Laser projectors general? [zero lag & zero blur!!!]

[spacediver]

wow, they actually did show it with film projectors it seems!

And varsity cinema was one of them!! (both varisty and scotiabank are 5 min bike rides for me). Seems that it was only showing on these special projectors until the end of December, although I'll double check.

[Light23]

I'm at an impasse...



My speckle free screen only worked for back projection.
Until now.

After 3 weeks of messing around with various screen materials...I now have a speckle free method for front projection. (COMPLETELY different setup and screen materials. )

I can get bigger sizes than I could with the back projected monitor and my colors and contrast are better. (The scan lines are gone as well with this new screen.)

Sooooooooo...I'm not sure what direction I want to take this project now.

I could sell a pair of modified Sony projectors (Lens removed, different wavelength lasers put in for increased color gamut, brighter laser diodes, PBS cube, ect) and sell the front projected screen all as a package.

Or

I can just sell the screen.

Or

I could continue work on the back projected monitor, but the back end yoke will be long as hell.
(Not saying, just saying.)


You guys let me know...because right now I'm thinking front projection is the way to go. (Especially with HDR.)

Light23, can you share some photos of some of your work?

[TimothyLottes]

Front projection with a pair of devices at better than 60Hz with laser frequencies designed to minimize problems with metamerism would be top on my list. Don't care at all about resolution. Don't care if it is only 8-bit per channel per device, temporal dither both devices differently should solve that issue, and I doubt one can get ambient room light levels low enough with these small laser devices for bit depth to be a problem.

Some comments on prior discussion in this thread:

(1.) Consumer HDR is power limited at the display sizes pushed for TVs. LG is supplying all the OLED panels, all WRGB based, with very strong global dimming, and such blue white points that they get a lot darker after manual D65 calibration. OLED TVs don't get very bright, and yet they look way better than the competiting LCDs, thanks to black level.

(2.) The real win in "HDR" is better black levels, in part due to better anti-reflective screens. Actual contrast ratio in most consumer homes is massively limited by ambient light level in the room. I'm guessing it is hard to simultaneously optimize a laser front projection screen to minimize ambient reflection, maximize direct reflection at head on angle (aka the road sign effect), and remove moire. Guessing practical contrast will continue to be a major downside to these pico projector based laser setups.

(3.) The LG OLEDs I've seen suffer badly what is likely electronic compensation of variable pixel element long term decay rates. So blacks get something which could be described as "burn-in". This is even more visible than plasma thanks to ultra low OLED black levels. Can only guess "HDR" will only make this worse. Also the market focus on HDR might make it very hard to sell a consumer low-persistence OLED. Hard to know where this ends up, hard to know how long economy of scale takes to bring costs down.

I'm still looking for a CRT replacement.

Could be interesting building a diffusion screen for outdoor/indoor LED sign screens, something to hide the black dead space between mm spaced pixels. Unfortuntely seems like while LED modules are driven at super high frame rates, the video to LED hardware is all designed to strobe the screen multiple times per frame.

[spacediver]

light23, I meant photos of your actual hardware, that you have actually taken, please

[mirrorneuron]

bigger & better is the way to go

[RLBURNSIDE]

Light123, if your screen tech is solid, then PATENT IT IMMEDIATELY.

Do NOT share your designs on the internet before you do that.

THEN you can turn around and sell this screen to IMAX + other laser cinemas (which is a market that's bound to take off, including the home market eventually). And of course sell smaller versions for use with these pico projectors. Especially if you can add some gain to them to make the image brighter and without making the speckle visible.

A speckle-free screen that's semi-affordable (it also needs to be acoustically transparent, for commercial cinemas all the front speakers are behind the screen) for large areas would be terrific. From what I understand silver screens like they use for 3D movies (passive polarized 3D) increase hotspotting and would probably make speckle worse too.

Front projection is, and will always be, 100x more common and profitable than rear projection, especially in an age with larger and larger OLED panels (including transparent ones...tough to beat that with a projector). But front projection still has a good few years of life in it, at home. And the cinema who knows, some people say movie theaters are going to go out of business one day but that doesn't mean in the meanwhile they won't all be upgrading to laser-illuminated projectors, and if there is even minute speckle then maybe your screen tech can make you some dough.

I don't know how much money you could make with hacked pico projectors, it wouldn't be zero but it wouldn't be huge either.

[flood]

light23, I meant photos of your actual hardware, that you have actually taken, please

yea i'm not going to get my hopes up until i see actual pictures of anything. it seems odd that there's been so much discussion without any actual pictures of what the thing looks like.

[RLBURNSIDE]

Front projection with a pair of devices at better than 60Hz with laser frequencies designed to minimize problems with metamerism would be top on my list. Don't care at all about resolution. Don't care if it is only 8-bit per channel per device, temporal dither both devices differently should solve that issue, and I doubt one can get ambient room light levels low enough with these small laser devices for bit depth to be a problem.

Some comments on prior discussion in this thread:

(1.) Consumer HDR is power limited at the display sizes pushed for TVs. LG is supplying all the OLED panels, all WRGB based, with very strong global dimming, and such blue white points that they get a lot darker after manual D65 calibration. OLED TVs don't get very bright, and yet they look way better than the competiting LCDs, thanks to black level.

(2.) The real win in "HDR" is better black levels, in part due to better anti-reflective screens. Actual contrast ratio in most consumer homes is massively limited by ambient light level in the room. I'm guessing it is hard to simultaneously optimize a laser front projection screen to minimize ambient reflection, maximize direct reflection at head on angle (aka the road sign effect), and remove moire. Guessing practical contrast will continue to be a major downside to these pico projector based laser setups.

(3.) The LG OLEDs I've seen suffer badly what is likely electronic compensation of variable pixel element long term decay rates. So blacks get something which could be described as "burn-in". This is even more visible than plasma thanks to ultra low OLED black levels. Can only guess "HDR" will only make this worse. Also the market focus on HDR might make it very hard to sell a consumer low-persistence OLED. Hard to know where this ends up, hard to know how long economy of scale takes to bring costs down.

I'm still looking for a CRT replacement.

Could be interesting building a diffusion screen for outdoor/indoor LED sign screens, something to hide the black dead space between mm spaced pixels. Unfortuntely seems like while LED modules are driven at super high frame rates, the video to LED hardware is all designed to strobe the screen multiple times per frame.

Wow, Tim Lottes posting here, you're a legend man. We study / use your work at the office all the time.

I think raster-based laser displays have some ideal characteristics like low persistence, perfect blacks, etc, but the issue is that the brightness can't go very high before you hit the massive brick wall of FCC regulation. So that means front projection of raster laser is probably out.

However for smaller / lower power devices like VR helmets I definitely think there is a future there and especially take one of the negatives (narrow bandwidth) and turn it into a positive (high efficiency, 43% passive 3D that doesn't require a silver screen or polarizing filters, so long as each projector has an independent set of primaries which can be occluded by optical filters well enough).

I believe there have been some projection based VR helmets in the past but this is well beyond a home project. 60hz is too small, and 8 bits is terrible for static images. Of course temporal dithering sounds like a good hack to mask banding but I thinkNVidia's on to something with their statiotemporal superresolution demo posted in the other thread. If you can indeed double the input framerate with stacked projectors then the Sony Pico models could do 120hz and achieve higher resolution and double the lumens too. And of course boost the lumens and change out the lasers so that you can use cheap 6P 3D glasses taken from a Dolby Cinema and there you go, high efficiency passive 3D that works on any surface (including presumably Light123's innovations).

To fight metamerism I think a display based on painter's gamut using a bunch of independent lasers, one for each primary spread out evenly across the spectrum, would do the trick. Plus some mixing / color processing obviously. You could literally do the entire possible gamut of humanly visible colours that way, without metamerism. For computer games that could be insane, send your colors over the wire in float16 wrapped as int16 (for transmission over HDMI). XYZ color support has been implemented in HDMI since 1.3 I think.

[spacediver]

To fight metamerism I think a display based on painter's gamut using a bunch of independent lasers, one for each primary spread out evenly across the spectrum, would do the trick. Plus some mixing / color processing obviously. You could literally do the entire possible gamut of humanly visible colours that way, without metamerism. For computer games that could be insane, send your colors over the wire in float16 wrapped as int16 (for transmission over HDMI). XYZ color support has been implemented in HDMI since 1.3 I think.

Pointer's gamut was based on samples of real world pigments, so while it reflects a good estimate of the range of real world surface colors, it doesn't capture the full range of possible colors available to human vision.

How many primaries were you thinking you'd need to significantly reduce observer metamerism?

[flood]

pointer's gamut doesnt have much to do with metamerism