Film-Tech Forum ARCHIVE: Rear projecting 3D

This is topic Rear projecting 3D in forum Film Handlers' Forum at Film-Tech Forum ARCHIVE.

To visit this topic, use this URL:
https://ft-forum.com/ft/cgi-bin/ubb/ultimatebb.cgi?ubb=get_topic;f=1;t=001884

Posted by Tom Sauter (Member # 616) on 01-31-2001, 07:43 AM:

I'm assisting a colleague of mine with a virtual reality setup. The plan is to rear project video (dual projector) with polarized filters.

This got me thinking: what factors would one consider when setting up a 35mm rear projection 3D system?

Posted by Darryl Spicer (Member # 711) on 01-31-2001, 08:25 AM:

Here is the problem that I see with this type of setup. In order for 3D to work the screen has to be a high gain silver screen. When the imagage is projected from the front the light is reflected back at you with more intensity there for allowing the 3D to work. I do not see how you can get this reflective action by projecting from the rear.

Posted by John Pytlak (Member # 331) on 01-31-2001, 08:44 AM:

3D projection of a left eye and right eye image would needs to isolate the appropriate image to each eye.

Polaroid glasses are often used with film systems, but the polarization of the light needs to be maintained by the screen, which is usually an aluminized high gain front projection screen. AFAIK, most materials used for rear screen projection would diffuse the light and render polarization useless.

You could use LCD shuttered glasses to separate the left and right eye images sequentially. The video projectors would need to alternatively present the left and right eye images, and the LCD glasses would need to be syncronized with the display, presenting the sequential images to the correct eye. A similar system is used for the high-end IMAX PSE (Personal Sound Environment) headsets.

The other method would be to separate the images by color (anaglyph), but this is not an optimum way to present color images. The glasses typically have a red filter for one eye, and a cyan/blue filter for the other.

------------------
John P. Pytlak, Senior Technical Specialist
Worldwide Technical Services, Entertainment Imaging
Eastman Kodak Company
Research Labs, Building 69, Room 7419
Rochester, New York, 14650-1922 USA
Tel: 716-477-5325 Cell: 716-781-4036 Fax: 716-722-7243
E-Mail: john.pytlak@kodak.com
Web site: http://www.kodak.com/go/motion

Posted by Gordon McLeod (Member # 33) on 01-31-2001, 08:51 AM:

Rear screen will diffuse the polarization the alternate sync method will work but Imax I believe has several patents involved and a very big legal department as well

Posted by Tom Sauter (Member # 616) on 01-31-2001, 09:04 AM:

Thanks, John. I suspected the diffusion of the screen would be problematic for polarization. I hope the screen material used was designed to overcome this, since they purchased a screen _before_ asking for my help. When the project gets a little further along, I'll post back with more details.

Posted by John Pytlak (Member # 331) on 01-31-2001, 10:15 AM:

3D LCD shutter viewing systems are available from others besides IMAX Corporation:
http://www.ahhz.com/docs/ecs075.htm
http://www.crpc.rice.edu/CRPC/idesk/tech.html
http://www.stereographics.com/
http://www.stereographics.com/html/whtpaprs.html
http://www.imax.com/innovations/theatre/IMAX_3D/index.html
http://www.imax.com/innovations/theatre/IMAX_PSE/index.html

There are also polarizing systems for computer displays:
http://lxdinc.com/Stereo_Graphics.htm

Posted by John Pytlak (Member # 331) on 01-31-2001, 10:27 AM:

I forgot to mention, Kodak's Dynamic Imaging Group (part of Entertainment Imaging) also offers a unique 3D imaging system that has already found its way into theatres for advertising displays and promotions:
http://www.kodak.com/US/en/motion/dynamic/

They can put full motion or 3D images (no glasses needed) onto posters, signs, and cards. They've been very successful with their sports memorabilia, with action scenes from historical sporting events.

Posted by Gordon McLeod (Member # 33) on 01-31-2001, 10:48 AM:

At one time I remeber (mid 70's)seeing a lenticularized rear screen and I wonder if you could use the russian lenticular system to project from different angles into the screen to produce the stereoscopic pairs

Posted by Jerry Chase (Member # 660) on 01-31-2001, 12:19 PM:

Rear screen 3-D projection is not only possible, it is AWESOME! How do I know? I've done it off and on for 25 years.

People in the industry natter on about the 3-D "window" but few have ever really seen a true 3-D window in all its glory. Front projecting makes the ideal viewing angle impossible, since the projector has to be close to centered. Problems with light drop off at the edges, specular reflection at the center, etc. make the silver screen less than ideal.

Rear projecting polarized 3-D requires a special plastic material that is fairly expensive, and strange stuff to work with. It does not de-polarize the light as plain frosted plastic or groundglass does, and for some happy reason it doesn't give as much of the usual "hot-spot" where there is a direct alignment through the lens to the projector bulb. The plastic has what appears to be a greenish tint, and is gummy rather than being slick like polyethelene or other plastics. It also acts as a neutral density filter, soaking up some of the light that hits it.

I've done the variations of the virtual reality thing using this plastic and slides, and the plastic and dual super 8 3-D. The slides worked much better, both because of synch problems (arrrgh!!!) and very low light levels from the super eight.

The slides were nothing short of spectacular. I used a section of plastic about a foot on a side, used black matte board as a frame, and projected both prism 3-D and dual projector slides. Because of the small screen, the light levels were very high, and because I am nearsighted, I was able to remove my regular glasses and use the polarizing glasses to get an effect that was as close to looking out a window (with corrected vision) as I've ever seen, including Kodak's 24 image lenticular system.

The effect only works because of the high light levels - giving more realistic contrast levels, the high resolution of color slide film - giving the detail and visual interest that is required for extended 3-D viewing, and the blackout of all extranious distractions.

My best setup was in my basement, where I could baffle the spill light from the projectors and keep any daylight or other light from entering. This made a huge difference compared to when I tried to set up in a dimly lighted room. John's "contrast killers" totally spoil the effect.

The reality was so well recreated that when I had a photo of something that went through the screen plane, it was possible to reach out and expect to touch a solid object. One shot of a horse's nose was particularly effective, with the hair clearly visible. The screen effectively disappeared on many shots, only becoming noticable when there were large areas of solid color and dust.

Trying to do an effective virtual reality with current generation video is, IMO, doomed from the start because the resolution isn't there to carry it off.

I'll tell you what I have done, which is as close as you'll get. There is a $500 3-D alternate scan LCD adapter that can go on the front of most camcorders.
http://www.3-dvideo.com/nuview.html

If you have a 53" or larger rear screen projection TV with a good comb filter (My Hitachi is very good), you can set the zoom on your camcorder to match the screen width. In other words, with the 3-D attachment on the camcorder, sit in your normal viewing position in front of the tv, attach the output from the camcorder to the input of the tv, then zoom in and out until the image of the frame of the picture area exactly matches the actual frame of the set. Rough set the convergence so that the frame of virtual tv set appears superimposed in depth on the actual tv frame.

Precisely adjusting the convergance of the 3-D adapter is next. Use a label to set up markings around the convergance dial. Go outside, set the camera on a tripod, and shoot some shots of an object over 1000 feet away, while adjusting the convergence and announcing the convergance setting.

Play this back on the set, and measure the distance between the two images you see without the LCD shutter glasses. When you get to a setting where the distance is exactly the distance between your eyes, that is _your_ personal perfect setting. It will vary from person to person, set to set, and camera to camera.

Set the convergance to that setting, lock the zoom to the proper setting, and go out shooting scenes using the exact same shooting angle as your viewing angle watching the set. (You'll quickly realize most couches and chairs put your head a little too high when viewing tv.)

When you play back the video tape you shot, you will have achieved as close to 3-D virtual reality with video as you can get with current equipment. It is a far cry from 3-D rear projected slides, but it is quite good, and won't cause much eyestrain.

Eventually, Ultra high definition video and special screens may approach the quality of 3-D rear projected slides, but I don't expect anything like that to hit the market for years.

In case you haven't figured out by now, I am _the_ 3-D nut.

Posted by Gordon McLeod (Member # 33) on 01-31-2001, 03:47 PM:

What is the name of this material and source?

Posted by Jerry Chase (Member # 660) on 01-31-2001, 05:21 PM:

>What is the name of this material and source?

Frustration! I bought mine from Edmund Scientific a number of years ago, yet I don't see any rear screen projection screen listed in their online catalog.

You might try:

http://home.t-online.de/home/Screen-Tech/ST-Professional-W.htm
http://cescompany.com/html/draper_diamond.html

For a site about a 3-D virtual reality video experiment: http://www.ukans.edu/~mreaney/reaney/vrscenog.html

Polaroid Land corporation would know of some sources. There used to be polarized 3-D rear projection viewers for photomicography and medical uses.

The key is that the system uses lenses rather than random difraction. I suspect that the material I have contains micro-droplets of a different type of plastic. Edmund may still have a manufacturer on file.

I'm a little surprised I can't find a source with a net address.

Posted by Tom Sauter (Member # 616) on 02-01-2001, 07:31 AM:

I did more research and it appears Stewart Screen now manufactures an RP surface (I think its blackhawk or something-black), which is designed specifically for use with circular polarized rear projection. It just came onto the market in the last year or so, and is certainly not cheap.

Posted by Jerry Chase (Member # 660) on 02-01-2001, 08:58 AM:

I'm beginning to think I never throw away anything. My 1988 Edmunds catalog shows "Projection Material for Daylight AV presentations" - items # X70,934 1x1' sheet of rear screen plastic $11.95 #X70,935 3x3' for 54.95 and X2508 6' wide sheet at $39.95 per linear foot. It doesn't mention the 3-D properties, but I'm sure it was the same stuff I had purchased earlier.

I'm wishing I still had the Edmunds catalogs from the 60's and 70's to sell on eBay!

Posted by Gordon McLeod (Member # 33) on 02-02-2001, 01:42 PM:

I remember useing one of there screens once I found I could keep polorization either only in the centre area or on a very narrow viewing angle. THis was a long time ago

Posted by Jerry Chase (Member # 660) on 02-02-2001, 02:45 PM:

Then my guess ist that it was probably a different material. The screen I have is quite forgiving. I can see how a fresnel based screen would have that type of problem. This screen has no apparent lens structure.

Posted by Gordon McLeod (Member # 33) on 08-01-2001, 05:40 PM:

Harkness Hall has listed a polarized rear screen material in there latest flyer

Posted by Tom Sauter (Member # 616) on 08-28-2001, 01:51 PM:

Our interactive rear-projection 3D system is now up and running. The system is utilizing a stewart polarized rear projection screen, two 3000 lumen lcd video projectors (stacked), and circular polarized lens filters and glasses (passive 3D). Aside from a bit of tweaking that needs to be done on the projector alignment, I almost walked right into the screen while checking it out toady!

Anyone who is in the Western New York (Buffalo) region can contact me and we'll set up a test drive.

Posted by Jerry Chase (Member # 660) on 08-28-2001, 03:06 PM:

Tom, how about posting a couple of pictures of the setup and giving a description? Some comments from people that have seen it might be fun too. I'm particularly interested in how the screen material worked. Are there any hot spots, ghost images, etc.?

Posted by Bob Maar (Member # 763) on 08-28-2001, 03:16 PM:

Tom your post reminds me of the time I first saw Doug Trumbull's Showscan. The film took place in a theatre and the projectionist went behind the screen and turned the light on.

I went crazy trying to figure this out, I would have bet all my money that he was really on stage.

Posted by Tom Sauter (Member # 616) on 01-29-2002, 09:19 AM:

There are now some pictures and technical details of our rear projection 3D system:
http://www.ccr.buffalo.edu/anstey/VR/LOW_COST_VR/index.html