Originally Posted by derek
And yes you do loose light to your eye through the glasses but we're talking output of the projector.
I'm talking specifically about LCD shutter glasses
. Some DLP projectors have used LCD shutter glasses and you still
need a black frame (doesn't have to be symmetrically the same size as visible frame). Basically, a black frame of about 0.5 milliseconds to accomodate the LCD shutter open/close time. That said, the longer you wait for the shutter to close, the more closed it is (more opaque), so there's favour in waiting a bit longer. The opaqueness of LCD shutters increases roughly logarithmically versus time and plateaus out gradually, 0.5ms is not necessarily the "fully closed" LCD shutter point, at least for many models of LCD shutter glasses. High speed video of LCD shutter glasses shows that an LCD shutter is only about ~90% opaque after 1ms, that still produces 10% crosstalk. There's a crosstalk-versus-brightness tradeoff. Shorter black frame, more crosstalk. Longer black frame, less crosstalk. Law of physics. Doesn't matter what projector technology, if you're using LCD shutter glasses
(I actually have some old high speed video footage, which I should be posting. Some readers here are already familiar with my high-speed video of LightBoost
, but I also captured some high speed video of LCD shutter glasses too (not currently published yet). The ones I have, nVidia 3D glasses, do NOT
fully reach >99%+ opacity in 0.5ms. It does mostly go opaque in less than a millisecond, but then gradually becomes more and more opaque as the microseconds tick by, in essentially what resembles like an approximate logarithmic curve... the less opaque, the more crosstalk will occur. If you are only 95% opaque, you will have 5% crosstalk leaking through -- e.g. light gray showing through a white-black vertical edge boundary, etc). Rating of LCD response is often measured to a certain completeness, otherwise, a lot of 2ms LCD's would measure more than 5ms, etc. To minimize crosstalk, you need to choose a "picky" completeness requirement -- that 0.5ms LCD can become a 2ms LCD when you're measuring to, say, 99.5% completeness of transition (0.5% crosstalk leakage), instead of a 95% completeness (5% crosstalk leakage), etc. Voila. Bigger black frame obviously needed.
There are faster LCD shutters that claim 0.1ms open-to-close speed, but I'd really love to see that "light leakage curve" at the microsecond scale. Is it truly all the way to 99.5%+ opaque in 0.1ms? (even 0.5% is still visible crosstalk). It's wholly possible the projector manufacturer decided to add a bigger black frame interval anyway (0.5ms of black frame and get about ~8% dimming relative to 8.3ms (1/120sec) refresh cycle, or 2ms of black frame, and get about ~33% dimming (relative to 8.3ms refresh cycle). Since 99% opacity vs 99.5% opacity is a MASSIVE difference in crosstalk (twice as much crosstalk -- leakage of 0.5 IRE on a full-white, versus leakage of 1.0 IRE on a full-white). I'd love to see the high speed video footage of what is happening on the projector screen...
For *LCD shutters
....more black interval during eye swap interval (shutter open/close), less crosstalk.
....less black interval during eye swap interval (shutter open/close), more crosstalk
....It's the law of physics
....and it shows in high speed 1000fps video footage (of high speed camera pointed through a shutter glasses).
This does not apply to polarized 3D glasses, which some DLP projectors are using (e.g. alternating polarization). For that, you don't have the concept of shutter open/close time.