- This topic has 3 replies, 2 voices, and was last updated 7 years, 11 months ago by
.
-
Posts
-
Hi everyone.
First of all, I’m really sorry for the confusion and excuse the bad use of terminology.
1. Let’s say I want to calibrate for REC709.
According to the following thread, “the ICC profile always maps from (measured) tristimulus values e.g. CIE XYZ – which define the display colorspace – to display RGB and vice versa.” Original thread here: https://hub.displaycal.net/forums/topic/what-color-space-to-choose-a10e2111/
As I understand, the ICC profile won’t map the TV colorspace from REC709. Actually, there’s no color space choice for the ICC. If you have a wide gamut TV, you won’t actually see only REC709 “colors”, after calibration. Did I understand correctly?My Panasonic TV offers two gamuts: native and standard. I took measurements in HCFR and the native one is about 70% of Adobe RGB and the standard is slightly smaller than REC709. If I’m going to grade for REC709, shouldn’t I set the TV to standard, then calibrate, since standard is closer to the actual REC709 gamut? This way, I won’t oversaturate colors that won’t be able to be displayed.
I have generated a LUT for madVR, installed it, and tried that in HCFR with madTPG. Even with the native (wider gamut) of my TV activated I get close results to REC709.
Or… If all the above is pretty much correct, a LUT with REC709 source colorspace for Davinci Resolve shouldn’t be the best solution?
2. 0-255 Test Pattern vs. TV (16-235)
I’ve found an interesting thread, and I just wanted to share it. I wonder if it is accurate.
According to the thread, a 16-235 test pattern on a RGB Limited TV is doubled mapped: “limited range, inside an already limited range”. To correctly calibrate a limited RGB TV, you must set a 0-255 test pattern. Unless you set your TV to Full. You can find the original thread here (easy to read): http://www.avsforum.com/forum/139-display-calibration/2042442-pc-test-pattern-tv-16-235-a.htmlThank you.
Hi,
As I understand, the ICC profile won’t map the TV colorspace from REC709.
A color transform always involves two color spaces, in the case of ICC color management in the form of two profiles. A 3D LUT is basically nothing else than a baked transform of one color space to another.
If you have a wide gamut TV, you won’t actually see only REC709 “colors”, after calibration.
You always need software or hardware that does the color transform. ICC profiles and 3D LUTs are just files, they do nothing on their own.
If I’m going to grade for REC709, shouldn’t I set the TV to standard, then calibrate, since standard is closer to the actual REC709 gamut?
For any form of color management (including 3D LUTs), it is usually desirable that the display color space is larger than the intended target, so that there is no unnecessary clipping.
I have generated a LUT for madVR, installed it, and tried that in HCFR with madTPG. Even with the native (wider gamut) of my TV activated I get close results to REC709.
madTPG applies the installed 3D LUT to the patterns. If you want to measure the native gamut again, you’ll have to disable the 3D LUT in madTPG.
2. 0-255 Test Pattern vs. TV (16-235)
It is important to note that while there may be multiple possible combinations of settings (e.g. 0-255 pattern -> graphics card driver conversion to 16..235 -> TV, or 0..235 pattern -> no graphics card driver conversion -> TV), there is only one correct outcome (the one where blacks are neither lifted nor clipped, and white is neither reduced nor clipped).
Thank you very much for your fast reply.
I’ve just calibrated my TV and then ran a test in HCFR (with madTPG). As reference, I set REC709 and ITU-R BT. 1886 in HCFR. Everything looks allright, good results, except the primary colors and saturations with 4 to 8 DE values. I think this is due the fact that HCFR use REC709 as a reference and the TV has a larger gamut (it is set to native – wider gamut). Is that accurate?
