- This topic has 10 replies, 3 voices, and was last updated 5 years, 2 months ago by
.
-
Posts
-
Hello!
So this monitor can save calibration to hardware directly which is one of the reasons I got it.
I am assuming there is no way to do this when using DisplayCal so my only option is their own program?Thanks
A better calibration options is to use an external LUT box. I believe, this would bypass any limitation of the internal calibration and file created by the native software: 1 D LUT, ICC vs 3D LUT external installation on the LUT box.
However, I have no experience with BENQ display, but it’s well known that the majority of proprietary calibration softwares are short in features and flexibility and compared to third party calibration apps.
Well the internal software is indeed driving me crazy. Like if I calibrate the monitor, I would expect that when setting up to run DisplayCal when you are supposed to adjust RGB , it will be very close to perfect if not perfect already yet is not even close. Also the program has its own validation which tells me is great after I run it but later on if I retry the validation the TEmp is reading like 2000k or so which I dont see as real at all so must be one weird bug.
Anyways, I am down to using xrite i1 software instead and it seems to be giving me good results for now. But for that I may as well keep the PD3200 if I am not using the hardware calibration.
Maybe is time to save for a Eizo lolWhen display calibration doesn’t match, which is the case with displays using different technology the only solution is perceptual calibrations.
There is no CalMAN, Light Illusion nor DisplaCAL that is going to fix maramerism by using just numbers.
Have you tried perceptual calibration?
I am assuming there is no way to do this when using DisplayCal so my only option is their own program?
Thanks
AFAIK there is no 3rd party for SW320 and there is no public SDK (like the ones HP or Dell have) for the possibility of a 3rd partey GNU app.
Well the internal software is indeed driving me crazy. Like if I calibrate the monitor, I would expect that when setting up to run DisplayCal when you are supposed to adjust RGB , it will be very close to perfect if not perfect already yet is not even close. Also the program has its own validation which tells me is great after I run it but later on if I retry the validation the TEmp is reading like 2000k or so which I dont see as real at all so must be one weird bug.
Benq software, like other similar brands have 2 sources of error regarding calibration results with i1d3 devices:
1- Measurement errors because wrong spectral corrections. Benq never cared about this in SW-series (other brands do). They use a RGBLED correction that is not even close to a GB-LED, QLED (SW2700PT) or WLED PFS phosphor (SW240 supposed to use it)
2- its own internal simplifications during calculations
2000K seems to high, even for Benq’s faulty software. Check which correction you used for validation. Since Benq software uses RGBLED, try to use it for validation even you know it’s wrong. Just to test if Benq Software made other mistakes play his game with the same rules.
IDNK which backlight type uses SW320 (not RGBLED for sure) but GB-LED (RGphosphor correction) would be my 1st guess if you want to do a 1DLUT calibration in graphics card with DisplayCAL… or to correct using GPU the results of Palette Master Elements calibration in white point, grey or gamma.
QLED seems limited to their AUO panels in 27″ screens.Also you should know that there seems to be a bug/discrepancy when validating PROFILES made with Palette Master Elements (ICC v2) and DisplayCAL. DisplayCAL reads as actual display profile whitepoint “D50” (and maybe DisplayCAL is right and profile stores D50 without CHAD tag) and measures a typical calibration target of D65 (with more or less deviation form target). DisplayCAL shows that it is “white” (close to daylight curve) but reports a huge diference between profile WP and measured WP.
Is that issue what you report? If it is that, don’t care about it (measured white vs profile white) as long as you aim for a daylight white and DisplayCAL measurement shows white close to that target.Maybe is time to save for a Eizo lol
IMHO NEC/Eizo is the sensible choice for AdobeRGB-like widegamut displays. A working calibration solution is good but color uniformity is the real reason for choosing them.
When display calibration doesn’t match, which is the case with displays using different technology the only solution is perceptual calibrations.
There is no CalMAN, Light Illusion nor DisplaCAL that is going to fix maramerism by using just numbers.
Have you tried perceptual calibration?
Although you are right and a perceptual approach will do the job, maybe the reason for a relatively huge mismatch in LGabrielPhoto is not a metameric failure (which is possible too!)… but just a wrong configuration because wrong spectral corrections for his colorimeter.
For example a well behaved i1d3 colorimeter should not vary too much in dE numbers while measuring white in a GB-LED/QLED display if you use a custom GB-LED CCSS, custom QLED, RGphosphor or even RGBLED, but such small dE difference (<3) could be placed in a* axis (pink -green) giving that display a distinctive but slight tint towards those opposite colors. That issue won’t be a observer metameric failure, just a wrong configuration.
Calibrite Display Pro HL on Amazon
Disclosure: As an Amazon Associate I earn from qualifying purchases.Although you are right and a perceptual approach will do the job, maybe the reason for a relatively huge mismatch in LGabrielPhoto is not a metameric failure (which is possible too!)… but just a wrong configuration because wrong spectral corrections for his colorimeter.
On this case, probably the workaround for the solution of the mismatching between displays would be, one, to recalibrate the colorimeter, second, to create a profile with a spectrometer for the colorimeter and the target display. Of course, this requires an additional cost, minimum investing on the i1Pro 2.
Hello,
Just a small update.
I will be returning the SW320. I tried once again the Hardware calibration and in a greyscale test image in photoshop I see this faint vertical bands of magentaish color cast. If I do a software calibration with i1 Profiler then it looks fine. But at that point, I may as well just keep my PD3200u that I was able to calibrate and get between ..45 to .85 edelta according to xrite i1 Profiler so slightly better than the SW320.
My main thing was thinking hardware calibration was going to be better but nothing like that so far.
Also screen uniformity is actually worst on the SW320..maybe I just got a lemon with the red dead pixel and all.Thanks!
Hello,
Just a small update.
I will be returning the SW320. I tried once again the Hardware calibration and in a greyscale test image in photoshop I see this faint vertical bands of magentaish color cast.That could be caused by
-small number of patches by PME if your monitor has some issues in grey uncalibrated (HW should be capable, but PME softwaqre ruins it) <—- Main source of errors, usually
-or some profile issues (smooth color cast with equal TRC profile => wrong PME calibration… but “solid band” color cast => could be caused by 3xTRC profiles)
-or some non linear calibration loaded in GPU with some HW or with some GPU LUT loaders. Your next paragraph seems to discard this issue.
If I do a software calibration with i1 Profiler then it looks fine. But at that point, I may as well just keep my PD3200u that I was able to calibrate and get between ..45 to .85 edelta according to xrite i1 Profiler so slightly better than the SW320.
If you suffer grey issues you want to validate with DisplayCAL and take a look on a*b* deviations vs neutrality ( a*=b*=0) in grey scale on measurement (right side of report). “range a* b* ” in DisplayCAL report also show this to some extent.
Raw dE00 values are not going to show those issues unless the numbers were “big”. You could validate a profile under 1<dE00 and green magenta cast in grey could be spotted visually.
My main thing was thinking hardware calibration was going to be better but nothing like that so far.
Also screen uniformity is actually worst on the SW320..maybe I just got a lemon with the red dead pixel and all.Thanks!
AFIAK the problem should not be “HW LUTs” inside monitor. It is very likely thay it uses at least 3 x 1024 x 12 or 14 bit LUTs… and such LUT should provide smooth grey ramps.
The problem is Palette Master Elements: too few measurements of uncalibrated R,G, B and grey ramp in a monitor with mild to severe grey ramp tint in its uncalibrated state. It suffers other equaly bad issues, your the issue you repot is very likely caused by that.
Since i1Profiler takes more measurements of that uncalibated grey ramp (DisplayCAL is able to take much more!) … it is able to measure were those tints are located and correct it.If you want a “fast”, small set of measurements in a HW calibration solution then vendor should ensure that their displays have a VERY neutral grey out of the box uncalibrated. Otherwise such small calibration patch set is going to leave those magenta cast on not so neutral displays.
(Calibration patches are not the same as profiling patch set that you can configure in i1Profiler or other HW CAL tools derived from it: calibration patch set is FIXED/LOCKED but profiling patch set is configurable to some extent in most tools)Awesome explanation even if I have to google part of it!
I think you are right that the flaw here is the Palette Master software not able to take advantage of the hardware I paid for.
So for now, I will stick to my PD3200u which is looking good to me and keep an eye to test an Eizo eventually to see what I am missing. 🙂Awesome explanation even if I have to google part of it!
I think you are right that the flaw here is the Palette Master software not able to take advantage of the hardware I paid for.
So for now, I will stick to my PD3200u which is looking good to me and keep an eye to test an Eizo eventually to see what I am missing.
Before returning it, validate HW calibration profile with DisplayCAL. Take a look in grey ramp measurements and “range” to spot those magenta tints. They should be there.
You can do a GPU calibration (like in non HW calibration monitors) over the calibration made by PME. Do it, really, do it (DisplayCAL better than i1Profiler).
GPU calibration only can fix grey neutrality issues, gamma and white point… but that is your issue. GPU calibration may cause banding on some graphics cards or with some LUT loaders but since you do not see it with i1Profiler then it should not apply here.If you just made a HW calibration with PME, then reboot computer to avoid the issues I described before on other vendors. Then run a calibration in DisplayCAL, “single curve+ matrix profile” and try again your Photoshop test. Those magenta bands should be gone.. and if they don’t then the culprit is graphics card or some calibration loader conflict and you are going to suffer the same with a PD3200U (if it has the same issues in uncalibrated grey)
I mean… this is free to try and it solves your problem (IF you had a graphics crad that to not caused banding because of truncation)
