AVS Forum

AVS Forum (http://www.avsforum.com/forum/)
-   Screens (http://www.avsforum.com/forum/23-screens/)
-   -   RAY TRACING 101 & SPHERE INTERSECTION. (http://www.avsforum.com/forum/23-screens/19160-ray-tracing-101-sphere-intersection.html)

CINERAMAX 10-07-2000 02:25 PM

http://www.cc.gatech.edu/gvu/multime...y/vsray00.html

Now go do your homework....

[This message has been edited by CINERAMAX (edited 10-07-2000).]

KFung 10-07-2000 07:42 PM

Peter,

If you are going with the pre-formed plastic screen and then having it laser perforated, you don't need to worry about the pixel spacing. To find the spacing of the holes on a curved screen you would want to project the holes off a flat screen onto the curved screen and then flatten the curved screen and measure the distance between the holes. This is only necessary if you are perforating the screen prior to forming it. If you are getting the screen pre-formed and then using a laser to perforate it, you just have the laser "project" the pattern of the holes for a flat screen onto the curved screen. Do you see what I mean?

Regards,

Kam Fung

CINERAMAX 10-07-2000 08:22 PM

I have to think about this one Remember there is pincushion.

There has to be a hole on each corner of the pixel. The laser or water perforation machine cannot follow an equidistant spacing. towards the center of the screen yes. But at the edges there is pincushion distortion.

Actually the RAY TRACING being done is not Meeting a sphere but a TORUS.

Placing the screen face down makes much more sense.

"Where did you find a lens to focus a DLP on a parabolic screen?"

I saw an alice shooting a vutec with good overall focussing. Depth of focus may not be an issue. specially with an anamorphic lens.



KBK 10-07-2000 10:01 PM

Yes, I agree, lens-to-lens variations alone would give you the fits. There is no reason not to fall back to the idea of a standard perforation pattern. Acoustically speaking, only a thin material will do, anyway.

I thought about it for a second, and realized that all the ray tracing you need, including that for projector type, seating, angle, etc., can easily be handled by the proper usage of the sphere/sectioning capacities of a modern 3-d graphics package. Describe the curve you require in 2-d, then use the program's capacities to extrapolate to 3-d, do horizontal and vertical separately I suspect. mirror the two features, I expect, then meld together. Done. The painful part is teaching yourself to understand how to get the dang program to do exactly what you want.
------------------
---Place Signature Here---

[This message has been edited by KBK (edited 10-08-2000).]

[This message has been edited by KBK (edited 10-08-2000).]

CINERAMAX 10-08-2000 08:00 AM

Thanks Dave,

You made our work a little easier. We will work with an existing perf aluminum sheet. Then form it by rolling it atop a round plywood mold Clamping it down. Fiberglass reinforcing the back. painstakingly routing the fiberglass off the holes from behind and inserting the curve in a wood frame/shadowbox, which has the perifery of the parabola's shape routed in the frame . Thin strips of wood are then laid at the edges and the stappled to the inside of the bezel.


There is a new thread that will focus on this developement.

------------------
The Artist formerly known as PANARAMAX.....

Stew M 10-08-2000 09:54 AM

Cineramax,

I'm confused by your terminology - you use parabola, sphere and torus interchangeably, but these are three very different shapes.

A parabola (2nd order curve, here acting in reverse) will return parallel rays to the rear wall, after reflecting light from a point source (lens).

A sphere (or frustum thereof) will send rays right back to the point source.

A torus might create a linear image or "sweet spot" (depending on the 2-d shape you rotate around what axis)

Which do you mean?

------------------
Stew
The search for excellence is unending . . .

CINERAMAX 10-08-2000 04:29 PM

If you mean by a parabola like a satellite dish shape. Then the right word is a TOROIDAL screen. Thanks.

------------------
The Artist formerly known as PANARAMAX.....

Jeffrey 10-08-2000 07:17 PM

Hi Peter.

Have you scrapped the idea of perforations at the corners of the pixels? I think the tolerances would be very difficult to meet- the screen and the installation.

After two semesters of Vector Calculus, and plenty of homework done in vector notation and polar coordinates, I could probably calculate what you are looking for, that is, if I knew what you are looking for.

Thanks for sending the Lutron devices. They are going to work out great in my new HT (under construction). I've been enjoying the posts regarding the current venture. Kung Fu and KGB sound like the kind of guys with which to work.

Jeff

CINERAMAX 10-08-2000 08:32 PM

Jeff :

You can do ray tracing calculations? We need badly someone like you on board as the jet/laser perfing guys want FORMULAE.

I know it is not much, a late comer's slice, but put yourself down for 9% IP ownership of the Perforated TOROIDAL screen project. Welcome Aboard!

You do understand that if someone copies our concept, all four of us, may have to go into action, can you handle a fifty caliber? I was shooting skeet all afternoon.... Let them just try it. : )

I will e mail you details on the screen,



------------------
The Artist formerly known as PANARAMAX.....


All times are GMT -7. The time now is 12:46 PM.

Powered by vBulletin® Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
User Alert System provided by Advanced User Tagging (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.