An idea RE: 'transluscent' mirrors

[unpopular]

So I was thinking about the SLT thing and how that the concept could be improved. I was wondering what if they (ahem: sony) used liquid crystal to polarize light from the lens, and then pass it through a polarizing beamsplitter such that P-polarized light is transmitted to the imager and S-polarized light is reflected to the evf? The liquid crystal state would be synchronized to the shutter, such that it would direct 100% of the light to be transmitted or reflected to appropriate device as needed. If the shutter is open, the the light would be transmitted to the imager, otherwise it would be reflected to the viewfinder.

Unlike the polarizing filters we are familiar, a thin layer of liquid crystal should be fairly transmissive and could be integrated into the beamsplitter itself, so overall loss should be minimal.

Obviously, you'll have a permanent polarizer built in, which is a drawback that could be marketed as "better contrast" while circular polarizers could be used as a variable ND.

 

[KmH]

Thank you, my patent application is now pending. :lmao:

 

[unpopular]

You know. I'm happy. I always wanted a continuously variable ND filter. After considering this hypothetical, I know now that they're like WAY over priced!

So KmH - make your millions with Sony - I just saved myself ninety bucks!

 

[Derrel]

That near-sighted donkey avatar with thick glasses...that's VERY amusing!!!

"Transporter Room, party of three, beam us up NOW!"

(Yeah...like this polarizer system is not the one that Kirk saw the alien life forms using on' The Battle on DelusionalSix'!)

 

[unpopular]

<<<< Technically accurate ass.

 

[mjhoward]

I can think of 3 major issues with an idea like this:
1. The polarized image you saw in the EVF wouldn't resemble what would be collected when non-polarized light hits the sensor once the shutter is clicked.
2. Battery life, I would image, would be much shorter
3. Doesn't the response time of Liquid Crystal vary based on temperature?

 

[unpopular]

1) The light hitting the sensor would be polarized in the complimentary pole. What this means as far as appearance I am not sure. Keep in mind that the light isn't being filtered in the way a polarizer filter does, but rather taking randomly polarized light and aligning it. So Actually I am not sure that you'd see the same effects as a polarizing filter would.
2) The liquid crystal would not likely consume a lot of energy. There are some liquid crystal arrangements which maintain their last state without electricity, and others which will maintain one state normally and the other with applied current. So no. I don't think battery would be an issue, and would consume much less energy than a mirror movement mechanism.
3) This is not to replace the shutter, only the mirror. At normal operating temperature (according to wiki) ferro liquid crystal has a response time of 100µs. Even if that increases to 1ms in extremely cold conditions, it wouldn't be perceived by the operator.

 

[Helen B]

It's a great idea. I like it. Of course the light coming out of a correctly-mounted circular polariser should be, er, circularly polarised, which comes as a bit of a shock, I know. This means that your system wouldn't act as a variable ND filter when used with one.

I'm not sure about the effect that the liquid crystals would have. Don't they normally rotate the plane of polarisation when 'off' and leave it alone when they are 'on'? Do they actually polarise light on their own? Wouldn't you have to use one with a polarising foil in front? Would the birefringence cause problems or would they never be used in a birefringent condition? I'm not sure what would happen at your beamsplitter either - wouldn't the non-reflection of the p-polarised light be angle-dependent even if the beamsplitter was aligned to the axis at the perfect Brewster angle? These are just musings and questions I'm asking myself because there is a lot I don't know about LCs. As mj says the different polarisations of the viewfinder image and the recorded image might be a teensy-weensy problemette, but not if you included a quarter-wave delay plate in front of the liquid crystal (or in front of the linear polariser in front of the LC). Then you would have to use a linear polarising filter instead of a circ in front of the lens. I'll stop there.

 

[unpopular]

Don't they normally rotate the plane of polarisation when 'off' and leave it alone when they are 'on'? Do they actually polarise light on their own?

I think that if you used a ferrofluid, you could place a magnet in the P or S position and select one or the other

Ferro Liquid Display - Wikipedia, the free encyclopedia

Wouldn't you have to use one with a polarising foil in front?

Maybe I am not understanding everything correctly - the LC would rotate the light into one of two polarities, S or P and the beamsplitter would either reflect or transmit based on the polarity. The LC would act as a waveguide.

Would the birefringence cause problems?

Probably. But this could be compensated for, I'd imagine?

I'm not sure what would happen at your beamsplitter either - wouldn't the non-reflection of the p-polarised light be angle-dependent even if the beamsplitter was aligned to the axis at the perfect Brewster angle?

Again, the idea is that all the light entering the beamsplitter will be aligned into a predictable polarity. Provided that liquid crystal can actually do this, then only the waveguide and the beamsplitter would need to be aligned.

 

[rgregory1965]

Others could just steal the idea from Sony....worked for Bill Gates