3D projector ideas

[Chernobyl]

original thread: Home made projector!
post #59

Originally Posted by Áedán

Quote:
Originally Posted by Chernobyl
"I read that you can make a stereoscopic projector from 2 OHP setups with horizontal & vertically polarised filters (using polarised glasses too). I'm not sure how I would get a page flipped image out to each projector though."


Indeed you can, but you definately need a screen that will retain the polarisation for it to work well. It would appear that many surfaces will affect the polarisation, leading to poor image seperation.

One possibility would be to use the stereo switching signal to alternately blank the LCD panels, but again, you might end up with bleed through due to the slowish response of the LCDs. Otherwise, you're back to using some kind of demultiplexor/frame buffer to keep the images current. I wonder what would happen if you held off the sync pulses from one of the LCD panels for a single frame?

Can you tell this is something I've thought about a bit?

lol, yes!
thanks for the info on surfaces not working too well with polarised light, that will need investigation.
Sadly the second OHP has to go back, so I no longer have a means to try anything, it cant hurt to discuss though

Your technique of suppressing odd/even frames with the sync pulses could work? Extra framebuffers shouldnt be necessary as the LCD panel is itself memory. This is assuming that with no sync signal fed to the LCD, it will hold its current image. Perhaps this could be forced if it is an issue?
Blanking of the LCD panels will dim the light, so is best avoided.

This might change the project:
I recently heard from a knowledgable chap that NVidia are going to allow stereo signals to be fed to the 2 monitor outs on the graphics card. This would simplify matters somewhat!
Lets hope its true

I'm wondering if stereo vision with 2 OHP/LCD screens would work straight out of the box? There would need to be some compensation for the different angles the images will be projected at, as the pixels from both need to line up pretty well.

Do you consider having both polarised images on at the same time to be a problem? I feel it shouldnt be an issue but am interested in your thoughts.

Chernobyl

[Aedan]

If polarised light is in use, then there should be no issues with both projectors running images at the same time. My thoughts were to minimise the amount of electronics! The LCD drive electronics should already have a framebuffer, but I don't know how they'd handle no sync every other frame. This of course assumes that the signal is analog. I don't know enough about DVI to figure out how we could switch DVI.

The concept of independant left and right images coming from seperate ports would make things nice and easy. That would be ideal!

[Gizmo]

It's been a long time since I messed with this, but I was under the impression that reflection off a surface tended to polarize light anyway. If both projectors are reflecting off their mirror assemblies to project the image, the light is going to be polarized into the same plane, isn't it?

[Chernobyl]

I am not sure where the best place to put the polarisation filters would be.
My hope was it would be the last thing the light passes through before hitting the screen. I dont know if this is normal or a good idea.

Can a mirror really change the polarisation of light?

[Gizmo]

It doesn't change the polarization per se. What happens is that only the light that strikes the plane of the mirror at the proper angle gets reflected, hence only light that has a particular polarization will be seen. I'm not sure if the effect only occurs with glass surfaces though.

We did an experiment in high school physics where we used a polarized filter to look at light reflected from a sheet of glass. If you looked at the light source and rotated the filter (to change the angle of polarization), the light source remained unchanged. If you looked at the reflected image of the light source, the image would alternately fade out completely and then fade back in as you rotated the filter, showing that the reflected light was polarized. A lot of LCD calculators take advantage of the effect to improve the contrast ratio of their displays. The (better) glare filters that you put over CRT screens work on the same principle.

[Chernobyl]

Hi Gizmo
thanks for your experiences.
From what you are saying the polarised filters will need to be placed between the OHP mirror and the big screen.

By examination of what you have stated it may be possible to use the OHP mirrors as polarisation filters. One OHP would need to be mounted at 90 degrees from the other.
As I dont know the polarisation angle that will come off the mirror, maybe both OHPs will need to project at say 45 degrees either way of the centre? Hopefully 0 and 90 degrees is correct.

Do you agree this is a possibility?

[Gizmo]

Seems reasonable to me, but then again I'm no expert.

Check this link. It seems to contain a simple enough to understand explanation of what's going on.

[Chernobyl]

thanks Gizmo, good link.
You're right, its not surefire getting it to work using just mirrors for polarisation, so filters between the last lense and the screen will be needed I surmise.
I wonder what projection surfaces wont depolarise the light?

How I wish this second OHP wasnt going back.

[Gizmo]

Highly polished metal surfaces, for starters. That's one of the reasons that the silvered side of a reflecting telescope mirror is used directly even though it is very delicate; it avoids the polarizing and refractive effects that would occur if the light were to strike a glass surface before being reflected off the silvered surface of the mirror.

[Chernobyl]

While researching the OHP projector, I came across some useful info.

LCDs have a polarising filter in and another out as part of the function of the LCD crystal.
(One bad side effect is immediately losing 50% of the light just by only letting singularly polarised light through the panel)
In theory, the light leaving the LCD should be polarised!

To test the earlier polarising mirror theory and the polarised LCD theory, I rotated the LCD screen 90 degrees to see if there was any loss of light. I couldnt see any difference sadly.
I'm thinking that the mirror I have may not be a good polariser after all. I thought there may be a layer of glass on it, perhaps not.

Is there anything around the home that could be used to test the polarisation of light from the LCD and the mirror?

cheers

[Chernobyl]

Must've been drunk when I wrote that
See children its bad!

Thanks Gizmo, I'll polish some metal and see what happens.