- This topic has 10 replies, 3 voices, and was last updated 6 years, 5 months ago by
.
-
Posts
-
Hello there,
I am having real doubts about the i1Display Pro measuring the white point correctly on my wide gamut display LG 31MU97 .
I calibrated and profiled to D65 Gamma 2.2. After leaving the room and coming back, I thought that the white point looked more like 5500. So I got my camera and checked the white point with white balance. It showed exactly what I thought: 5500 not 6500K as it was supposed to be.
Then I switched to the Adobe RGB emulation on my display which is supposed to be precalibrated to D65 as well, and boom, it looked a lot more like 6500 to me. Again, I got my camera and this time it showed 6500K. Then I tested my LEDs with the camera white balance. They are supposed to around 5600K which indeed I measured with my camera.
According to http://www.prad.de the LG 31MU97 has GB-r-LEDs, so I used the “Spectral LCD RG Phosphor” as correction when I calibrated it. But also all the other corrections showed me the wrong color temperature. I really wonder why. I also did a Hardware calibration with True Color Pro from LG, setting D65, but the result was the same. So it must be the colorimeter. I read about the problems with Colorimeters on wide gamut displays, but I thought that the i1Display Pro was capable of wide gamut Displays.
Also I tried one more thing: I switched the i1Display Pro to surround measuring and I got the same results for the LED lamps as I got with my camera white balance. Then I opened up a white blank page on the display, pointed the i1Display Pro towards the display and did a surround measuring. It was wrong again. Really really confusing! The I calibrated my Laptop to D65 and checked with the camera white balance and it showed 6500K. So it just looks like the i1Display Pro isn’t suitable for wide gamut displays.
Did anyone else experience such a behaviour, or what do you think?
Calibrite Display Pro HL on Amazon
Disclosure: As an Amazon Associate I earn from qualifying purchases.Hi,
a camera, even a high quality one, is not a color measurement instrument, and completely unsuitable to check color performance. Also, comparing correlated color temperature on its own is an unsuitable way to check white balance, because it does only take into account the deviation in the red/blue direction (meaning a very green or very purple whitepoint can have the exact same CCT, but be far from the neutral white of the corresponding daylight or blackbody locus).
Hello Florian,
thanks for your reply. I assumed that my camera white balance won’t be accurate, but don’t you think that it’s odd, that I get almost identical white point measures with the camera and the i1Display Pro expect for the display? I even tested it with my calibrated Laptop display: Both, the colorimeter and the camera measured around 6500K which was supposed to be the color. When I do the same with my calibrated wide gamut display, the color temoerature varies around 1000K. So wouldn’t you agree that the i1Display Pro seems to have problems measuring my wide gamut dislay?
Color temperature is a useless measure unless referenced against the daylight or blackbody locus. To assess white, you need to measure the XYZ coordinates.
What do you mean by this? I am very sorry if this a stupid question, but how can I reference it to daylight? Maybe you can send me a good source that explains it.
A calibration software will do it for you. That’s what the daylight or blackbody choice on the calibration tab does (you can see it if you enable advanced options, the default is daylight reference), without such a reference, a color temperature target for whitepoint calibration would be pretty much useless.
OK, but I guess my camera is also referenced to daylight, otherwise it wouldn’t make any sense. Besides, I accept the fact, that the white balance in my camera is not exact. But still, it measures light more than a spectrometer than a colorimeter, right?
I just find it very confusing, that both, my camera and the i1Display Pro measure almost the exact temperature on everything like my LEDs, my calibrated Laptop, light coming from outside. But when it comes to my wide gamut display, the discrepancy between camera and colorimeter is almost 1000K. Doesn’t this imply that the colorimeter just isn’t capable of measuring the wide gamut display properly?
I mean, isn’t that what Graeme Gill talks about here ? I quote:
Why don’t Colorimeters work so well on Wide Gamut displays ?
As explained above, due to the imperfect match between the Colorimeter filters and CIE Standard Observer weighting curves, Colorimeters have calibration matrices that are created for “typical” CRT or LCD displays. A Wide Gamut display by its very nature has primaries that have narrower spectral characteristics than typical displays, and this spectral difference exacerbates the approximations and errors in the Colorimeter filters.Since Spectrometers have mathematically computed weighting curves, they are less sensitive to the spectral characteristics of the display primary colors, and generally work better on Wide Gamut displays.
The X-Rite i1 DisplayPro, Pantone ColorMunki Display and possibly the Spyder 4 may have such an improvement.
Well, to my experience at least the i1Display Pro and the ColorMunki Display may not have such an improvement.
But still, it measures light more than a spectrometer than a colorimeter, right?
No. A digital (photo) camera uses typically a Bayer pattern (filter) in front of monochromatic sensor (CCD). Some video cameras use three-chip CCDs with a light-splitting prisma. For all intents and purposes, cameras are closer to colorimeters than spectrometers.
the discrepancy between camera and colorimeter is almost 1000K
Coreelated color temperature is a less useful indicator (it lacks meaning as outlined above). I.e., you can’t use a camera to check how close the measured temperature is to the daylight locus. You can do that (within sensor and filter limits as well as used color matching functions) with a colorimeter.
Well, to my experience at least the i1Display Pro and the ColorMunki Display may not have such an improvement.
I see no indication of that. I wouldn’t trust a consumer camera (“ballpark”) over a colorimeter like the i1 DisplayPro.
Thank you very much for your reply!
Still, I got the Colormunki Photo (spectrometer) now and I have to sas that I am much happier with the results for my wide ganut display, although it‘s a little bit slower. „Adaptive high res“ is the correct setting, right?
Anyway, I‘m happy to make a little donation for your great work, Florian!
Still, I got the Colormunki Photo (spectrometer) now and I have to sas that I am much happier with the results for my wide ganut display, although it‘s a little bit slower.
You can have the best of both worlds by creating a correction matrix for the i1 DisplayPro using the ColorMunki Photo.
„Adaptive high res“ is the correct setting, right?
Adaptive (highres or not) is the recommended setting for best low light performance.
Anyway, I‘m happy to make a little donation for your great work, Florian!
Donations are always welcome 🙂
Some side notes:
-True Color Pro is unable to apply RG_phoshor correction. Shame to LG. It cannot be taken as reference for your results.
-RG_phoshor is a GB-LED spectral correction plus other types of non GBLED backlight. You can try to “clean” this CCSS (keep last 4 lines) although in a common well behaved i1d3 should not make much diffreence. There are also several GB-LED samples in DisplayCAL’s colorimeter database… and they are more or less the same as that RG one from xrite once cleaned.
-Your particular i1d3’s colorimeter “observer” is stored in its firmware and you can dump it for inspection with ArgyllCMS. There is a flag(#define) in its sources for that task. Also you can ask Mr. Gill for a dedicated executable for that task in future ArgyllCMS versions (or write one if you know how to do that).
When you have your colorimeter observer dumped in a cmf file then you can argue how bad it is and the kind of correction which is applied.
Another and easier way is to run a command line basic uncalibrated report with and without correction on a high level trace to calculate by yourself an “equivalent” XYZ to XYZ matrix correction to the CCSS correction applied (that will be portable to CalMAN too). Huge deviations from identity matrix (more than 0.1) may point a not so well behaved observer with that kind of backlight (from colorimeter’s observer factory data, which I assume to be as true -or even better- as the accuracy of the diffraction grating + sensor position from a low cost low end non-NIST traceable spectrophotometer).-Since you bought that low end low cost spectro, you could try go make a custom CCSS for that LG backlight, but it’s very very likely that it would look the same as the ones you had available for free before. As Florian pointed, since you have that two devices is a nonsense to not use the high speed of your i1d3, use both of them.
If you go to CCMX matrix branch instead of CCSS, that 10nm spectros from Xrite are likely to get a litle more warm white point than the actual one for a GB-LED backlight (Czornyj in LuLa and other sources)… but Munki’s inter instrument agreement play a huge role here, so who knows.
