The display itself is a very nice display overall. I used to have an U2711h before.
The UP2715K has a higher real contrast (Measured as 1:870 when hw-calibrated to AdobeRGB) and due to the "optical bonded" (glued on) antireflective front panel a much higher visual contrast. Also, there is no pixel distortion effect at all, which was extremely annoying with the U2711 due to heavy anti-glare coating.
Colorwise it is all okay, but no calibrated native color gamut yet. Meaning, I currently lost a bit of color space when compared to the U2711 setup.
Overall, the colors "feel" smoother than with U2711. Visually the picture is much better due to contrast
There is one tiny thing with the whitepoint. The whitepoint varies from -190 K to +190 K when comparing left and right side of the panel to the measurement in the middle. For the Dell Calibration Solution this is still green because its "all-good value" is everything below 250 K. But the effect is clearly visible.
I know that the higher Nec or Eizo have a compensation (DUE) for this kind of thing, but the UP2715K has not. Not even the static version some Dells have.
I tried a second UP2715K which had the same (visually, not measured) effect. Question is, if this is a panel or electronics issue.
I've got this display several days ago and was a bit dasappointed. I was expecting it to work fine out of the box. but it's a lettle yellowish in all modes. Nice white (for my eyes) is only in 9300k and 10000k modes. Both precalibrated profiles look the same and not quiet good. I'll get an xrite in several days, maybe it'll get better. Was your display good calibrated out of the box?
Yep true whit different to everyon, but it just looks too yellow for me.
I've got another question. I've read in one of the discussions, that someone got different calibration results on every half of the display due to two controllers after hardware calibration. Results are on this image
The out of box calibration looked okay to me visually. I did not measure it.
The HW-calibration is a bit different. One thing is that the HW-calibration results in 6650K and not 6500K as it is supposed to.
Frankly, most people that have the same experience like you are usually first timers with displays calibrated to 6500K which is "the usual" whitepoint. They are just used to colder, bluer white points like 9000K or 10000K and need some time to adapt.
You will see which is it when you get the xrite.
Strangely enough there is not "the one, right white". There are a lot of them and all in different colors :)
Yes, white point is a matter of taste, but I still feel it's to yellow for me.
Onother important question. I've read in discussion, that someone have got different calibration results on every half of the monitor after hw-calibration due to two controllers.
It happened to me one time after a first start of monitor and PC that the calibration was obviously not loaded into one of the controllers. So, one side of the display had a clearly different brightness.
That was easily fixed by invoking the "Cal1" mode again (which is the one I use) using the Dell Display Manager.
This never happened again.
The other interesting thing here is, as I mentioned the white point differs across the panel (and did just the same for a second unit), which could also be a two controller issue or a panel issue. I am thinking it is more a panel issue, because it is not really a left/right side issue, but gradually changing across the screen.
As you can see in the homogeneity measurement I did. (The SW keeps taking German to me, even if the system is set to English ... the system is located in Germany, okay.)
DCCS' "Native" target is a WRONG choice for 99% of users. It's true that it aims for native gamut but also to native white (GB-LED white is a green-white @7000K, out of daylight locus).
My advice is to set manually R,G,B xy coordinates, and your desired gamma in "Custom" DCCS preset. I think that i posted a year ago how to get native RGB coords but as a summary:
-open "driver's" ICM, if you did not overwrite it with a profile inspector (DispcalGUI, ... othe rtools). You're looking for examplo for "red iluminant relative xy" NOT "PCS relative"
OR
-run ArgyllCMS/dispcaGUI, USE RG_phosphor spectral correction (GBLED's spectra) and run a tools->uncalibrated report (i think that this is the option, I've not a DispaclGUI instalation right now). In its log file it will plot RGB+White+black XYZ coordinates. x=X/(X+Y+Z) y=y/(X+Y+Z) for each R,G,B primary
Also remember that Firefox does not like "Table" profiles from DCCS, a "profile only" (DO NOT CALIBRATE, just make an ICM with current monitor's behaviour) + XYZLUT profile in dispcalgui will fix it.
I understand what you want me to do. I can derive RGB values (an maybe even the right one) by looking into the supplied profile, I can also create a profile from EDID using DisplayCalcGUI.
Both gamuts are different and both are bigger than AdobeRGB. Which is the right one, I will not know. I am also not 100% sure which are the correct values.
I also tried to measure the right gamut using DisplyCalGui, but I failed upto now. Just doing a profile does not work. The process fails with an integer out of range error message and the report you mention does not seem to contain the color values I am looking for.
Let us say I am able to derive some values. When I am using the way you propose, will I not instruct the DCCS to setup an emulation for the requested color space? If I am lucky and my color value are very close to the real values, that might be okay, but still this might also lead to a quite bad calibration ... i would very much like to have a working native color mode in DCCS.
Coming back to this mode for a moment. At least with the UP2715K the color gamut in "native mode" is much much smaller than AdobeRGB, therefore this is really not as expected. When you say, it will calibrate to the native white point, this is another thing I would not have expected.
Coming back to the native calibration, I did not know that it will do native white point. That seems to be a bad idea.
Native is native:) Like in "AdobeRGB" DCCS does not ask you for Whitepoint or gamma, it takes waht "AdobeRGB" means. Same aplies to here, native gamut, native white and native gamma ~2.26 not constant but more similar to sRGB (lighter near-blacks, darker near white than true constant 2.26).
Try as RGB coordinates: Blue=AdobeRGB blue green= AdobeRGB green red= 0.680, 0.310
DCCS is not such accurate to be confused with "out of gamut" red or green. That will aim to native gamut.
BY THE WAY... UNLESS you work with eciRGBv2 (91% coverture... and L* gamma) or Phophoto (that has a not useful gamma) it is pointless to use native gamut in a 8bit workflow. You wont be able to SEE 256 steps in black to sRGB or AdobeRGB red, because you have 256 steps going from black to "native" red. Unless in: Windows, have Photoshop, have FirePro or Quadro vidoe card and use Displayport ... not very wise to aim to such gamuts. Best stay with gamut limited to largest gamut you edit (AdobeRGB for most people)
When Native is native, what is actually calibrated I wonder :)
After some playing around I was able to finish a profile in DisplayCalGui. When I load this in information, there a number of RGB values, but there are quite close. I ended up using the values ended under "colorants (pcs-relative)". Is this the right one?
I used a set of them and gave it a try. The result seems to be okay :) I need to invest further, but the color gamut is visually bigger as it was before in AdobeRGB.
I understand that you advise against using >AdobeRGB if not in a working 10-Bit setup. I am not sure if the U2715K even supports 10-Bit, it is not mentioned anywhere.
However, I needed to get a new graphics card for using this monitor and decided actively against a professional line card and ended up with a GTX970.
I using mostly Lightroom for photography, which is internally using a prophoto gamut and does not support 10-Bit anyway. But special test pictures that show 8-Bit-display-banding in Photoshop do not do so in LR, because LR uses dithering for screen display.
The gamut in my RAW pictures is bigger than AdobeRGB and everything that is out of screen gamut looks like overexposed in a picture, but is not. I worked for more than 3 years in the native color gamut with the U2711 and had no banding issues. On the other hand, I saw quickly that the AdobeRGB full red aso. with the UP2715K seemed a little washed out.
In native you calibrate GREY NEUTRALITY, what calibration is main for. Gamma is secondary since it is managed by color managed apps and white point is a matter of taste and situation (you may want a non daylight white if you exactly know your "paper+normalized lamp" white coords).
As I said before DO NOT USE PCS. PCS (profile conection space) is monitors colorspace translated (matrix) to D50, that space is what color managed apps use. Is said before to use ILLUMINANT RELATIVE.
GTX 970 OR ANY GAME NVIDIA is a BAD advice to use in photography. If you cannot afford a Quadro with the same computing power, get an AMD. Otherwise you'll end with banding (very bad banding in widegamut presets) each time you modify GPU LUT.
Your last paragraph regading desaturation is related to OS default profile (it needs to be the profile you get for THAT specific OSD mode) and the fact that you choose PCS coords which are wrong. Make sure that you fix that 2 issues.
BTW it is IMPOSSIBLE that a widegamut (U2711) in its full gamut configuration (Standard or custom) does not show banding in black to white gradient (I mean in profile colorspace) AFTER you modify GPU LUT in order to calibrate with a gamer nvidia or intel GPU. IT IS NOT POSSIBLE since that hardware has 8bit LUT.
That apply for ALL GB-LEDs (green blue led with phosphor coating for red and 99% AdobeRGB coverture)
It won't solve WP missmatch since it is DCCS fault, the way program "guess" which gain to apply ro RGB native primaries. It does a lot of guessing instead of a iterative approach after measuring PROPERLY native gamma ramps for 3 channels. Thats the reason because DispcalGUI gets so close to desired white when GPU calibrating (even without OSD's RGB gain access)
Your issues are related to DCCS not to your monitor which is fine and has a suberb color uniformity that I've not seen on a dell before yesterday
"DCCS' "Native" target is a WRONG choice for 99% of users. It's true that it aims for native gamut but also to native white (GB-LED white is a green-white @7000K, out of daylight locus)."
Is this true also for me and my UP2414Q?
"My advice is to set manually R,G,B xy coordinates, and your desired gamma in "Custom" DCCS preset."
Do you know the values for my screen? This will solve the blackbody locus issue? If I remeber correctly by default all those coordinates are set to zero but one that is set to 1, not sure what that means.
Thanks Yumichan, this adds another piece to the puzzle. So I should try and get the coordinates for my monitor instead of using Native, understood. Will try the process you described even if it sounds complicated for my pc skills (lack thereof).
So I am left with these issue to solve:
1) the WP mismatch
2) low contrast after calibration
3) I get some banding in the greys also after the internal calibration (no GPU)
I guess they are all DCCS fault so no point returning the monitor?
I am glad you find the monitor uniform because, before applying the uniformity compensation during the calibration, it was not uniformat all, far right was quite bluish, I think that will show up again if I use any of the other present modes.
Do you mind jumping to my other thread? Have left some questions for you plus I posted the results of a DCCS calibration + DispCal profiling, would be glad to have your view on that vs DCCS only.
dopeless
5 Posts
0
February 3rd, 2015 16:00
Hi, I don't have an answer for your question, sorry, but I've got a question for you )
Are you generally satisfied with this display? I mean color quality first of all.
brightcolor
1 Rookie
•
19 Posts
0
February 3rd, 2015 22:00
The display itself is a very nice display overall. I used to have an U2711h before.
The UP2715K has a higher real contrast (Measured as 1:870 when hw-calibrated to AdobeRGB) and due to the "optical bonded" (glued on) antireflective front panel a much higher visual contrast.
Also, there is no pixel distortion effect at all, which was extremely annoying with the U2711 due to heavy anti-glare coating.
Colorwise it is all okay, but no calibrated native color gamut yet. Meaning, I currently lost a bit of color space when compared to the U2711 setup.
Overall, the colors "feel" smoother than with U2711. Visually the picture is much better due to contrast
There is one tiny thing with the whitepoint. The whitepoint varies from -190 K to +190 K when comparing left and right side of the panel to the measurement in the middle. For the Dell Calibration Solution this is still green because its "all-good value" is everything below 250 K. But the effect is clearly visible.
I know that the higher Nec or Eizo have a compensation (DUE) for this kind of thing, but the UP2715K has not. Not even the static version some Dells have.
I tried a second UP2715K which had the same (visually, not measured) effect. Question is, if this is a panel or electronics issue.
dopeless
5 Posts
0
February 4th, 2015 02:00
Thx,
I've got this display several days ago and was a bit dasappointed. I was expecting it to work fine out of the box. but it's a lettle yellowish in all modes. Nice white (for my eyes) is only in 9300k and 10000k modes. Both precalibrated profiles look the same and not quiet good. I'll get an xrite in several days, maybe it'll get better. Was your display good calibrated out of the box?
dopeless
5 Posts
0
February 4th, 2015 14:00
Yep true whit different to everyon, but it just looks too yellow for me.
I've got another question. I've read in one of the discussions, that someone got different calibration results on every half of the display due to two controllers after hardware calibration. Results are on this image
ic.pics.livejournal.com/.../84470_original.jpg
have you got this issue?
brightcolor
1 Rookie
•
19 Posts
0
February 4th, 2015 14:00
The out of box calibration looked okay to me visually. I did not measure it.
The HW-calibration is a bit different. One thing is that the HW-calibration results in 6650K and not 6500K as it is supposed to.
Frankly, most people that have the same experience like you are usually first timers with displays calibrated to 6500K which is "the usual" whitepoint. They are just used to colder, bluer white points like 9000K or 10000K and need some time to adapt.
You will see which is it when you get the xrite.
Strangely enough there is not "the one, right white". There are a lot of them and all in different colors :)
dopeless
5 Posts
0
February 4th, 2015 15:00
Yes, white point is a matter of taste, but I still feel it's to yellow for me.
Onother important question. I've read in discussion, that someone have got different calibration results on every half of the monitor after hw-calibration due to two controllers.
Have you got this issue too?
ic.pics.livejournal.com/.../84470_original.jpg
brightcolor
1 Rookie
•
19 Posts
0
February 5th, 2015 00:00
It happened to me one time after a first start of monitor and PC that the calibration was obviously not loaded into one of the controllers.
So, one side of the display had a clearly different brightness.
That was easily fixed by invoking the "Cal1" mode again (which is the one I use) using the Dell Display Manager.
This never happened again.
The other interesting thing here is, as I mentioned the white point differs across the panel (and did just the same for a second unit), which could also be a two controller issue or a panel issue.
I am thinking it is more a panel issue, because it is not really a left/right side issue, but gradually changing across the screen.
As you can see in the homogeneity measurement I did. (The SW keeps taking German to me, even if the system is set to English ... the system is located in Germany, okay.)
yumichan
3 Apprentice
•
739 Posts
0
February 10th, 2015 11:00
DCCS' "Native" target is a WRONG choice for 99% of users. It's true that it aims for native gamut but also to native white (GB-LED white is a green-white @7000K, out of daylight locus).
My advice is to set manually R,G,B xy coordinates, and your desired gamma in "Custom" DCCS preset.
I think that i posted a year ago how to get native RGB coords but as a summary:
-open "driver's" ICM, if you did not overwrite it with a profile inspector (DispcalGUI, ... othe rtools). You're looking for examplo for "red iluminant relative xy" NOT "PCS relative"
OR
-run ArgyllCMS/dispcaGUI, USE RG_phosphor spectral correction (GBLED's spectra) and run a tools->uncalibrated report (i think that this is the option, I've not a DispaclGUI instalation right now). In its log file it will plot RGB+White+black XYZ coordinates.
x=X/(X+Y+Z)
y=y/(X+Y+Z)
for each R,G,B primary
Also remember that Firefox does not like "Table" profiles from DCCS, a "profile only" (DO NOT CALIBRATE, just make an ICM with current monitor's behaviour) + XYZLUT profile in dispcalgui will fix it.
brightcolor
1 Rookie
•
19 Posts
0
February 10th, 2015 14:00
Hi, thanks for your answer.
I understand what you want me to do. I can derive RGB values (an maybe even the right one) by looking into the supplied profile, I can also create a profile from EDID using DisplayCalcGUI.
Both gamuts are different and both are bigger than AdobeRGB. Which is the right one, I will not know. I am also not 100% sure which are the correct values.
I also tried to measure the right gamut using DisplyCalGui, but I failed upto now. Just doing a profile does not work. The process fails with an integer out of range error message and the report you mention does not seem to contain the color values I am looking for.
Let us say I am able to derive some values. When I am using the way you propose, will I not instruct the DCCS to setup an emulation for the requested color space?
If I am lucky and my color value are very close to the real values, that might be okay, but still this might also lead to a quite bad calibration ... i would very much like to have a working native color mode in DCCS.
Coming back to this mode for a moment. At least with the UP2715K the color gamut in "native mode" is much much smaller than AdobeRGB, therefore this is really not as expected.
When you say, it will calibrate to the native white point, this is another thing I would not have expected.
Coming back to the native calibration, I did not know that it will do native white point. That seems to be a bad idea.
yumichan
3 Apprentice
•
739 Posts
0
February 10th, 2015 15:00
Native is native:)
Like in "AdobeRGB" DCCS does not ask you for Whitepoint or gamma, it takes waht "AdobeRGB" means. Same aplies to here, native gamut, native white and native gamma ~2.26 not constant but more similar to sRGB (lighter near-blacks, darker near white than true constant 2.26).
Try as RGB coordinates:
Blue=AdobeRGB blue
green= AdobeRGB green
red= 0.680, 0.310
DCCS is not such accurate to be confused with "out of gamut" red or green. That will aim to native gamut.
BY THE WAY... UNLESS you work with eciRGBv2 (91% coverture... and L* gamma) or Phophoto (that has a not useful gamma) it is pointless to use native gamut in a 8bit workflow. You wont be able to SEE 256 steps in black to sRGB or AdobeRGB red, because you have 256 steps going from black to "native" red.
Unless in: Windows, have Photoshop, have FirePro or Quadro vidoe card and use Displayport ... not very wise to aim to such gamuts. Best stay with gamut limited to largest gamut you edit (AdobeRGB for most people)
brightcolor
1 Rookie
•
19 Posts
0
February 11th, 2015 00:00
When Native is native, what is actually calibrated I wonder :)
After some playing around I was able to finish a profile in DisplayCalGui. When I load this in information, there a number of RGB values, but there are quite close.
I ended up using the values ended under "colorants (pcs-relative)".
Is this the right one?
I used a set of them and gave it a try. The result seems to be okay :)
I need to invest further, but the color gamut is visually bigger as it was before in AdobeRGB.
I understand that you advise against using >AdobeRGB if not in a working 10-Bit setup. I am not sure if the U2715K even supports 10-Bit, it is not mentioned anywhere.
However, I needed to get a new graphics card for using this monitor and decided actively against a professional line card and ended up with a GTX970.
I using mostly Lightroom for photography, which is internally using a prophoto gamut and does not support 10-Bit anyway. But special test pictures that show 8-Bit-display-banding in Photoshop do not do so in LR, because LR uses dithering for screen display.
The gamut in my RAW pictures is bigger than AdobeRGB and everything that is out of screen gamut looks like overexposed in a picture, but is not.
I worked for more than 3 years in the native color gamut with the U2711 and had no banding issues.
On the other hand, I saw quickly that the AdobeRGB full red aso. with the UP2715K seemed a little washed out.
Thanks again.
yumichan
3 Apprentice
•
739 Posts
0
February 11th, 2015 01:00
In native you calibrate GREY NEUTRALITY, what calibration is main for. Gamma is secondary since it is managed by color managed apps and white point is a matter of taste and situation (you may want a non daylight white if you exactly know your "paper+normalized lamp" white coords).
As I said before DO NOT USE PCS. PCS (profile conection space) is monitors colorspace translated (matrix) to D50, that space is what color managed apps use. Is said before to use ILLUMINANT RELATIVE.
GTX 970 OR ANY GAME NVIDIA is a BAD advice to use in photography. If you cannot afford a Quadro with the same computing power, get an AMD. Otherwise you'll end with banding (very bad banding in widegamut presets) each time you modify GPU LUT.
Your last paragraph regading desaturation is related to OS default profile (it needs to be the profile you get for THAT specific OSD mode) and the fact that you choose PCS coords which are wrong. Make sure that you fix that 2 issues.
BTW it is IMPOSSIBLE that a widegamut (U2711) in its full gamut configuration (Standard or custom) does not show banding in black to white gradient (I mean in profile colorspace) AFTER you modify GPU LUT in order to calibrate with a gamer nvidia or intel GPU. IT IS NOT POSSIBLE since that hardware has 8bit LUT.
yumichan
3 Apprentice
•
739 Posts
0
February 11th, 2015 03:00
That apply for ALL GB-LEDs (green blue led with phosphor coating for red and 99% AdobeRGB coverture)
It won't solve WP missmatch since it is DCCS fault, the way program "guess" which gain to apply ro RGB native primaries. It does a lot of guessing instead of a iterative approach after measuring PROPERLY native gamma ramps for 3 channels.
Thats the reason because DispcalGUI gets so close to desired white when GPU calibrating (even without OSD's RGB gain access)
Your issues are related to DCCS not to your monitor which is fine and has a suberb color uniformity that I've not seen on a dell before yesterday
Mais1978
1 Rookie
•
82 Posts
0
February 11th, 2015 03:00
"DCCS' "Native" target is a WRONG choice for 99% of users. It's true that it aims for native gamut but also to native white (GB-LED white is a green-white @7000K, out of daylight locus)."
Is this true also for me and my UP2414Q?
"My advice is to set manually R,G,B xy coordinates, and your desired gamma in "Custom" DCCS preset."
Do you know the values for my screen? This will solve the blackbody locus issue? If I remeber correctly by default all those coordinates are set to zero but one that is set to 1, not sure what that means.
Mais1978
1 Rookie
•
82 Posts
0
February 11th, 2015 03:00
Thanks Yumichan, this adds another piece to the puzzle. So I should try and get the coordinates for my monitor instead of using Native, understood. Will try the process you described even if it sounds complicated for my pc skills (lack thereof).
So I am left with these issue to solve:
1) the WP mismatch
2) low contrast after calibration
3) I get some banding in the greys also after the internal calibration (no GPU)
I guess they are all DCCS fault so no point returning the monitor?
I am glad you find the monitor uniform because, before applying the uniformity compensation during the calibration, it was not uniformat all, far right was quite bluish, I think that will show up again if I use any of the other present modes.
Do you mind jumping to my other thread? Have left some questions for you plus I posted the results of a DCCS calibration + DispCal profiling, would be glad to have your view on that vs DCCS only.
Many thanks