On Tuesday, January 15, 2019 at 5:33:51 AM UTC+1, Phil Hobbs wrote:   
      
   >    
   > Yup.  A Glan-Taylor is good for 1:10,000 if you use it right.  Almost as    
   > good as a Wollaston, at least in the transmitted beam.  A "beam    
   > splitting Thompson" improves the refracted beam a fair amount, but    
   > they're not that common.  Stick with the transmitted one.   
   >    
      
      
   I became curious and I read a bit on Wikipedia. The reflected (I think you   
   mean reflected, right?) beam is only partially polarized in these kind of   
   polarizers because they work by total internal reflection of one polarization   
   (which is then completely    
   absent in the transmitted beam) but the other polarization is also reflected a   
   bit (so it is present in the reflected beam).   
      
   I have also read about the difference between Glan-Thompson (is this the one   
   you called "beam splitting Thompson"?) and Glan-Taylor, but to figure it out I   
   would have to spend some time with calculations, so I stop here for the moment.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)