- Posts: 5
- Joined: Thu Dec 12, 2019 4:15 pm
- Real Name: Tobias Schachinger
Sat Mar 07, 2020 10:53 am
Hi everyone,
just when I thought I understood the basics of color management, I looked through the HDMI specifications to find out what the difference between the versions was, and found something quite confusing: the whole thing about "supported color spaces".
This is not really any kind of problem that I'm having with my workflow or anything, I'm just interested in getting a broader understanding of video technology, and I'm hoping someone can help clear up my confusion.
This was my prior understanding of what color spaces mean in the context of video post-production:
The display device receives a signal that corresponds to certain RGB values, and it uses those values to control the intensity of light sources that are located behind color filters to produce a color image. What that image looks like, i.e. what primary colors and resulting color gamut the monitor can reproduce, depends on the physical properties of those light sources and color filters.
Color spaces like Rec. 709 or DCI-P3 are standardized sets of primary color chromaticities (that were defined together with a number of other properties) that I can calibrate my monitor to if I want to make sure that the image that I'm seeing will look the same on someone else's display device that was calibrated to the same standard. And if that other person's display device was calibrated to a different standard, I can apply a color space transformation to my footage to achieve the same result. Those two scenarios - calibration and transformation - are the only times when color space has anything to do with the actual signal values. But basically, a video signal is just a series of values (be they RGB or subsampled YCbCr) that is processed by the display device, and color space is not an inherent property of the signal (though, of course, the intent to show the image on a display device that was calibrated to a certain standard kind of makes that standard a property of the signal, but that's not something I can read out of the signal itself, only possibly out of metadata).
So far so correct?
If so - what does it mean that HDMI versions up to 1.4b don't "support" Rec. 2020? If my display device is capable of reproducing the full Rec. 2020 spectrum (which is... unlikely, I know), and it's calibrated to that standard, and it receives a signal that corresponds to 10-bit RGB(1023,0,0), then the chromaticity output will be very close to the primary red chromaticity defined in Rec. 2020 - what does that have to do with the HDMI ports in the signal chain? What am I missing here?
Thanks to all those of you who have read this entire novel without giving up, and thanks in advance to those who will actually take the time to answer me
just when I thought I understood the basics of color management, I looked through the HDMI specifications to find out what the difference between the versions was, and found something quite confusing: the whole thing about "supported color spaces".
This is not really any kind of problem that I'm having with my workflow or anything, I'm just interested in getting a broader understanding of video technology, and I'm hoping someone can help clear up my confusion.
This was my prior understanding of what color spaces mean in the context of video post-production:
The display device receives a signal that corresponds to certain RGB values, and it uses those values to control the intensity of light sources that are located behind color filters to produce a color image. What that image looks like, i.e. what primary colors and resulting color gamut the monitor can reproduce, depends on the physical properties of those light sources and color filters.
Color spaces like Rec. 709 or DCI-P3 are standardized sets of primary color chromaticities (that were defined together with a number of other properties) that I can calibrate my monitor to if I want to make sure that the image that I'm seeing will look the same on someone else's display device that was calibrated to the same standard. And if that other person's display device was calibrated to a different standard, I can apply a color space transformation to my footage to achieve the same result. Those two scenarios - calibration and transformation - are the only times when color space has anything to do with the actual signal values. But basically, a video signal is just a series of values (be they RGB or subsampled YCbCr) that is processed by the display device, and color space is not an inherent property of the signal (though, of course, the intent to show the image on a display device that was calibrated to a certain standard kind of makes that standard a property of the signal, but that's not something I can read out of the signal itself, only possibly out of metadata).
So far so correct?
If so - what does it mean that HDMI versions up to 1.4b don't "support" Rec. 2020? If my display device is capable of reproducing the full Rec. 2020 spectrum (which is... unlikely, I know), and it's calibrated to that standard, and it receives a signal that corresponds to 10-bit RGB(1023,0,0), then the chromaticity output will be very close to the primary red chromaticity defined in Rec. 2020 - what does that have to do with the HDMI ports in the signal chain? What am I missing here?
Thanks to all those of you who have read this entire novel without giving up, and thanks in advance to those who will actually take the time to answer me