I watched Moon
tonight and I couldn't ignore the fluctuating black luminance. So, when I finished the movie, I muted the sound, placed my spectro to the lower black bar and I started to play with the seek bar while I look at the readings.
I could pick up some scenes with 0.04
, 0.05 , 0.07 , 0.09 , 0.10
cd/m^2 (never in between these discrete points and I couldn't catch a scene in this movie with 0.12 - the highest I ever measured with a specific test pattern...)
But there is a tiny problem with repeatability. I saved two screenshots for each of the five black luminance scenarios. When I started to measure these static images (starting from 0.04 ones) I didn't get any higher readings than 0.07
First I thought the error is in my spectro (like be the black calibration went out very fast, etc...).
But I was sure I saw the difference with my eyes, so I started the playback again and I easily caught some scenes with 0.10 again. Ok, I saved new screenshots. But when I tried to measure those screenshots, the readings were 0.07 again.
So, it seems like there are some thresholds in both directions. If you hit a point where the black goes as high as 0.10, then it remains that high until there is a bigger change. I think it's a safety wall against crazy flickering (like jumping up-down-up-down... many times in just a few seconds...) And it proved to be hard to catch a frame which triggers the 0.10 black luminance. But it's there and if you play the movie you will have 0.10 blacks many times (because it's not too hard to trigger and once it's that high, then it possibly remains there until you hit a significantly lower APL scene...).
And there is some connection between the APL and the black luminance, but it's far form linear. Sometimes I could produce high black levels with relatively low APL but keep low blacks with relatively high APL.
But I guess it's a factor of the above mentioned thresholds (once it went up, it waits up there until it "feels safe" to get lower again...).
This is crazy! I effectively have a display with 0.10 cd/m^2 black level + some stupid "black level extender adjustment" like the back-light control of the LCDs.
I always turned it off on LCDs because it's easier for me to tolerate the high black levels than got mad about the fluctuations.
And you know, an LCD doesn't have ABL. So, if you count that too, then the worst-case contrast ratio of this plasma (with a very high APL scene) is something about 600:1 (~0.1 black and ~60 white).
My old LCD had a stable contrast ratio of ~900:1 (WP calibrated to D65 - this PDP can't even be really calibrated anymore...)
I am proud of myself that I upgraded (ok, there were some other reasons to change - but also other PDP manufacturers for that matter...).
One thing I am not sure about
- Is this some new "rising black" (where the MLL remains low but the peak black level rises with the panel's age)?
- Did blacks got worse with the last firmware update? (Is it easier to trigger the high black levels?)
- Was it always like this and I just didn't noticed it back when the grays also fluctuated (and the old "floating gamma" covered the "floating blacks")?