CurtPalme.com Home Theater Forum

deronmoped · Joined: 03 Nov 2006 Posts: 1154 Location: San Diego

Yeah it's a BOC album, but I want to explore something different here.

Why are curved first surface mirrors not being used? They seem like the perfect solution for lesser PJ's and 1080P.

Take one PJ that is not really capable of 1080P, add one curved first surface mirror, stretch the image and voila.

Deron.

deronmoped · Joined: 03 Nov 2006 Posts: 1154 Location: San Diego

Here is a link to a mirror. It looks pretty good, only $18.00 for a 2'X4' sheet, could be a fun experiment.

http://home.att.net/~cleardomesolar/solarflex.html

Deron.

garyfritz · Joined: 08 Apr 2006 Posts: 12088 Location: Fort Collins, CO

Chiem on the A site once explained the physics behind this. Bottom line: won't work.

See e.g. the thread starting at http://archive2.avsforum.com/avs-vb/showthread.php?p=6487559#post6487559. Unfortunately the images he posted are gone, but the text is still there explaining why it can't work.

deronmoped · Joined: 03 Nov 2006 Posts: 1154 Location: San Diego

Did anyone actually check to see if a curved mirror would work? From what I read it was just all formulas. My common sense tells me it would work.

Deron.

Gino · Posted: Mon Jan 28, 2008 1:43 am Post subject:

ecrabb · Posted: Mon Jan 28, 2008 2:04 am Post subject:

If you understood the basic optical theory, your common sense wouldn't be wrong! Wink

The problem with the curved mirror is that the projected focal plane becomes a curve. So, to have the entire screen in focus, your screen would have to be curved like the mirror. Basically, the curved mirror throws a huge optical wrench in to the works. IIRC, the curve wasn't even simple curve (like your common sense would tell you it would be.)

Also, Chiem's thread wasn't just formulae. He used some whiz-bang optics software to design and simulate the optical system, and rendered the simulated optical performance. Those were the images that are now gone. As I remember it, parts of the screen were in focus, and parts weren't. He came up with a design where you could use two mirrors to mostly correct the effect, but then the system became so large, it would be totally impractical in an HT.

But, feel fry to spend your money on a mirror to experiment with and report the results here. I'm curious, too!

Wink

SC

deronmoped · Joined: 03 Nov 2006 Posts: 1154 Location: San Diego

OK

Then tell me why did my PJ focus onto a curved screen.

My curved screen and a curved mirror are creating the same problem. The distance from the phosphor to the screen is changed by the curve. The center is farther away in each case.

Sure if you have a monster curve you may not be able to correct the focus, but there might be something to gain from a slight curve, not sure how much or if it's worth the effort.

Deron.

spitz · Joined: 26 Oct 2007 Posts: 26 Location: Indianapolis, IN

Doesn't Draper have something called Mirror Manager that warps the mirror to adjust it to the screen?
Is that what you are after?

http://www.draperinc.com/RearProjectionScreenSystems/MirrorManager.htm

-Matt

**AnalogRocks** · Posted: Tue Feb 05, 2008 5:07 am Post subject:

You know my reflector telescope has little adjusting screws on the back of the mirror to make the image perfect over the whole surface. Same idea. I like it. I'd hate to see the price though.

deronmoped · Joined: 03 Nov 2006 Posts: 1154 Location: San Diego

Some of the newer telescopes (monster size) adjust the mirrors to adjust out distortions in the atmosphere. They use a guide star, created by a LASER. How the hell do these guys figure all this stuff out Question

Deron.

MotoGeek · Joined: 07 Feb 2008 Posts: 2 Location: Redmond, WA

You can get the image close to the correct shape with simple mirror. It would have to be rescaled though to correct the nonuniform horizontal stretch. I'm sure the projector could handle the pincushion.

I think the problem is that the reflected lines don't converge to a focal point behind the mirror. I don't know much beyond that though.

ecrabb · Posted: Fri Feb 08, 2008 3:52 am Post subject:

Nice drawings, but now make that work for 3-lens projection with three different projection angles (offset and toe-in)! How large would the mirror have to be? Finally, as you indicated, the focal plane is not a plane anymore, but a warped plane (rectangular section of a cylinder). You'd be able to properly focus the edges but not the center, or the center but not the edges. If you think about it, your drawing actually illustrates the complexity of the problem and why it won't work for a CRT projector.

SC

PS - Welcome to the forum! You a CRT'er?

deronmoped · Joined: 03 Nov 2006 Posts: 1154 Location: San Diego

I still say you can do it, CRT's are capable of focusing on a curved surface. Horizontal curved screen, vertical curved screen or even a torus screen is no problem.

Deron.

Dave Lister · Posted: Fri Feb 08, 2008 9:24 am Post subject:

It might work if you can get the right lenses from a rear projection CRT TV, wouldn't they be made to compensate for the distortion?

**AnalogRocks** · Posted: Fri Feb 08, 2008 5:19 pm Post subject:

Or along the same lines you could start playing with disasembled lenses to see how adding more elements warps the picture.

MotoGeek · Joined: 07 Feb 2008 Posts: 2 Location: Redmond, WA