DLP projector for stereograms

Jeffrey Weil · Post by **Jeffrey Weil** » Sun Jan 30, 2011 4:56 pm

Brand new the projectors are something like 800-900 bucks, but the dlp development boards I've seen are thousands of dollars. I have no idea why. They don't even have light sources. I hope new ones are out that are cheaper than what I saw a year or two ago.

On the pico projector, check the res of that unit. Unless your stereograms are going to be very small it might not be sharp enough for you. Large pixels are much more noticeable on a stereogram than a projected moving image like the projector was designed for. Look at computer displayed images vs printed ones. Take an 800x600 image and view it full size on your screen, it looks great. Now print that picture the same size as your screen, it looks like crap!

How large are you going to make the holograms? That'll determine how much res you need. I would keep it to over 200-250 dpi if you want it to look ok, over 300 for it to look good. For example an 800x600 dlp chip shouldn't be projected more than 4x3 inches. Smaller would be better. Those are just starting numbers, you'll have to experiment to find where the limits really are. Plus its subjective. You might like large images with visible pixels.

The lcos will reduce the diffraction issue. I don't know how much of problem it's going to be with a dlp, maybe not bad at all. Don't know. But the lcos will have less.

Jeffrey Weil
NorthBeach Holography Inc.

Justin W · Post by **Justin W** » Sun Jan 30, 2011 11:43 pm

Well, I'm aiming to create masters that are 8x10.

It's really beginning to look as though using a projector for multiplex work is going to be absurdly expensive and possibly unsatisfactory even after all the investment of time, money and effort that it sounds like will be involved.

I think I'm going to begin considering just using an LCD panel backlit with my laser light. Unless, of course, the refresh rate of these also presents a problem....
but we'll save that for another topic.

Thank you to everyone who chimed in on this subject and helped me steer clear of a costly and disappointing mistake.
I'll be back again with another thread to start asking advice on the matter of LCD panels when my eyelids aren't drooping so much. It's late here.

Dinesh · Post by **Dinesh** » Mon Jan 31, 2011 1:35 pm

favalora wrote:I don't recall where the right place would be to put an aperture to block the diffracted light.

Without the aperture itself diffracting. Put the aperture too near and you may get Fresnel ringing. Too far away and you'll get Fraunhofer. I would suggest some kind of variable ndf at the edges.

jnhong · Post by **jnhong** » Mon Jan 31, 2011 10:39 pm

Justin, sorry to jump in so late into your thread. The picoprojector is likely to work. Do a search on MIT Media Lab papers. They have one that describes their hologram-making experience using the TI DMD projector, way back around 2000-2002. They didn't find any issues from the grayscale modulation of the micro-mirrors and there was no diffraction issue. In your case, when you illuminate the DMD with the laser and it throws a nice, crisp image onto your screen, you'll quickly see if diffraction is distorting your image or not.

If you shape your input beam properly, you shouldn't have a problem. You will need to expand the input beam slightly to fill the first lens into the optical engine, so that the DMD will be properly filled. The optical engine has a microlens screen to diffuse and homogenize the illumination. It will probably scramble the phase, and that should take care of diffraction artifacts.

It's pretty hard using an LCD to generate laser-backlit images. The transmission efficiency is 2% at best, and then you need to find a way to match polarization of the LCD and the reference beam, against the film plate. And then if you need to buy a new LCD or HDTV to sacrifice to the optical altar that could be more expensive than a picoprojector anyway.

Joe

holorefugee · Post by **holorefugee** » Mon Jan 31, 2011 10:46 pm

I have been waiting for Ed to jump in, he has used the DLPs as well.

How large are you going to make the holograms? That'll determine how much res you need. I would keep it to over 200-250 dpi if you want it to look ok, over 300 for it to look good. For example an 800x600 dlp chip shouldn't be projected more than 4x3 inches. Smaller would be better. Those are just starting numbers, you'll have to experiment to find where the limits really are. Plus its subjective. You might like large images with visible pixels.

Thank you Jeffery. I had not realized that the resolution requirements were so high. What does Liti Holographics use? The resolution is pretty good on the Kiss hologram that I have. There are some artifacts on the first and last frames from image processing, some frame to frame noise as well as time smear but if I could make that image I would consider it a success.

http://www.litiholographics.com/samples.htm

Ed Wesly · Post by **Ed Wesly** » Tue Feb 01, 2011 8:54 am

I am just back in town and missed the beginning of this thread, but we used at CFC/Northern Banknote a TI DMD with at first 600 by 800 pixels, and it got pixelly when blown up to about 6" by 8". For the typical security size holo it was just peachy. We wanted to go to XGA resolution, 1024 by 768. For some reason TI used a third party to sell the developer kits, like $10k, and that didn't include a board for continuous tone. Another $6k. Go figure!

Good luck!

PinkysBrain · Post by **PinkysBrain** » Tue Feb 01, 2011 9:01 am

I wonder how hard the hogel approach would be to DIY. With a small modulator like a DLP or LCOS the optics get a bit complex, but with a LCD it should be easier.

Could you just push a collimated beam through a LCD (with the pixel intensity being the rear plane color of the last plane intersected from a ray from that hogel to the "sky" in that direction) and project the result onto the plate with a fresnel lens (slightly off focus), interfering it with an off axis beam to create the hologram? You could use a modified old pen-plotter to step the setup around the plate.

PS. of course getting it stable enough to do it without a decent power pulse laser might be problematic.

holomaker · Post by **holomaker** » Tue Feb 01, 2011 10:06 am

Heres a posting i made over at ning a while back, the MIT link is still working !

I am currently trying to use the afore mentioned projector to try and digitize my holographic images, so i thought i would start a thread to help myself and others that are to make this semi jump into digitized holography. I purchased this unit a few years ago thru ebay and since it had some issues i decided to disassemble the unit to help to understand its workings. I now have purchase two more working units an plan on adapting one for my holo-table. Very little information is available about the workings of these units, so any information is worth its weight in gold. To my surprise while doing searches for this unit i found this link from MIT lab

http://www.media.mit.edu/spi/SPIPapers/ryder/SPIE99.pdf

holomaker · Post by **holomaker** » Tue Feb 01, 2011 10:54 am

where i left off was; haveing the quartz bulb lit in box, the little color wheel spinning on some foam rubber bed and the board with the DLP running, as a stand alone (needs optical train now),if i could only fool the projecter into thinking the bulb was runnning, i may try and see if a hene tube will fool it!

here is a picture of the Taiwan school's table setup

http://api.ning.com/files/DSGM345U4yP3m ... G_0490.jpg

Justin W · Post by **Justin W** » Tue Feb 01, 2011 4:21 pm

Well, Hurray!

This is an unexpected treat. Just as I had given up hope on what I thought was going to be an excellent method of managing source images, along comes fresh insight!
Beautiful.
A Pico projector truly does sound like a much more fun, compact and generally cooler option than prying apart (and possibly ruining) a big, expensive LCD panel only to have the barest fraction of my object light make it through. Then there's the polarization issues: Ugh.

So I am truly happy to encounter the MIT paper on the subject. I think I just might give this a try.