RE: preliminary summary on ghosting

[P3D Rafael Ramirez <rafaelramirez@xxxxxxxxxx>]

> Date: Tue, 29 Jul 1997 11:06:48 -0700
> From: P3D Peter Abrahams  <telscope@xxxxxxxxxx>
> Subject: preliminary summary on ghosting
> 
> There has been a good amount of input on ghosting, and this is a first
> attempt at summarizing it for the archives.  Comments and additions are
> solicited.
> --------
> If the polarizing filter is placed between the film and the projection
> lens, any depolarizing quality of the film will have no effect.  It can
> also be placed on the outside of the projector, where it is subject to
> scratches (which depolarize as well as scatter light), if it is not
encased
> in glass, like a camera filter.   Many projectors focus by rotating the

Obviously, for lowest ghosting, this is how all polarizers should be
mounted, as the very last element of the projection path, although they are
usually placed before the slide to reduce the amount of heat on the precious
slide itself.

> Screens made of tiny glass beads or of white paint will scatter and
> depolarize the light.  Metallic, silver screens are required.  Ghosting
can
> be stronger at the corners of a screen, or to a viewer who is not directly
> in front of the screen.

For this reason, flat metallic screens should be preferred (light permiting)
over lenticular ones.

> When using Polaroid filters, some amount of ghosting is inevitable.  When
> two sheets of the best Polaroid material are sandwiched so that the
> orientation of the polarization of one is at 90 degrees to the other, a
> maximum of 99 percent of the light will be blocked, and one percent will
be
> transmitted.  Commonly used Polaroid material transmits more than one
> percent when crossed in this way, and this quality can degrade with time,

I have reached the conclusion that the best polaroid filters can
theoretically block OVER 99% of the light on PERFECTLY aligned 90 degrees
polarizers, maybe about 99.9% on most light waves, but even 0.1 percent on a
high contrast subject can stand out as much as a baby light in the dark. 
Needless to say you NEVER get PERFECT alignment between projection filters
and audience glasses, so the amount of light leaking grows EXPONENTIALLY and
even a 1 degree head-tilt increases ghosting by an order of magnitude. 
Also, extintion is not the same for all wavelengths.  For some, mostly in
neigbourhood of color red, it is usually big in the most common types of
polarizers, say 99.99%, while it may only be say 99.8% on the blue area. 
Moral:  if you must do high-contrast subjects, make sure they are redish :-)

Now that I think about it, it is not that much of a bad advice.  A slightly
redish (orange?) filter on the camera could possibly reduce ghosting a few
percent points.

Also you are very right on the H type of polarizing films (the
polivinyl/iodine streched type) degrading over time.  They DO degrade, they
do it MORE than believed or subjectively perceived, and even REALLY good
ones do when subjected to light sources of, say, 250W or more over a 30mm
square area.  This is actually a big problem on LCD panels, which rely on
polarizer filters to do their magic.  This usually results in a lower
contrast range for the panels over time, so it is not seldom perceived.
Large research on this subject has come out with some Japanese polarizers of
great transmission/extintion ratios (>40% >0.001) that degrade twice as slow
as the best polaroids.

And don,t even think filters don,t degrade if not used, because the aligned
molecules in their insides might decide to go out on a date if they are
bored.

> causing worse ghosting.  If the Polaroid is not held flat, but is curved
> during manufacture, or bulges in the center from the method of mounting,
> extinction is less complete.  Because the plastic is heated during
> projection, flatness is best maintained if the filter is sandwiched
between
> glass.

Good filters for projection are the kind sandwiched between thick plastic
sheets or glass DURING MANUFACTURE.

> A small amount of depolarization is caused by the curved surface of any
> lens.  This is probably inconsequential for projection, but is a problem
> discussed in texts on polarizing microscopes, where the lenses are spheres
> of very small radius.  Condenser lenses can have quite steep curves.
> Coating the lens also minimizes the effect.

Best thing in this case seems to be advice #1:  Place flat polarizers as the
LAST element of the projection chain or the optical path.  Not much can be
done if polarized light must passed through a complex and very "bending"
optical path.

> Circular polarizers have been used for stereo projection, but probably do
> not give more complete extinction.  The typical circular polarizer is a
> linear polarizer adjacent to a quarter wave plate.  I believe I've heard
> that the problem with circular polarizers is that you can tilt your head
> with no noticeable effect, but over time the rotation of the eyes that you
> have to maintain with a tilted head will cause eyestrain.

Circular polarizers are almost as good as linear ones, except generally a
little more expensive (say 10%),  little less transmission (say 7%), little
less extintion (say 1%), all this due to the added 1/4 wave retarder.  The
BIG plus is that extintion is mostly unaffected by head rotation (cross axis
alignment), so they would be preferred for large audiences looking at large
or curved (panoramic) screens from wide range of angles, moving audiences
(walking around), the complex optical paths mentioned above, etc.

> Nicol prisms of calcite were used for stereo projection at the turn of the
> century.  For a long time, I searched for Nicols for my microscopes,
hoping
> that they would give greater extinction (thus a blacker background).
> Nicols deteriorate with time, and it took me much effort before I finally
> got a good pair.  They did not give greater extinction than quality
> Polaroid material.  

It seems the best anybody has been able to do is about 0.0005% leakage on
controlled conditions on a specific wavelenght, so we can expect no more
than say 0.01% on the average.

> Finally, there are grades of polarizing material, and it is possible that
> some laboratory grade Polaroid gives more complete extinction.

Their transmission/extintion properties do vary.  I would like to know which
ones have the greatest extintion/transmission ratio, with emphasis on
extintion.  I have the figures on some Polaroids and the best, for
microscopes, seems to be HN22 (22% transmission, 0.0009% leakage), for 3-D
glasses seems to be HN38S (about 38%-0.005%) or HN35 and for 3-D projection
I would probably go for the HN42HE (42%-0.005%).  Of course my choices are
based on different parameters including wavelenght uniformity or price and
matching polarizers in glasses and projection is a factor to study when
deciding what to use.

> Questions for the list:
> --what films are known to depolarize?
> --the screen used at NSA is a very unusual lenticular screen.  Is it
> possible that this optical element, with very small refracting elements,
> introduces a depolarization?  I imagine the lenticular elements
> (lenticles?) are not hugely larger than the glass beads used on screens.
> --anybody tried 'research grade Polaroid', or is there such a thing?

And I add the question:
--What circular polarizers where used in the SPIE convention and what was
the result?  Which circular filters/glasses are best for 3-D?
--What polarizers does IMAX use for their 3-D theaters (any IMAX people on
the list?  Noel?)


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