Polarization Mystery

I recently saw two 3D movies using ‘RealD’ technology. The effect was very good. I concomitantly acquired two, presumably identical, pairs of ‘polarized glasses’. The technology is ‘RealD. The document pushes RealD but gives some information on two current alternate technologies. This gives some more detail. These ‘passive’ glasses ‘cost’ $1.50 extra and have no moving or electronic parts. (Some variations requires batteries.) Slide 19 of the first document describes Real D as using circular polarization which I had already deduced. The first chart does not even mention linear polarization.

What’s the Mystery?

I experimented with the glasses with results that I cannot explain. Here are the results broken down by several factors. In each case I send unpolarized light thru lenses to see how the light is attenuated.

('F') “Forwards” means the light travels thru the lens as it would in a theater;
('B') “backwards” is in the opposite direction.
('T') “Turned” means one lens is turned 90 degrees relative to the other.
('P') “Parallel” means not turned. In all cases that I tried 90 degrees and −90 give the same result.
('L') “Left” means left lens and
('R') “Right” means right.

FRPFR	o	Light going thru two parallel right lenses in theater direction is not much attenuated, at least not much more than one lens, but it is given a slight orange hue ('o'). A single lens does not color the light.
FRTFR	b	The same lenses when turned give a slight blue hue ('b'), yet little attenuation.
FyPFz	o	Both lenses do the same, even mixing left and right.
FyTFz	b	ditto. (4 cases)
ByPBz	o	I think thermodynamics requires this considering FyPFz.
ByTBz	b	ditto (no one way mirrors)
xyPxz	o	In general (8 cases)
xyTxz	b	ditto
FRPBR	c	Light travels thru second lens backwards. The hue is gone ('c') “clear”. Very little additional attenuation.
FxPBy	c	four cases here.
FxTBy	0	Black! only a percent or so of blue light gets thru. Notice so far no distinction between left and right lenses!
BxPFx	c	clear again if we match lenses.
BxPFy	p	dark purple if x≠y. Perhaps 20% of light gets thru.
BxTFx	0	black again as for FxTBy.
BxTFy	c	Very clear if x≠y! no coloration.

The lens order is sometime significant which would not be the case for a simple filter. FxTBx is dark while BxTFx is clear. The lenses are not simple circular polarizing filters!

I am confused!

Further experiments would involve a mirror which I think I understand!

Circular polarization