Yellow Cyan ShowWX+ display?

[tao]

Hello all!
I wanted to run an idea past you to see if it would be feasible.

I am combining 5 showwx+ laser pico projectors to create a brighter image.
I also wanted to create a wider color area of display by using a yellow laser.
One of the problems here is that adding a fourth channel would require a 4-dimensional color space to fully utilize that extra channel. To really utilize this sort of new feature would require a whole new image recording system which I am working on. (The ideal would be that all color information in video would be in device-independent xy color space instead of RGB.)


Is it possible to replace one of my red laser diodes with a yellow and replace my blue with a 490nm cyan laser diode for better color control? My other laser projectors would be displaying RGB as well, just one or two would be displaying yellow for green/red and cyan instead of blue. (A turquoise primary would increase the RGB gamut significantly.)

Y in YUV isn't Yellow, it's Luma. Still, the imperfect conversion between YUV and RGB means that a fourth primary could make it possible to more accurately show YUV data on an RGBY display.

(The forth color is useful if the signal is coming in as CYMK, but its not, and it isn't switching to CYMK internally, its RGBY which is some completely different color model.)

I know the showwx+ analog drivers are ultra fast but was wondering if there is a yellow diode (and cyan) I can get to work with this application and its on board drivers.

Thanks!

 

[ZRaffleticket]

You can get yellow DPSS at 589nm... you're looking about a grand for a 100mW lab unit
You can get a cyan ~490nm laser diode at around 200mW for "between $5k and 10k". I'm afraid to figure out the exact figures...

The only practical purpose for those two being high powered diodes would be for your purpose... and that said, it's not that practical at all...

 

[tao]

So will a yellow DPSS diode plug in play with the DPSS green I have in the showwx?
If so, I can just replace the green. Neat!
I think I can find a used 490nm laser for cheaper than 7k. But does it need to be single mode like the blue in my pico projector to work with the onboard drivers?
I am assuming yes.

What brands of diodes would you suggest for this project? And will they be the same small size that goes into the showwx+? If not is there a way around that?

 

[Things]

You won't be able to use just any DPSS laser in this. DPSS lasers _hate_ modulation, especially analog and high speed, they tend to mode hope like mad and you'll get visible artifacts in your image. The Corning green modules used in the projectors were specifically designed for these high speed applications, they use a somewhat different process to the usual green DPSS. I don't think you'll be able to find a yellow laser that'll meet the required specs, unfortunately.

 

[tao]

Oh bummer about the yellow.:yabbem:

What about a 490nm cyan diode instead of the blue diode in one of the projectors? Would that work?

I want to use the Sony DSC-F828 with 4-Color RGBE CCD to film a special color movie for playback on a homemade RGBE small rear projection laser based (showwx+) "TV".

Most normal cameras only use R/G/B subpixels to make up an image, the F828's sensor uses four colors, Red, Green, Blue and Bluish-Cyan.

If I have a 4-color "RGBE" (red, green, blue, 'emerald' -- actually more cyan) display method it will increase the color range (gamut) towards the center of human vision: 555nm, and match the RAW data coming from the camera.

I know RGB covers almost all colors, but adding cyan allows linear combination to cover more of the possible stimulation. It is primarily the near UV purplish blue below 440 nm and the warm reds near IR that can not be reproduced. (It would be cool to have one pico displaying Cyan, UV, & I.R. overlapping another pico doing Red, Green & Blue.)

By using cyan or near 490 nm (blue green) creates a better white and black image also.

EDIT: I was wrong about yellow but right about cyan!
The only way more primaries could expand the presentation of an image is if it expands the color gamut. If the Cyan used is within the Blue-Green primary gamut of the show WX+, then it is pointless. But it turns out it's not...

490nm all by its lonesome...

Yellow is represented VERY well in the showwx+ gamut, so it would not make sense to add a 589nm yellow primary color.
But 490nm Cyan DOES make sense as it falls outside 450nm blue and 530nm green. (And 442nm & 532)

One of the things that has been toted a lot about Laser based projectors is the 'larger' number of colors. The issue is difference between primary colors and secondaries that fill in the color space. The primary colors are VERY vibrant, however the spaces between has a lot less energy. It is like having a very loud sound system that only have 32 narrow bands of power. Yes it is loud, and can extend from 20-20k possibly, but that doesn't make a nice flat response. The level difference between the RGB spikes and the spaces is around 90-95%. Yes that may make the image look punched, but it has to be. The color spaces that are not covered end up over saturated because there is nothing else the projector profile can be programmed to do. Proper multi color laser projectors need to have filters on the peaks and start to include secondary emitters to ensure a flat base light before the imaging section of the projector. So that means the inclusion of not only RGB, but CMY as well.

All this may be moot since the showwx+ is 16bit color...

But on the other hand anything that uses JPEG compression is guaranteed to not represent true color because of the 16 million colors in the 24-bit true-color pallet, JPEG can only store about 2.3 million colors. (That's about 14% of the available color space!) This includes MPEG videos, AVI, DVD and Blu-Ray videos, and High-Definition TV. Add to the fact that most laptops screens have a 1:1000 or lower dynamic range, and while some do have better displays, most use a TN-LCD which may only deliver 6-bits per pixel, rather than the 8-bits (for 24-bit color) that you normally expect.

The altered 4 color Showwx+ would still look better at color reproduction at 16bit back projected closely on a high gain screen at laptop screen sizes than most conventional viewing options when you take all these other things into account.

So could a cyan (490nm) single mode replace the blue single mode in one of my showwx+ projectors for this purpose?
Does anyone even make a 3.8mm 490nm cyan diode?

 

[tao]

(BUMP)
Does anyone even make a 3.8mm 490nm cyan diode?:thinking:

 

[tao]

Ok I have made a breakthrough.
I am buying an Osram 60mw 488nm cyan laser diode to replace one of my blue diodes in the showwx+.
The diode is the plt5 488.
It has high modulation bandwidth and can do cw and pulsed operation.

The only downside is its a TO56 package instead of a TO38 package.
(I think it should still work though since the optics do there magic BEFORE it hits the MEMS mirror.)

I would clone the displays and run a RGB projector and coverlap that with a RGC projector.
It should create the widest color gamut display ever created.

Does anyone see any problems with this before I lay down the cash to test this theory out?

 

[crazyspaz]

Aren't those several thousand dollar diodes?

 

[rhd]

Ok I have made a breakthrough.
I am buying an Osram 60mw 488nm cyan laser diode to replace one of my blue diodes in the showwx+.
The diode is the plt5 488.
It has high modulation bandwidth and can do cw and pulsed operation.

The only downside is its a TO56 package instead of a TO38 package.
(I think it should still work though since the optics do there magic BEFORE it hits the MEMS mirror.)

I would clone the displays and run a RGB projector and coverlap that with a RGC projector.
It should create the widest color gamut display ever created.

Does anyone see any problems with this before I lay down the cash to test this theory out?

One problem might be the existing dichro. If the curve is between 450 and 488 (rather than above 488) you may end up directing the cyan beam the wrong direction.

The other problem is cost If you've found the PLT5-488 for under $1200, please let me know. Seriously.

 

[tao]

Well damn. ..I didn't think about the dichro.
I don't mind buying another dichro and getting it cut to the same specs as the original blue one. (Just have to find some suppliers who carry a 490nm dichro lol)

And no...sadly I haven't found a cheaper source for the plt5 488. BUT its totally worth it to me to get this diode because it will serve two functions.
One is for my laser projector setup and the other is for my full color holograms. (It will give me a better white color)

 

[rhd]

Honestly, I don't really see how this gives you better color reproduction, given that source material doesn't contain a cyan channel. It's kind of like those yellow channel television sets someone tried to peddle a few years back. Everything I read said that there was no actual positive impact on picture quality. There shouldn't be, if you think about it. Interpolating a new color channel from two existing color channels shouldn't give you any better coverage than could have been rendered by using those color channels to begin with.

But this really isn't my area, so maybe I'm missing something.

 

[tao]

I can display content that will benefit from the extra coverage in the gamut though. ProPhoto RGB, Sony DSC-F828 with 4-Color RGBE RAW output, Adobe RGB & Rec. 2020 (UHDTV) color space to name a few.
(Most normal cameras only use R/G/B subpixels to make up an image, the F828's sensor uses four colors, Red, Green, Blue and Bluish-Cyan. By using cyan or near 490 nm (blue green) creates a better overall white image also.)

If I added a yellow laser for example it would be a waste as yellow is already covered VERY well by the showwx.

So I have many good reasons for picking 490nm.

 

[Cyparagon]

Using cyan instead of blue will decrease gamut by definition. Are you using in addition to blue?

 

[tao]

Yes I am using it in addition to 450nm blue and 442nm bluish purple.

Red. 642nm & 638nm
Green 530nm and 532nm
Blue 450nm and 442nm
Cyan 490nm

 

[crazyspaz]

Honestly, I don't really see how this gives you better color reproduction, given that source material doesn't contain a cyan channel. It's kind of like those yellow channel television sets someone tried to peddle a few years back. Everything I read said that there was no actual positive impact on picture quality. There shouldn't be, if you think about it. Interpolating a new color channel from two existing color channels shouldn't give you any better coverage than could have been rendered by using those color channels to begin with.

But this really isn't my area, so maybe I'm missing something.

Kinda what I was thinking...There was a booth (panasonic, maybe?) at CES that had "8K" TV's that were just 4K with a yellow subpixel. I was talking to a very helpful and knowledgeable Sony rep who said all the yellow pixel does is decrease performance since they take away from the true function of the other main pixels (color mixing), so its actually more difficult and less effective.
(sorry for tangent)

 

[rhd]

Yes I am using it in addition to 450nm blue and 442nm bluish purple.

Red. 642nm & 638nm
Green 530nm and 532nm
Blue 450nm and 442nm
Cyan 490nm

That seems entirely unnecessary to me:

- 450 and 442 is an irrelevant difference. If anything, why not 450 and 405?
- 532 and 530 is an irrelevant difference. If anything, why not 532 and 515