Rec 709 & sRGB

Apologies if this has come up before but I didn’t see anything in my searches.
I recently purchased 2 BenQ PD2500Q monitors and they have (among others) a Rec 709 and sRGB modes...

The Rec 709 mode has ALOT of contrast compared to the sRGB. I realize my monitor hasn’t been calibrated outside of BenQ’s factory setting (though they say it’s individual, I got a certificate and everything lol)

If I grade for Rec 709 it looks OK on my phone but very flat on my TV and my wife’s laptop.

Anyone have any ideas where I could improve to be able to grade so it looks acceptable in both Rec 709 and sRGB?

Ursa Mini Pro RAW 3:1 footage rendered out of a Mac Pro 5,1 with data levels on full (otherwise it renders washed out)

Thanks All!

Variants of this question have been covered to death on this and every other color grading forum. The answer is always the same.
The only way to get a REC709 image on screen from Resolve that you can trust is to run SDI out of a BMD card (or BMD Thunderbolt box) to an accurately calibrated reference monitor.
Grading by viewing the image in the GUI just doesn't work.

As I understand it, the gamma is exactly what’s different between sRGB and Rec. 709.

There’s no way for the same file to change gammas depending on what device it’s on. You just have to choose one.

You could also just split the difference, knowing that the blacks will be a little crushed on one and a little lifted on the other.

FWIW, I think the industry standard is moving toward BT.1886, which is close to 2.4, so you might just grade to that, and take comfort knowing that you can’t control exhibition out in the world, once it’s released.

My philosophy is to grade for the discriminating consumer, and reward that consumer who HAS taken the time to set everything up properly.


Sent from my iPhone using Tapatalk

As I understand it, the gamma is exactly what’s different between sRGB and Rec. 709.

There’s no way for the same file to change gammas depending on what device it’s on. You just have to choose one.

You could also just split the difference, knowing that the blacks will be a little crushed on one and a little lifted on the other.

FWIW, I think the industry standard is moving toward BT.1886, which is close to 2.4, so you might just grade to that, and take comfort knowing that you can’t control exhibition out in the world, once it’s released.

My philosophy is to grade for the discriminating consumer, and reward that consumer who HAS taken the time to set everything up properly.


Sent from my iPhone using Tapatalk

Thanks so much for the reply.
Yea the file never changes just my view of the grade.
I’ve read a lot of forums and people always seem to say they there isn’t really a difference so when I saw one I thought I’d throw it out there.

Thanks for taking the time to answer!
Best,


Sent from my iPhone using Tapatalk

Variants of this question have been covered to death on this and every other color grading forum. The answer is always the same. ...

The answer can actually be a bunch of different things.

Most probably in this case is that BenQ monitor's Rec 709 profile is mapping 1 to 16 to black and 235 to 255 to white. While in sRGB only 0 is black and only 255 is white. That would explain a huge contrast shift.

What profile should you use? Which profile matches the delivery medium the best? You have test out to how people will actually view the image.

The answer can actually be a bunch of different things.

It could be a million different things IF you're grading in the GUI. That's why the first answer to the question is simple: Don't grade in the GUI.
Run SDI out to a calibrated monitor so that you can first be certain that your signal is right, then you've got a reference from which you can try to determine why your rendered grade looks different on different screens.

Some people just don't want to hear this, so they stick their fingers in their ears and pretend that it's possible to grade directly in the GUI on cheap computer monitors. But it's a fantasy. No colorist would spend thousands of dollars on i/o cards and reference monitors if they weren't absolutely necessary. Telling the OP that he can determine the cause of the discrepancies he is seeing without first having a solid reference image on an external monitor from which to make comparisons is simply sending him off on a fools errand.

+1

The answer can actually be a bunch of different things.

It could be a million different things IF you're grading in the GUI. That's why the first answer to the question is simple: Don't grade in the GUI.
Run SDI out to a calibrated monitor so that you can first be certain that your signal is right, then you've got a reference from which you can try to determine why your rendered grade looks different on different screens.

Some people just don't want to hear this, so they stick their fingers in their ears and pretend that it's possible to grade directly in the GUI on cheap computer monitors. But it's a fantasy. No colorist would spend thousands of dollars on i/o cards and reference monitors if they weren't absolutely necessary. Telling the OP that he can determine the cause of the discrepancies he is seeing without first having a solid reference image on an external monitor from which to make comparisons is simply sending him off on a fools errand.

I'm not telling him that he should grade w/o an external monitor connected to card or breakout box. What I am saying is that the cause of playback not being what's expected can be a MULTITUDE of things. It's a very complicated issue.

Yes, if you remove OS, program and monitor profiles from the equation, which using an external calibrated monitor and BOB does, then you eliminate all of those things. But it doesn't explain what was going on with the OP's setup nor does it answer his question about why his monitor is showing such a huge difference between Rec 709 and sRGB. And even if he uses a Decklink card to connect the monitor, the disparity between the monitor's Rec 709 and sRGB settings will still remain.

Peter Moretti is correct. The BenQ monitor is expecting bit 16-235 video levels for the rec709 setting and 0-255 for the sRGB setting. These correspond to the Resolve 'Video' and 'Full or Data' output settings. Rec709 rendering from Resolve should always be set to Video levels, sRGB rendering should be Data.

Yes, if you remove OS, program and monitor profiles from the equation, which using an external calibrated monitor and BOB does, then you eliminate all of those things. But it doesn't explain what was going on with the OP's setup nor does it answer his question about why his monitor is showing such a huge difference between Rec 709 and sRGB.

Exactly. And it seems pretty clear from his setup that he's grading in the GUI. So, first the OP needs to be running SDI out to eliminate all those variables. Once he's got a signal on a reference monitor that he can trust, only then can he begin to troubleshoot what he's seeing on other screens. That's all I'm saying.

The BenQ monitor is expecting bit 16-235 video levels for the rec709 setting and 0-255 for the sRGB setting. These correspond to the Resolve 'Video' and 'Full or Data' output settings. Rec709 rendering from Resolve should always be set to Video levels, sRGB rendering should be Data.

I doubt that's the issue. If you choose "auto", virtually all video file formats exported out of Resolve will have flags in the header of the wrapper that denote how the file should be displayed. Manually setting to "video" or "full" can screw that up. The colorists at MixingLight.com (who I trust about such things) recommend sticking with "auto" unless you are rendering out image sequences or intermediate files for another software or hardware that you know for sure can't or won't read the flag properly.

Thanks so much for all the responses! Very helpful tips and things to continue to learn. Definitely putting a good reference monitor on the priority list. Still have more research on the whole “video” vs. “full” data thing. Auto always washes out my video but I haven’t tinkered with the setting for the timeline or anything yet, I just let Resolve do it. I think my issues can be tracked to the monitor and grading in the GUI but I was surprised at how big a difference these monitors had in the Rec 709 & sRGB spaces.


Sent from my iPhone using Tapatalk

It could be a million different things IF you're grading in the GUI. That's why the first answer to the question is simple: Don't grade in the GUI. Run SDI out to a calibrated monitor so that you can first be certain that your signal is right, then you've got a reference from which you can try to determine why your rendered grade looks different on different screens.

Some people just don't want to hear this, so they stick their fingers in their ears and pretend that it's possible to grade directly in the GUI on cheap computer monitors. But it's a fantasy. No colorist would spend thousands of dollars on i/o cards and reference monitors if they weren't absolutely necessary. Telling the OP that he can determine the cause of the discrepancies he is seeing without first having a solid reference image on an external monitor from which to make comparisons is simply sending him off on a fools errand.

I think this is very wise advice.

It could be a million different things IF you're grading in the GUI. That's why the first answer to the question is simple: Don't grade in the GUI. Run SDI out to a calibrated monitor so that you can first be certain that your signal is right, then you've got a reference from which you can try to determine why your rendered grade looks different on different screens.

Some people just don't want to hear this, so they stick their fingers in their ears and pretend that it's possible to grade directly in the GUI on cheap computer monitors. But it's a fantasy. No colorist would spend thousands of dollars on i/o cards and reference monitors if they weren't absolutely necessary. Telling the OP that he can determine the cause of the discrepancies he is seeing without first having a solid reference image on an external monitor from which to make comparisons is simply sending him off on a fools errand.

I think this is very wise advice.

Agreed


Sent from my iPhone using Tapatalk