- This topic has 8 replies, 2 voices, and was last updated 7 years, 2 months ago by
.
-
Posts
-
Hello Florian
I’ve noticed some changes (Particularly with gamma) to my calibrations since the last couple of updates and I’m wondering if any changes were made that might be causing this. .. My settings are always the same:
Default settings plus…
6500, 2.2, 120, speed ‘low’, xyzl lut + matrix, using ciecam02 perceptual, lumanace matched, darkened room lighting, 11,000 patches.
I’ve noticed some inconsistent results. Sometimes highlights are blown out (gamma too high at high end) and other times I get lower contrast (maybe normal) contrast throughout.
Also, I’ve noticed that skin tones have changed as well. Not necessarily bad, just different.
Is it better to use ciecam02 settings during the first run, or is it better to apply it to an existing profile? (Would there be any difference?)
Next, is there such a thing as calibrating and existing calibration and or profile to further ‘fine tune’ that calibration\profile? … If so, could you tell me how to accomplish this in dispcal. is it a good idea?
Finally. I’ve noticed that if I want to create a new calib\prof but use the same filename as the last (current) one, dispcal creates a file called ‘xxx.cal’ in the settings drop-down box of the old one.. What is this? Why is it retained? How to remove it?
Thanks so much. 🙂
Hi,
I’ve noticed some changes (Particularly with gamma) to my calibrations since the last couple of updates and I’m wondering if any changes were made that might be causing this.
No.
I’ve noticed some inconsistent results. Sometimes highlights are blown out (gamma too high at high end) and other times I get lower contrast (maybe normal) contrast throughout.
Is the display stable? 11000 patches seems a lot for an RGB device, and in case it’s not stable, reducing the number of measured patches may help.
Is it better to use ciecam02 settings during the first run, or is it better to apply it to an existing profile? (Would there be any difference?)
It’s not possible to “apply” CIECAM02 gamut mapping to an existing profile – the profile is always regenerated completely from scratch using the measurement data.
Next, is there such a thing as calibrating and existing calibration and or profile to further ‘fine tune’ that calibration\profile?
It is possible to update the 1D calibration for an existing profile, but that’s not for ‘fine tuning’ but for saving time – if (e.g.) the display has only drifted slightly, updating the calibration may be enough.
Finally. I’ve noticed that if I want to create a new calib\prof but use the same filename as the last (current) one, dispcal creates a file called ‘xxx.cal’ in the settings drop-down box of the old one..
This only happens when you profile off of an existing calibration.
How to remove it?
Just select and click the trashcan icon.
Thank you, so can the ciecam02 gamut mapping be generated during the first run by setting it up in the cog box, or do I recreate it using the measurement data later? …Same result??
Doesn’t matter, the result will be the same. Apart from that, if you already know you want CIECAM02 there’s no reason not do it in one go.
Cool…So what about going in the opposite direction? ie, take a profile which already has ciecam02 applied and create a profile from measurement data without using. (Bringing it back to ‘normal’)
Also, While doing a calibration, a random window popped up behind the measurement swatch…Does this sort of thing have any effect on the readings? …Even a little bit? (Light coming from beyond the sides of the swatches perhaps, or from underneath.)
And finally, this ‘black point’ adjustment thing. What exactly does it mean? Is it just the ability to adjust the rate at which the gradation slopes toward the lowest dark reproducible on a given monitor? Can I play with it? Where would you start (What settings other than default? … What symptoms should I look for to determine which direction to adjust? …I want perfection!!! LOl,,,
Cool…So what about going in the opposite direction? ie, take a profile which already has ciecam02 applied and create a profile from measurement data without using. (Bringing it back to ‘normal’)
Not needed. The CIECAM02 mapping is only available via the “perceptual” and “saturation” rendering intents and respective tables within the profile. The colorimetric table is unaffected.
Also, While doing a calibration, a random window popped up behind the measurement swatch…Does this sort of thing have any effect on the readings?
As long as it’s really behind the measurement window, it shouldn’t, unless the display reacts to displayed content in some way (e.g. auto-dimming, ABL…).
And finally, this ‘black point’ adjustment thing. What exactly does it mean?
I’m assuming you’re talking about black point correction on the “Calibration” tab. The percentage chosen determines how much the calibration tries to make black the same hue as white (if at all, default is 0% = off because hue changes near black are less perceptible and maintaining contrast is usually more desirable). The “rate” setting determines the blend to black. The default setting of 4 blends to the native black hue relatively close to the black point. See the documentation on black point correction for more info.
Can I play with it? Where would you start
Personally I’d leave it at default.
Thank you!
My monitor’s native gamma is about 2.0 (HDTV) … Do you think it would be better to calibrate to 2.0 to reduce banding and increase quality? … Does this mean that I won’t be viewing the sRGB 2.2 standard? (Wont match the original?)
And along these same lines, would it be better to use the tv’s default white point (7000K something) and allow the calibration to adjust it to 6500K, or should I use the RGB (high) sliders? … The sliders work ok, but I do get some banding especially in the greens.
Finally, what is considered the native setting on an HDTV if there are choices to choose from? ie, I can choose presets … They can’t all be native, right? 6500, 7000, etc…
My monitor’s native gamma is about 2.0 (HDTV) … Do you think it would be better to calibrate to 2.0 to reduce banding and increase quality?
Probably not. Moderately modern AMD and nVidia cards have >= 10 bit videoLUTs, so videoLUT-induced banding may not even be an issue.
And along these same lines, would it be better to use the tv’s default white point (7000K something) and allow the calibration to adjust it to 6500K, or should I use the RGB (high) sliders?
You’ll have to figure this out by experimentation because differnt displays may react differently.
Finally, what is considered the native setting on an HDTV if there are choices to choose from? ie, I can choose presets … They can’t all be native, right? 6500, 7000, etc…
The one with the highest brightness (when only adjusting the TV’s whitepoint) is probably native, but that doesn’t mean it’s the best mode to use.
