Originally Posted by Wizziwig
Have you missed the links where Darin admits that letterbox bars were darker on a 3000:1 on/off (1000:1 ANSI, iris open) DLP vs. a 50000:1 JVC on many scenes? The same thing is happening in the rest of the image and causes visible washout of black areas.
Sounds like you are talking about a split screen comparison I mentioned. This doesn't mean it ends up being nearly as visible when a person is watching one projector at a time, although that applies to lots of image characteristics like the black floor also.
In general the bars are bigger and easier to spot differences in than in smaller black areas within the image area. Masking can make it harder to see the differences also.
I recall one scene in one of the James Bond movies from a few years ago in a desert where a guy had a black suit on and this black suit held better with my Planar than JVC, but in general I had to work much harder to find scenes where I could see advantages to the ANSI CR beyond a point. Even then I doubt I did enough testing to make 100% sure that what I was seeing was ANSI CR differences, although it likely was. It definitely wasn't that kind of difference that would pull me out of a movie like on/off CR between those 2 projectors was more likely to.
Also, when we actually tried doing some of this testing where a viewer didn't know whether they ANSI CR was low or not the subject failed pretty badly at being able to see differences. I would put a screen up behind them that would lower the ANSI CR off the screen and they weren't able to tell me whether the ANSI CR was higher or lower. We only tested some scenes and with more material may have found things where it would be easy to tell.
By using one projector (a Sharp DLP) and lowering ANSI CR by adding reflections behind the viewer we were keeping many parameters other than ANSI CR constant. That is hard to do when comparing to different technologies.
I know you've said you can see the differences on the ANSI CR checkerboard, but it would be interesting if you tried the same kind of thing with a single projector and just adding white surfaces behind you. Could you quickly tell when the system ANSI CR was higher or lower on 2.35:1 content without screen masking so you could watch the black bars? With screen masking so you couldn't focus on the bars?