Originally Posted by Dominic Chan
You won't be able to measure the colours accurately until you first fix the grey scale (i.e., measured Y matching Target Y in addition to D65 colour temperature).
Thanks. I had asked about this previously. After running the gamma auto-cal, the gamma that was in my last zip file is the gamma I get with gamma set to "Normal" and no other tweaks to it. I don't know what the measurement is because I can't tell how to get that number in HCFR, but I can tell from the graph that it is likely 2.1 or maybe even lower.
I have to change the JVC gamma to Custom with a correction value of 2.4. THEN it measures along that gamma curve perfectly.
Also, I thought you were calibrating HDR?
I am. OK, so if you are asking me this then I take it you see something that I do not have set up properly in HCFR? I use BT2020 both for HDR and SDR, as I have other sources like a settop box and Nvidia Shield that output SDR BT2020. In that case the only thing I will change is gamma from Custom/2.4 (SDR) to Custom/Import to get to my custom curve for HDR.
I did notice that the colors measure wacky if I calibrate around the custom import HDR curve. Maybe I have to adjust something in HCFR to tell it to do HDR? Where/what setting is that please? In the meantime I have just assumed that I could calibrate BT2020 with filter and SDR, and when I put it into a gamma mode for HDR everything will still be calibrated, no?
In the GDI dialog box you can change the stimulus level. I usually use only 2 levels.
So you change the stimulus level there from 100% to 50%? And you change the "4" levels measure to 10, in order to get the 10% increments, so you can get 50-100% if making the custom profile (the measures needed for Chad's spreadsheet)?
Do not touch the Axis adjustments.
When you adjust hue and saturation based on the CIE diagram, keep an eye on the delta-L. If it changes (most likely it will), adjust the colour brightness and iterate.
I spent a lot of time experimenting with the CMS today. Basically I used the CIE chart to gauge if any colors Saturation needed to be increased or decreased, and watched the dE for 0, 25%, 50%, 75% and 100% for RGBCYM and tweaked until I had the dE as low as possible.
There is an icon for the gamma, next to the luminance.
Does that icon tell me what the gamma actually measured? I can't find where HCFR provides the gamma measure, only can see the gamma curve itself.
Can you tell me where specifically the gamma icon is? I wasn't sure what you meant by "next to the luminance".
I just looked around for a while and don't see what you're referring to.
Are all the gamma controls (Picture tone, Dark Levels, etc) all in their neutral positions? Make sure HCFR Reference is set for gamma 2.2, not BT.1886 (the default).
Yes all gamma controls were neutral. But BT1886 was set as the default, and even with it changed to 2.2 as the target, I had to change the JVC gamma setting from "Normal" to "2.4" and then it matched the 2.2 curve. From there the colors measured much better.
It’s best to measure 20 points as the anomalies tend to be in the low end, between 5 and 10%. You can also make a Near Black run to check in finer steps.
If there is an anomaly in the low end, is there anything we can do about it anyway?
After hours of experimenting today, here's where I am:
1. I have completely given up on trying to use BT2020NF. First, my net brightness gain after calibrating each mode to D65 is only 8% without the filter. Even tho I am STARVED for HDR nits, that's just not enough to really make much of a difference. And second, the colors are a mess. The gamut is screwy with BT2020 without the filter, and the success I had with the CMS and excellent dE's is no where in reach in BT2020NF. So especially given its only a 8% net gain for me in nits, its not worth the poor dE (color check 2.5 - 5 for many colors).
2. On the other hand, I had great success getting an excellent calibration for BT2020 with the filter. I did the autocal and then used the RGB gains to dial in 100%, which gave me a grayscale dE of 0 - 1.2ish IIRC. Then I made some tweaks with the CMS and got the color checker to mostly be under 1.0 dE across the board. There are a few color checker points that were maybe 1.5ish but for the most part the results are about as good as when I used to use CalMAN and a Lumagen with 3D LUT on my previous projector VW95 about 4-5 years ago.
Can you please have a look at the attached chc file, which is my latest. This is the i1D3 measuring off the screen, with everything dialed in an the JVC set for a 2.4 gamma. I measured basically everything, including many of the color checker options, grayscale, primary/secondaries/saturations, near black/white etc. I'd love to know if you think I've nailed it or if you think it still needs work.
And there's also the question now I have why you are asking "I thought you are calibrating HDR?" so I am thinking I setup HCFR wrong or perhaps there's a way to set up HCFR so I can use my custom gamma curve with HCFR in the calibration/testing?