U2413, Real gamut coverage

OK, I did not read  the text in TRC windows combo box in the lower part of the window. Forget that picture.

Ok, then.

I have calibrated the monitor in other PC (with and old 8600GTS, I don't have any AMD). I did as you explained without installing anything. The coverage is the same (1% more ~92%).

I asked Xrite for a diagnostic on the i1Display Pro and they addressed me to software i1Diagnostics, but after downloading, I see that even the last version 4.1 (16/6/2014) do not support i1DisplayPro !

I asked them for an alternative or what to do, but they give silence as answer...

On the other hand I have asked Amazon for a 4th (and last) replacement. The third arrived as bad as the second replacement, so I don't have hope.

Dell support here, in order to exchange one of the monitors asks me for a 'purchase order number' (number issued by Dell when Amazon bought the monitor stock from Dell) which Amazon should give me, but Amazon says they do not know what I am talking about...

If only I could test the i1Display Pro...

Hi

DELL has said they are going to send another monitor. When I receive it I'll report.

At side of the TRC misunderstanding, I am trying to verify everything involved and I may have the chance to get an AMD FirePro V4900, would it be ok? There is preference to be FirePro or Radeon?

I would need to know before four hours from now, if possible in order to have it tomorrow.

Thanks!

FirePro: 10bit under Windows, Photoshop, Displayport and these kind of Dells. Fully GPU calibratable without banding. More expensive for the same computing power. Certified drivers for Autodesk suite and such CAM programs.

Gamer AMD: 8bit workflow. Fully GPU calibratable without banding (due its 12bit LUTs, for "Custom Color" mode). Cheaper than FirePro for the same computing power (OpenCL, games...)

Thanks for you input!

Likely I am wrong, as I don't know too much, but I was thinking yesterday about that TRC and the calibrations performed.

1. I have seen that the TRC in absolute shows three different lines when calibrating with D65 or 6500K as whitepoint, but that TRC curve in absolute, is virtually a single curve (the trhee ones are superposed), when I calibrate with D50 or 5000K.

I interpret that curves as the correction that must be made to each channel to reach the wp desired. If that is correct (?) then that made me think (maybe wrong) that  the native white point of the monitor in let's say Cal1 (this is the mode I have calibrated more) after a reset, is something closer to 5000K, so the calibration only needs to make slight corrections.

2. Additionally, when I calibrate for WP of 5000K the coverage of AdobeRGB space increases in 3-4% respect the coverage obtained with 6500K.

3. But on the other hand, when I calibrated the monitor in native ot test, (DCCS) and 'As measured' (DCGUI Cal&Prof), in preset Cal1 after a rest, both find something around 6400 / 6500 K as native wp, but far from 5000K.

At this point points 1 & 2 make me think that the native WP is around 5000K, but the point 3 seems to contradict it. Likely I am wrong with tha logic, but I don't know. Do you know how to interpret that, or I am messing with things wrongly?

On the other hand, I still do no t have answer from Xrite related to diagnostic of i1Display Pro. I have seen from its box / package that my unit is from 11/2010, which seems too old for GB-LED compatibility?. I was asking myself if the results obtained could be explained by a bad i1Display Pro unit, maybe it sees colors with a bluish redish dominant?

NO, natie WP is at R100 G100 B100... as most GBleds it should be a cool "cian-white".

As I said in my previous post, I didn't read text in combo boxed. Right one is relative. What you saw in absolute is related to PCS transformations. Read about Profile Connection Space (PCS) first, it is always D50 no matter what WP your monitor has and is not related to your issues. Forget TRC in absolute, is not what you meant.

Ok, I think I understand it more or less. Thank you for clarification.

I got yesterday a brand new AMD FirePro V4900 and tested with a fresh Windows 7 x64 install, only with mobo drivers and AMD drivers. I tested as you described without installing anything (dispcalGUI from ZIP, edr file imported from i1profiler.exe installer, and i1Display Pro simply plugged -no DCCS install, nor i1profiler install). The results are more or less the same (plus / minus 1% max of coverage vs what was found with the nvidia).

I tested also in another PC with an nvidia 8600 GTS, with same results (maybe 1% more max of coverage).

So, a priori, it does not seem related to the graphic card, or PC. I got still a brand new i1Display Pro to see if the fault comes from the one I used (which was also brand new, but who knows). I'll try to test with the new one today.

I have read around that these monitors U2413 (GB-LED based) have a green/cyan - white tint in native (without calibration) due to the GB-LED nature. Nevertheless the monitors I received in native have a WP around 6400K (in DCCS) and closer to 6500 from dispcalGUI cal&prof, and visually I don't see at all any green/bluish tint, I would say that visually they are ok, and quite neutral (thinking on a 6500 K temp).

That makes me think that maybe the edr / ccss file from i1profiler/DCCS does not fit anymore, as it did with the U2413 that showed native-greenish tint until now (maybe these are new DELL revisions, after all they are manuf in 2014 but Rev A01, or maybe the panel LED supplier has changed slightly led manufacture/specs ?). 

I guess that to clarify that, I should make a spectra measurement with some labo grade spectrophotometer (3 nm step or so, I think), but this is out of my possibilites (to acquire or to know someone that could do it for me), even an i1Pro 2 is out of my reach (which I understand has 5? or 10? nm steps) which could miss the green peak, except that this 'eventual' revision of monitor would have a thicker/less narrow green peak, which I strongly doubt as led emission always produce peaks (anyway if i1Pro was ok I could try to look for a rent).

At this point, I wonder if I could test with the ccss/ccmx files in dispcalGUI colorimeter correction database for U2413 -files from 2014-, which may be related to these 'less green' new 'eventual' revisions. What I wonder is how are these ccss/ccmx files built? With an i1Pro?, then the remedy could be worse than the problem? And if the result is ok with these 'third-party' ccss (green xy closer to aRGB green xy), how to be sure it is real?

So, what do you think about that ? Could it be the cause? Do you recommend me to test recently created/uploaded ccss ccmx corrections?

CONTINUE IN NEXT POST

In fact I am referring to Chris-M post here http://en.community.dell.com/support-forums/peripherals/f/3529/t/19525792#, where the .edr correction for U2413 plotted belongs to ~6700K, meanwhile the native of my U2413's monitors (three of them) is around 6500 (*) or slightly less. So the correction -I guess- should be different...(?)

(*) Value measured with i1Display Pro so with the .edr/ccss correction from DELL/XRITE, so if my U2413s have another SPD (so the reading is not so good) which is the real native temperature of these monitors? Would it be higher or lower than the value reported?

Another thing is that if it has really a different SPD, should it still be able to cover a 99% of aRGB?

Maybe Chris-M could apport some info or factory confirmation about these -presumably- new U2413 units with warmer native WP and its specs (native spectra and aRGB coverage mainly)?

I saw the U2413 review at tftcentral, it shows one unit with native WP of ~6390 and still it covers nicely aRGB space in native. This review is with i1Pro and LaCie's Blue Eye Pro, so it is not influenced by colorimeter correction files (although it should be limited by the nm band precision of i1Pro), nevertheless their results are ok. So it makes me think that the problem could be in the colorimeter correction or in the colorimeter itself (will see this afternoon). I don't know if my problem is that (3 monitors in a row with same problem seems strange), or if the tftcentral review uses a cherry picked monitor to make a nice review but it is not real...

I don't know if I am alone with this problem, neither.

The beauty and power of spectral corrections is that WP chromaticity coord does not mater. ANY color which is linear combination of 3 channels described in EDR could be measured properly. In fact, RG_phosphor RGBW entry for GBLED (last four) have a WP of 5800K non daylight.

Thats why no CMX are needed. You have 3 functiosn described by a graph (CCSS): R(lambda), G(l), B(l).

Any color in gamut is described by :

Color = GAIN1* R(l) + GAIN2*G(l) + GAIN3*B(l), where gain are not OSD gains. So while GBLED spikes locations and rise and fall are the same as monitor's, RG_phosphor are valid no matter what native WP edr as or has your unit. There are several CCSS on DispcalGUI database for GBLEDs and many are wrong: thery are obtained from an emulated (limited gamut) which is a mistake (clipping problems), they are 10nm resolution which are not so accurate... Despite this, if they are obtained from natvive gamut it only make 1-2dE drift from RG_phosphor EDR.

In your case, B & R coordinates seems right, so the problem is that G(lambda) function in your U2413 integrated over CIEXYZ obverver functions does not give the same (after normalization) coordinates as "normal" (not "previous") U2413s. There are NOT ccss files (AFAIK and I seen many up to June 2014) in DispcalGUI database that behaves like this, maybe you could check is there is one. REMEMBER TO AVOID emulated gamut CCSS (ther are easyly spotted in a graph since they have one or more channels that are a lineal combination of native itself an the other channels.).

My avice after discard GPU, Computer and i1DisplayPro is to return your U2413 for refund to amazon an buy (if you want to) a new one from aonther reseller in your country since your units may be defective and habe been returned many other times (I've not seen ths behavior on brand new U2413/U2713H bought from resellers in last months). As a las chance buy it from dell.es but I think that great overprice over resellers is not justified.

If you wish to make a custom ccss for your U2413 with a i1Pro, REMEMBER to make it IN NATIVE GAMUT (custom color OSD mode, RGB gain 100, even perform a factory reset) and using high res driver from ArgyllCMS(3.3nm) since I1Pro "as is " has a low spectral resolution for LEDs (10nm)

Hi

Thank you very much. I followed your advice, returned all monitors back to Amazon for a refund, and bought another monitor from another reseller (they just received new stock). I have just opened it, the light bleeding is much better, and the OSD responsiveness to the finger buttons is much better, I have had no time yet to measure gamut coverage, but I found that both presets, adobeRGB and sRGB are the same,

Is that normal? I would say no, I would expect that when selecting sRGB, the reds (for example) to be less saturated (in general, all colors), but it is the same.

This happenned also with the last 3 monitors form amazon (not with the first one). And as explained before, when I did a 'color settings' reset in OSD, they became corrupted. I have not done a reset in this one yet.

I use a test image in PS to check, I am missing some configuration in order to use the sRGB and aRGB? I have not found in internet any input about some DELL with aRGB and sRGB presets looking the same... which seems strange to me, as I have received 4 in a row. Is this my fault?

I'll test gamut coverage, if I have some time.

Thank you. I tested and there is no difference in any color ramp between aRGB and sRGB (with the U2413.icm) in Windows CP. Even slightly.

Although I activated the 'Use Windows screen calibration' in order to try to be sure the U2413.icm is loaded by OS, I ran also the DPGUI calibration and icc profile loader, but nothing changes.

When changing from aRGB or sRGB to CAL1 or Custom / Standard all become more saturated, specially reds.

If so, then this is monitor fault? I can NOT believe that bad quality or to have so much bad luck... 

Nobody has bought and U2413 from july or later and found this behaviour ? I wonder what can I do.

Measure it, green primary for each precalibrated mode. Is hard to believe such behaviour.

I measured with DCGUI the presets AdobeRGB and sRGB in native (so only setting the brightness to 20 to keep it at 120 cd/m2), the rest is left 'as measured'.

AdobeRGB preset:

[View:/cfs-file/__key/communityserver-discussions-components-files/3529/Measurement-Report-2_5F00_5_5F00_0_5F00_0-_1420_-DELL-U2413-Sept14-AdobeRGB-Native.zip:550:0]

And the sRGB preset:

[View:/cfs-file/__key/communityserver-discussions-components-files/3529/Measurement-Report-2_5F00_5_5F00_0_5F00_0-_1420_-DELL-U2413-sRGB-preset-Native.zip:550:0]

I see that both are pretty the same (!?). REALLY STRANGE, specially when this is the 4th with that behaviour... Monitor fault? It is incredible...

I am afraid of making a reset, likely the presets will become corrupt, as with the other three monitors.

As a side note: I calibrated also with DCGUI in Custom, and, for the first time, the RGB Gains are more inline with your suggestions some post above, around R100 G91 B98. Giving aRGB coverage of 97.3% (6500K), or even 98.0% if wp is 6400K... I have still to calibrate it with DCCS, which in the other monitors gave always 2 or 3 % less of coverage, than DCGUI (why would it be like that?)

Make sure Dell Display manager is not doing something weird. Set it to manual.