P3D Re: more on circ polarizers

[Tom Hubin <thubin@xxxxxxxxx>]

Hello Bob,

> I've had a message from the fellow who is pursuing the
> recording project that I've mentioned previously. He says...
> 
> ">Well, I have tried them, but they do not eliminate blue wavelengths
> >entirely.  Consequently, there can be color fringing on the highlights when
> >cross polarizing.  Zeiss plano optical glass with a linear polarized
> >material sandwiched in between is the best that I have ever tested, bar
> >none.  The absorption curves from Polaroid bear this out.
> >
> "
> 
> This is interesting, no? Has anyone here seen this effect using
> circular polarizers (ie, is it generally strong enough to be
> objectionable?).

This is not surprising. The circular polarizer is made of a layer of
linear polarizer and a layer of quarter wave retarder. Both layers must
perform well at the wavelengths of interest. The simplest of quarter
wave retarders works properly at only one wavelength. It actually causes
more of a time difference in a pair of linear phases. This time
difference is a quarter wave shift at only one wavelength. At other
wavelengths it can be pretty poor.

More complex materials permit a nearly quarter wave shift over a broad
spectrum. But eventually you reach their design limits and get shifts
other than quarter wave. When this happens the circular polarizer puts
out elliptically polarized light. This is a combination of mostly
circular and some linear. The receiving circular polarizer has a similar
problem. As a consequence some wavelengths will leak through. Usually
very blue and very red.

You need to either filter out the blue so that it is not a problem or
get a broader spectrum circular polarizers. Get graphs of phase shift
versus wavelength for your circular polarizers. It would be good for
your scientific documentation of this project.

As an alternative, you can use a several very good narrow band circular
polarizers with color filters so that only light within each polarizer's
design spectrum is photographed. Photgraph different bands of light then
combine the results digitally. If the bands are narrow enough you can
even use monochrome film and still produce a digitized color photo. 

Tom Hubin
thubin@xxxxxxxxx
AO Systems Design


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