Trying to calibrate ASUS VG27AQL1A with Spyder4

[Orangelynx]

Hi,

first of all, thanks to all developers and contributors of Displaycal, it’s great to be able to get such professional software open source and free. I’ll happily donate once I got a satisfying result.

I got a ASUS VG27AQL1A which, according to Rtings, has poor out of the box color accuracy but can be calibrated really well. The monitor has sRGB mode, in which basically all controls (even brightness) are locked. This mode looks okay-ish but is too bright in the evening. Hence I’d like to calibrate one of the more user-configurable modes to get the best of both worlds. In the rtings article, they use Racing mode for that, but this has a ASCR active, which I believe is not good (please correct me if I’m wrong). Hence I intend to use the Scenery mode, which out-of-the-box looks horribly over-saturated.

I have a old Spyder4Express kit, that came with my previous monitor, a factory calibrated ASUS PA248QA. I tried the stock software and it produced complete garbage, hence it started to read into the topic and use DisplayCal. I did about 10 calibration runs now, and they all turn out quite differently, and I’m sure I’m not doing it right OR the colorimeter doesn’t deliver good readings. I’m hoping you can help me figure out, which is the case. I also have the option to rent a i1 Display Pro for a day for about 20€, but I want to make sure my Spyder4 is not up to the task before that.

I hope to create three calibrations / profiles. 1) For the monitors sRGB mode, reducing the deltaE of that mode, 2) a correction for the scenery mode with a high brightness (daytime) and one 3) for low brightness.

Here’s what I did:

1. install DisplayCal + ArgyIICMS + Drivers, installed colorimeter corrections from installer of spyder4 software, set to white LED.
2. Used these settings (common for all 3 cases): Full Range Output, No Correction, White / Black Level As Measured, sRGB Tone Curve, High Speed, 0% Black Point Correction, XYZ LUT + matrix profile (high quality), CIE 1931 2° Obsever.
3. For sRGB i disabled interactive calibration, as I can’t do anything anyway and tried a fixed white point and “as measured”. I used a black point rate of 3.0, as previous test showed significant brightning of dark colors.
4. For Scenery mode I used interactive calibration with measured whitepoint and 3.5 / 3.8 black point correction rate.

The results produced always a (to my eye) slightly tinted color, but not really consistency. I also noticed a severe impact of ambient lighting on the result (leakage?). The verification log also shows rather poor delta E for dark patches and checking on LCD monitor test images, the results don’t really look satisfying.

So I’m not really sure what I can do to improve the results. Any tips on what I could try would be appreciated. Also I’m wondering:

1. Does sRGB mode usually limit the gamut to the sRGB space?
2. Should I calibrate in a completely dark room or have some ambient light?

Thanks,
Max

Attachments:

You must be logged in to view attached files.

Calibrite Display Pro HL on Amazon  
Disclosure: As an Amazon Associate I earn from qualifying purchases.

[Vincent]

1) If it is a widegamut, yes.
Instead of using scenery mode just by name, choose an OSD mode that allows you to access RGB gains for whitepoint and RGB/RGBCMY saturation controls to try to reduce gamut to sRGB, so you can use it as a common sRGB monitor (games and non color managed apps). Check manual, check which controls are available in each OSD mode.
DisplayCAL’s cousin, “HCFR” can help you with that task. Once you get sRGB-like colorspace run DisplayCAL calibration as usual and if all goes as expected you’ll end with a display with sRGB colorspace (and only that), D65 white and 2.2. gamma as a typical configuration sample (you may choose other).
If you do not care about non color managed apps, or want to calibrate an additional different preset with native gamut and its own ICC, skip that HCFR part and do not modify saturation controls in that OSD mode.

Also you need a colorimeter correction, but IDNK which backlight is used in that display. If you know that it is some P3 gaming display, the only one available with Datacolor corrections that will look “remotely close” is RGB LED (if it is a P3 gamer display it cannot be “White LED” in datacolor correction naming, it needs to be something else). If once numerically matched to D65 white looks off, from that state try to use visual whitepoint editor till you get a white that looks white to you.
Same applies to i1DisplayPro, it’s a shame rtings do not provide a spectral power distribution sample in most of their reviews. It will greatly simplify things for users without spectrophotometer.

Very dark patches error can be caused by Spyder, so ignore them or replace it with an i1d3. If you keep Spyder then using single curve + matrix profiles can solve some issues in color managed apps “IF” those near black greys have visually no color issues.

[Orangelynx]

Thanks Vincent!

sorry, I didn’t make that very clear, but scenery mode is in fact the mode which allows to disable all dynamic image improving wizardry and has access to RGB controls for color temperature and saturation / contrast / brightness controls available. The saturation slider is however not RGB, just scalar. I will check out HCFR! I wouldn’t have guessed that it’s even possible to reduce the gamut to sRGB using those controls.

If you do not care about non color managed apps, or want to calibrate an additional different preset with native gamut and its own ICC, skip that HCFR part and do not modify saturation controls in that OSD mode.

This startled me a bit. So are you saying that non-color managed apps will always look weird on a wide-gamut display because they are made for sRGB? I guess that makes sense, but is there some middle way to get the both of both worlds, the wide-gamut and color accuracy for non-color-managed apps?

Also you need a colorimeter correction, but IDNK which backlight is used in that display. If you know that it is some P3 gaming display, the only one available with Datacolor corrections that will look “remotely close” is RGB LED (if it is a P3 gamer display it cannot be “White LED” in datacolor correction naming, it needs to be something else). If once numerically matched to D65 white looks off, from that state try to use visual whitepoint editor till you get a white that looks white to you.

Thanks, also very interesting! I just read on the spec-sheet that the monitor has a white-LED backlight, so I assumed that correction was the obvious choice. I will experiment as you suggested.

[Vincent]

Thanks Vincent!

sorry, I didn’t make that very clear, but scenery mode is in fact the mode which allows to disable all dynamic image improving wizardry and has access to RGB controls for color temperature and saturation / contrast / brightness controls available. The saturation slider is however not RGB, just scalar. I will check out HCFR! I wouldn’t have guessed that it’s even possible to reduce the gamut to sRGB using those controls.

If you do not care about non color managed apps, or want to calibrate an additional different preset with native gamut and its own ICC, skip that HCFR part and do not modify saturation controls in that OSD mode.

This startled me a bit. So are you saying that non-color managed apps will always look weird on a wide-gamut display because they are made for sRGB?

Yes, although actually not “made for sRGB” but “for whatever content colospace was encoded for”.  It’s sRGB 99.9%  of time

I guess that makes sense, but is there some middle way to get the both of both worlds, the wide-gamut and color accuracy for non-color-managed apps?

Using different OSD presets for each type of application.

Also you need a colorimeter correction, but IDNK which backlight is used in that display. If you know that it is some P3 gaming display, the only one available with Datacolor corrections that will look “remotely close” is RGB LED (if it is a P3 gamer display it cannot be “White LED” in datacolor correction naming, it needs to be something else). If once numerically matched to D65 white looks off, from that state try to use visual whitepoint editor till you get a white that looks white to you.

Thanks, also very interesting! I just read on the spec-sheet that the monitor has a white-LED backlight, so I assumed that correction was the obvious choice. I will experiment as you suggested.

If it is P3 it cannot be “White LED” in datacolor (or Xrite) correction naming it needs to be something else because that means Blue led yellow phosphor sRGB like display. It can be a While led with additional phosphors like a W-LED PFS phosphor, but this is not a “white led” in Datacolor or Xrite naming.

• This reply was modified 3 years, 3 months ago by Vincent.

[Orangelynx]

Just to avoid missunderstandings, by P3 you mean a display made for DCI-P3 space? I guess that’s the case. Btw, the panel is a innolux M270KCJ-K7B.

[Vincent]

Just to avoid missunderstandings, by P3 you mean a display made for DCI-P3 space? I guess that’s the case.

Yes

[Author]

Posts