 100% pass AR coating and light not scattering inside the lens assembly we would still see something, the DLP micromirror chip itself illuminated. That's how the Avegant Glyph device works.

As the screen imperfections become more finer, it turns more and more into a mirror.

True, the question is how fine we can get without seeing the reflection of the projection lens and our eyes on the screen itself. I see no way to know this without testing etching or somehow texturing the screen such as micron sized glass beads on a surface acting as retroreflectors.

Silver screens made of silver have been used in cinema and still metal-containing screens are used, but they definitely are not pure mirrors although they have that glare from the lens (called a "hotspot") which can be reduced in software or a painter or somehow applied gradient mask.
Some cinema screens use simple white or grey matte or slightly glossy 1.0-1.2 gain screens which has no hotspot or limited viewing angle before it gets t0o dim but that in turn requires brighter projector to achieve same brightness as silver screens.

What I'm basically trying to do is figure out how much the textures on cinema silver or matte screens can be reduced without ending up with a pure specular mirror. Although I don't need to use silver and can use black color to improve contrast because the screen on my stereoscope is 100-1000x smaller than video projector screens so brightness doesn't matter which is one of the reasons silver screens are used in cinemas, to improve gain (brightness) in cost of viewable FOV. The pico projector I have can produce 300 ANSI lumens which is too bright as is for two 50x50mm screens.

 

[256bits]

Because the screen viewed through a stereoscope...

Ok. Thanks for the explanation. I think I get it now.
You need something with imperfection size a factor smaller than the pixel size of the image.

 

[nwytg]

You need something with imperfection size a factor smaller than the pixel size of the image.

I'm not sure it needs to be if the pixels were very small. But it definitely needs to be smaller than the human eye can detect. So for a 20/20 vision which is 60 pixels per degree roughly 50mm stretched to 90 degrees gives 1/108th mm or about 9 micron (50mm / 90 deg x 60 pixels) grain or scratch(?) size I think? (this is of course ignoring lens distortion messing with the pixels per degre value).

If at 9 micron the individual pixels can be seen I figured 5um is a safe bet to make them blend together. However 9 or slightly bigger may trick the brain into blending them as well. I don't know and there doesn't seem to be readily available research on exactly this so best bet is probably building such a screen and testing.

 

[CWatters]

I might be out of the ball park but some tablet and mobile phone cases have a smooth matt finish. The one on my phone has some texture but I can't see any on the one protecting my tablet. It's a plastic with a slightly rubbery feeling to it.

 

[nwytg]

I might be out of the ball park but some tablet and mobile phone cases have a smooth matt finish. The one on my phone has some texture but I can't see any on the one protecting my tablet. It's a plastic with a slightly rubbery feeling to it.

Thanks for the idea
I checked 4 matte silicone cases for iphones. Thw white matte one almost didn't appear to have any apparent texture after projected onto, same with gray, the two black ones have some speckle shine on them. My guess is all of them do just the white shine on the brighter ones is less apparent. Maybe something to do with size of colorant dye particles and/or some of them acting as retroreflectors.
But it's pretty darn close to what I hoped for.

I cut one up and it has a rubbery layer adhered to a more rigid plastic layer. Don't know if it really is silicone or that's just a marketing term. I've used elastic silicone sheets before for decorative uses and also some silicone covered frame resistant components in 3d printers, both cases silicone was pretty glossy.
Any ideas what's the actual material uses in these or how to find out? Maybe I can find a black sheet with a different colorant with less glow to it if I know what material it is.

 

[skystare]

15,000 grit sandpaper, available from automotive paint supply stores, produces scratches in the 1 micron range, just large enough to disrupt specular reflection of visible light. Cheap and easy to try, just start with full gloss plastic and wet sand. Might have to go with 12,000 grit, but even that shouldn't show when projected. DIY telescope supply stores carry superfine polishing grits as well, but are thinner on the ground than auto paint shops.

 

[CWatters]

Wow. I've seen some 3000 and 4000 grit and thought that was pretty fine. Apparently some people use it to polish chisels after sharpening the edge. Never knew you could get as fine as 15,000 grit.