Re: Shift lenses

[P3D John Bercovitz <bercov@xxxxxxxxxx>]

Wm. C. writes:
 
> Keep the film plane parallel to the screen...
 
John B responds:
 
>Can anyone explain how this cures keystone?  It's not intuitively 
>obvious to me.
 
Wm. C. responds:
 
> Only the vertical axis of the beam is affected by the wedge 
> prism. The horizontal axis is not (the vert. & horz. elements of 
> the image). Therefore as the beam is steered through the prism, 
> the further, less retarded part (in your case top) of the beam 
> (think of it as swinging through a hinge?) will be wider than 
> the bottom... keystoned.
>
> When this keystoned beam images on the screen, it will keystoned 
> oppositely. That is, the top now is narrower than the bottom. 
> The result is a cancellation of these two keystoning effects, 
> rendering a normal image. 
 
I sure don't follow this.  Can you explain it a second way?
 
> In fact I think you'd have a tough time creating a keystoned 
> image with any manner of mirrors and/or prisms as long as you 
> kept the film parallel to the screen.
 
I don't think I buy that.  I can think of lots of ways of using 
mirrors that would be equivalent to tilting the film plane 
relative to the screen.  Think of it this way: You can see the 
projector in a new position when you look at it in the mirror.  If 
that new position has a different orientation of the projector, 
the film will no longer be parallel to the screen.
 
> I just looked at my Melles Griot price list for '92-'93. Sheese! 
> You might think about making a prism out of plexiglass and 
> water!? 
 
Yeah, aren't they a trip?  I asked them about it maybe ten years 
ago but didn't get what I felt was a satisfactory answer.  Good 
quality stuff, though.  Of that there should be no doubt.
 
OK, the thought of calculating this problem was not something I 
looked forward to (can you imagine me not wanting to calculate 
something?).  So I used an analog computer: I found (That was the 
hard part; I _have_ to get organized.) a big light crown wedge 
prism in my junk box and did the experiment.  First I focussed the 
projector on the screen.  Next I stuck the prism in the beam and 
got an image at a different elevation on the screen.  I measured 
the keystoning.  Then I took out the wedge and changed the angle 
of the projector until its image was located where the image was 
when the beam went through the prism.  I again measured the 
keystone.  Got exactly the same number.  Now what did I do wrong?
I tried it with the apex line horizontal and up and with the apex 
line horizontal and down.  I'm stumped.
 
John B
 
PS: You know, since it doesn't make the situation any worse (at 
least if you use a good achromatic prism), and you may want to 
keep your projector level for some reason (like the feed of the 
reels) maybe the prism is a solution for a different problem other 
than keystoning.


------------------------------