Originally Posted by R Harkness
I thought that, provided a display with good contrast, you can calibrate for higher gamma and maintain shadow detail; the trade off not being shadow detail loss, but overall perception of a slightly dimmer image (but with a perception of greater depth, so long as the image is bright enough).
And that's actually one of the main reasons I recently bought calibration software and a Lumagen: so I could calibrate for a higher than 2.2 gamma (which I believe the person who calibrated my projector prefers), but maintain shadow detail.
(I wanted to calibrate for more than one gamma setting - one lower, another higher, to choose as I wish per content).
Is it true that going for a higher gamma in of itself *necessarily* sacrifices some shadow detail?
Originally Posted by Highjinx
I recall a good discussion, participants were Greg Rogers &( DarinP I think ), the general consensus was the higher the Native CR the higher the gamma could be without crushing shadow detail.
That's exactly it Highjinx, in order to not crush blacks on your display you need to have enough on/off so that for the gamma curve you are targeting the targets in the low end are not so close to your actual black level that you end up crushing shadow detail. The consensus was that in a light controlled environment you needed around 50000:1 effective on/off in your set up to target a 2.4 gamma.
And Rich, you are correct, going for a higher gamma in itself doesn't necessarily sacrifice some shadow detail, as long as the gamma you are aiming for is not too high regarding the actual on/off your display can deliver in your set up and your environment.
The beauty about BT1886 is that it takes your actual black level into account so that you can target a gamma curve as close as possible to a power 2.4 without crushing blacks.
For example, before BT1886, with my rs45 which would crush blacks using a flat power 2.4 curve, the only "standard" alternative was to use a power 2.3 or 2.35 gamma (so that the gamma targets at 5% were sufficiently above my black level) but that meant I also had to target a lower gamma for all the other targets.
Many calibrators and enthusiasts (including me), before BT1886 became a standard, would use a custom gamma curve which looked very much like BT1886 to come out of black faster, with custom targets at 5%-30% going slowly from 2.2 to 2.4 and then use 2.4 for the rest of the curve. It achieved the same thing, but there was no formula providing a way to calculate each target. Everyone defined their own targets. BT1886 makes it possible for everyone to use the same standard.
Higher model owners with higher specs shouldn't think they can automatically target a 2.4 gamma without crushing blacks. I get 40000:1 at mid throw with my iris fully closed on the rs45, but an rs55/56 owner at short throw with the iris fully open would get LESS on/off, around 30000:1. What matters is what you model delivers in your set up, not the maximum specs of your model which can only be achieved at long throw with the iris fully closed (and I mean fully closed, with the dual iris models, open the iris more than a few notches and most of the on/off advantage vs a single iris model goes away).
So to go back to BT1886, say your display has a theoretical black level of zero. The targets for the gamma curve will be identical to a flat power 2.4. As the black level of the display raises, the targets go down (higher luminance in the low end compared to 2.4) to make sure that you do not crush blacks. BT-1886 always allow you to get as close to power 2.4 as your display allows, which is why it's so great.
So as long as your meter is able to read your black levels (or you provide your software your actual black levels manually as discussed), there is nothing to lose from using BT1886, even with a display which can accommodate a flat 2.4, as with a PJ with very good black levels and on/off like the rs57, you'll get something very close to a flat power 2.4, except that it won't crush blacks even slightly.
For example, my rs45 at long throw would use a BT-1886 curve closer to 2.4 in the low end, because my absolute black level will be lower and my on/off would go up (from say 35000:1 to 50000:1).
Regarding having the same gamma as used during mastering, forget it, there is no standard between post-production houses, and even within the same post-production facility sometimes! Colorists use anything between 2.2 and 2.6, most using something around 2.35 or 2.4, but forget about reproducing what was done during mastering, there is no way to know. The only thing you can do is make sure you don't crush blacks (and using BT-1886 is the best way to achieve that) or reveal detail that is not meant to be shown (by raising brightness too much and showing Blacker Than Black levels). Hopefully BT1886 (rec 709 + gamma 2.4, they don't crush blacks on their expensive monitors) will become the standard in post-prod houses, so aiming for it should get you as close as possible to what was mastered, at least for new releases (not only new movies, it can be new releases of old movies).
I hope I haven't sent everyone to sleep, it would be a shame just before New Year's Eve! Edited by Manni01 - 12/31/13 at 8:12am