- This topic has 10 replies, 4 voices, and was last updated 3 years, 6 months ago by
.
-
Posts
-
Greetings everyone!,
I have a VA QDEF(Quantum Dot Enhancement Film) 4K display with 97% DCI-P3, HDR 1000 and a i1 DisplayPro for calibration. I am NOT targeting color accuracy for sRGB content but I am concerned with:
- Gray accuracy/balance very important DeltaE < 1
- Gamma curve accuracy (whether i choose 2.2/2.4/2.6)
- Best contrast ratio and dynamic range.
- red/green/blue primary accuracy INSIDE the dci-p3 space, so if red is shifted to magenta inside p3 i want to correct that (want to correct native gamut which is close to P3 to P3 target). (DeltaE<1.5)
- I want sRGB content to be extended to dci-p3 so that an sRGB red becomes a DCI-P3 red. But i want accuracy in the sense that if I am watching extended gamut content that content should be at least DeltaE<1.5 accurate.
These are NOT in HDR mode, I will think about that later, for now just standard sdr calibration.
I don’t see any settings in displayCAL to set the target gamut, does anyone know which settings i must use, if what I’m trying to do is even possible.thanks
Calibrite Display Pro HL on Amazon
Disclosure: As an Amazon Associate I earn from qualifying purchases.Displaycal makes a map of how the display responds.
A seperate Color_managed application (photoshop/premiere) given user input to use either p3/rec2020/rec709 maps its output TO the map that displaycal makes.
You’ve misunderstood the process. The display always does what the display does. It’s up to displaycal to Map what the display does, and subsequent software to use that map.
On the display end, your job is to figure out which MODE is the behavior most stable, for example, you have to disable dynamic contrast, or any post processing that the display does on its own, because displaycal Can not map the display if those behaviors are in play.
but why can’t there be a mode that extends lower gamuts.
With standard windows ICC:
chrome sends a 255 red thinking it is an sRGB red, display displays that red but oops colorimeter and my eyes see a wide gamut red.
With calibrated ICC profile that displayCAL creates:
chrome reads a 255 red in content thinking it is an sRGB red, but chrome uses icc profile and ends up sending 234(just an example) to display, display displays a red that the colorimeter and my eyes see as an sRGB red. accuracy! but boring sRGB red.
what I want:
chrome reads a 255 red in content thinking it is an sRGB red, then chrome reads the icc profile and adjusts it a tiny bit so it hits the red P3 primary target, ends up sending a value to the display that is close to 255 (so still well saturated) but has a lower DeltaE than native display P3 gamut. accuracy! vivid red! (not faithful to creator’s intent)- I want sRGB content to be extended to dci-p3 so that an sRGB red becomes a DCI-P3 red. But i want accuracy in the sense that if I am watching extended gamut content that content should be at least DeltaE<1.5 accurate.
Nonsense… you do not even know what accuracy is.
If you want to see things *wrong* so desperately, use DisplayCAL for grey calibration and IE explorer/VLC/Windows Photo app
what I want:
chrome reads a 255 red in content thinking it is an sRGB red, then chrome reads the icc profile and adjusts it a tiny bit so it hits the red P3 primary target, ends up sending a value to the display that is close to 255 (so still well saturated) but has a lower DeltaE than native display P3 gamut. accuracy! vivid red!That’s the same as no calibration, the display already does that with no intervention.
As Vincent said, you completely misunderstand the point of using displaycal.
what I want:
chrome reads a 255 red in content thinking it is an sRGB red, then chrome reads the icc profile and adjusts it a tiny bit so it hits the red P3 primary target, ends up sending a value to the display that is close to 255 (so still well saturated) but has a lower DeltaE than native display P3 gamut. accuracy! vivid red!That’s the same as no calibration, the display already does that with no intervention.
As Vincent said, you completely misunderstand the point of using displaycal.
The display does that but it does it wrongly, an srgb red becomes a magenta shifted red instead of a P3 red…..
I know it’s not accurate in the sense that the color displayed isn’t the color intended but there is an argument to make that there’s more and less accurate ways to extend sRGB content…
But as i see there is no way to do this and the only option is to go for sRGB being accurate I’ve given up on extending and am just calibrating normally. I tried a few colorimeter corrections but there’s no explicit QDEF spectral correction, but i did find a matrix correction for my display on the displaycal ccmx online rep, funnily enough using that makes everything shift green and look horrible compared with a standard LCD WHITE LED correction and compared with my calibrated TN panel that i have right on the side…
So now i’m just trying on corrections like pants and see what looks best perceptually since i can’t trust a colorimeter if the correction isn’t accurate… funny how im going back to my eyes after buying a supposedly objective measuring device. Here we are 2020 and color is still a mess, Windows needs os wide color management yesterday.what I want:
chrome reads a 255 red in content thinking it is an sRGB red, then chrome reads the icc profile and adjusts it a tiny bit so it hits the red P3 primary target, ends up sending a value to the display that is close to 255 (so still well saturated) but has a lower DeltaE than native display P3 gamut. accuracy! vivid red!That’s the same as no calibration, the display already does that with no intervention.
As Vincent said, you completely misunderstand the point of using displaycal.
The display does that but it does it wrongly, an srgb red becomes a magenta shifted red instead of a P3 red…..
I know it’s not accurate in the sense that the color displayed isn’t the color intended but there is an argument to make that there’s more and less accurate ways to extend sRGB content…
But as i see there is no way to do this and the only option is to go for sRGB being accurate I’ve given up on extending and am just calibrating normally. I tried a few colorimeter corrections but there’s no explicit QDEF spectral correction, but i did find a matrix correction for my display on the displaycal ccmx online rep, funnily enough using that makes everything shift green and look horrible compared with a standard LCD WHITE LED correction and compared with my calibrated TN panel that i have right on the side…
So now i’m just trying on corrections like pants and see what looks best perceptually since i can’t trust a colorimeter if the correction isn’t accurate… funny how im going back to my eyes after buying a supposedly objective measuring device. Here we are 2020 and color is still a mess, Windows needs os wide color management yesterday.Then you won’t be able to get what you ask…. all this threads seems to be a nonsense.
If windows had color mangement at desktop level converting untagged images/colors to sRGB… you won’t have what you ask. Untagged 255 red will be 255 sRGB, not P3.
My advice is that:
-you learn what is color management first… because it seems that you do not know, otherwise your text will be different
-then try Samsung QD bundled CCSS in DisplayCAL for your i1d3 (plot it, 3 peaks relatively broad compared with other P3 widegamut backlights, QD typical). If white does not look white use “visual whitepoint” approach on top of that CCSS correction… or buy/borrow/rent an i1Pro2 and make one
-after you calibrate white & grey with Displaycal you’ll get 1D grey calibration and an ICC profile storing that 1D calibration.
-if you want to see accurate colors you’ll need color managed apps and in most of them untagged images are sRGB/Rec709, so untagged 255 red will be 255 sRGB red, not native255
-if you want to see all things wrong, like you ask, just use non color managed apps or disable color management in those color managed apps. 255 red in whatever colorspace will be rendered to native 255 red.@ Nighteyes,
Put more simply, P3 and Srgb are standards.
They’re only Correctly displayed if the content was MADE FOR and under that standard.
Taking something made in Srgb (which is the majority of content), whether you stretch it to P3 or your Native display gamut (which isn’t P3, it’s just close to p3), neither is more correct than the other, they’re both VERY WRONG.
But if YOU like the way that looks, that’s fine. It’s your machine.
To achieve your description, as we’ve said, you don’t really even need displaycal, because it will pretty much look like that out of the box.
Thanks for calling me out on my bs guys
Yeah I understand that part, whichever way one stretches sRGB it will always be veeery wrong. It’s just that for non-creatives and non-professionals, people who are just enjoying their wide gamut displays, some might want stretching to wide gamut in a matter that is more natural than the native stretching the display does.For anyone else that is trying to achieve the same thing i advise calibrating normally and then you can use the nvidia control panel to stretch it to your native gamut, and then if u need accuracy again just turn turn it back to application-controlled.
I don’t know if every display i calibrate becoming subjectively green-tinted is my eyes being used to blue displays or if my i1Display’s filters are kicking the bucket. (it’s supposedly only a few years old). I have done 4 displays in my house, 1 VA, 1 IPS, 2 TN all become a bit greener with the TN’s seeming the least green.
I personally don’t believe it is necessarily a terrible idea to push the sRGB under certain conditions. I have a projector in my home theatre that does exactly this. This is a non-HDR projector (ca. 2009 or so) which is designed and calibrated for HDTV/Rec.709. It is dead-on accurate to Rec.709 up to 75% or so saturation in any color, but for more highly saturated colors, it pushes them out toward its native primaries. The result is a very accurate looking display, including skin tones and other perceptually important colors (for lack of a better term). Only highly saturated colors have more pop than they should, and this really does add something. It looks amazingly good, but it is not technically accurate. The result is much different than simply increasing the saturation or directly using DCI-P3 RGB values on an sRGB (or Rec.709) display. The key is the non-linear saturation push which only kicks in for the most saturated portion of the source gamut. A person far more clever than me could probably use DisplayCAL and/or ArgyllCMS to create a 3D LUT which achieves the same thing an any other wide-gamut display. This can’t be done with simple 1D LUT (gamma curves) because of the non-linear behavior. Only applications which can use a 3D LUT could then make use of that. An example is MadVR (for video playback).
A person far more clever than me could probably use DisplayCAL and/or ArgyllCMS to create a 3D LUT which achieves the same thing an any other wide-gamut display. This can’t be done with simple 1D LUT (gamma curves) because of the non-linear behavior. Only applications which can use a 3D LUT could then make use of that. An example is MadVR (for video playback).
No rocket science, we do not have to code it, Mr. Gill and Mr. Höch did it for us.
Just profile without calibration your projector, a big XYZLUT profile with 400, maybe more depending on how slow is your mesurement device. Then LUT3D maker, Rec709 source g2.4 in, your newly created projector profile as target, make it.I mean it’s easy, a very guided & automatized process… although if your extra saturation at 100% is not severe and you own a slow device it can be very time consuming so it’s reasonable to skip it. If you use madVR and can use an i1DisplayPro I’ll go for it.
