[Laser] Polarization Subtraction

[Dave]

On 2012-10-19 12:03 PM, laser-request at mailman.qth.net wrote:
> Date: Thu, 18 Oct 2012 21:48:56 -0700 (PDT)
> From: Tim Toast <toasty256 at yahoo.com>
> Subject: Re: [Laser] Polarization Subtraction
> To: laser at mailman.qth.net
> Message-ID:
> 	<1350622136.35828.YahooMailClassic at web161002.mail.bf1.yahoo.com>
> Content-Type: text/plain; charset=us-ascii
>
> More about the camera and adapting the concept to optical comms.
>
> At first i thought of the rotating or electronically switched
> (LCD)filters too but i guess it's not that complicated really.
> At least in the case of this special camera. Just a lens + plane
> beam splitter with two fixed polarization filters at right angles
> and two CCD's in their camera.
>
> The scheme appears to work over whatever the exposure time they
> used to take the photo's, 1/1000th to 1/10th a second or so. I
> suspect in the case of our optical comms, with photodiodes instead
> of CCD's, it would work continuously, with the two polarized
> signals being continuously subtracted in real time. And with
> enough signal to make the loss in light level tolerable due to
> the splitter (-3 dB?) and the filters (-4 or -5dB ?) I think some
> only let 30 percent of the light through. It seems that this could
> not work with very weak signals. It could be that this thing is
> only usefull in fog with moderate optical signals - wouldn't that
> be strange?!
>
> It may be that the path length from the single lens optic to the
> focal points of both beams are important. With a plane ordinary
> beam splitter, the beams are going to be exactly the same
> length assuming the photodiodes are positioned well enough
> - normally within a millimeter or so either side of the focal
> point. I think light travels about a foot per nanosecond. So the
> two beams would be well within that distance. There would need to
> be a whole foot difference in the path length to have a one
> nanosecond difference in arival times for the photons. This is
> very - very simultaneous! A few millimeters of light travel time
> - juat a few picoseconds.

One could also use fiber optics to route the light.  You'd probably want to
have the polarization filter before the fiber, since I'm not sure how badly
the fiber would mangle the polarization information.

> Not that the polarization in the scene is changing at ultra fast
> rates but if it were, this capture device should still get a
> "temporal frame of any length" nearly simultaneously - within
> a few picoseconds. I think most of the polarization info is
> almost static anyway and not changing very fast other than
> being jumbled around a bit by the air pockets or cells that
> cause scintillation sometimes. The fog or haze is scattering the
> light photons randomly but is it scattering the polarization
> angles too? Or is it affecting the scene polarizations in other
> ways? IDK. Maybe the polarization info in the image or scene is
> not affected by scattering fog at all. Meanwhile, the degree
> of simultaneousness in the camera exceeds all that by an
> enormous amount.

There are several magneto-optical effects which may (or may not) affect the
polarization of the beam:

http://en.wikipedia.org/wiki/Faraday_rotation
http://en.wikipedia.org/wiki/Voigt_effect
http://en.wikipedia.org/wiki/Cotton%E2%80%93Mouton_effect
http://en.wikipedia.org/wiki/Cotton_effect

Some of these may be sensitive as to whether the optical path is parallel
to the Earth's magnetic field or perpendicular to it.  Some will also depend
upon the material in the optical path (air, fog, etc.).

> -toast

Dave