U2713H + X-Rite i1 Help a newbie

1) Disable all color enhancements in your graphics card driver. Do not enable color vibrandce, digital vibrance or such things, all OFF

2) Disable "Smart video enhance" in your U2713H OSD

3) calibrate CAL1 in DCCS for example and your desired gamut. You do not need to set CAL1 on OSD before DCCS starts. It will prompt you for calibration storage later "calibration 1"=CAL1, same to CAL2.

4) Open this URL in a non color managed browser, this is important. It needs to be A BAD BROWSER, a non color managed one, like Internet Explorer or Opera. Do not use Firefox.
[View:www.lagom.nl/.../gradient.php:550:0]

Did you see a VERY noticeable colorarion (mostly red or green) => this is caused by DCCS and their useless dev team. It takes 4 x11 (RGBW) measures in black to RGBW ramps in order to interpolate correction. If your device presente some kind of anomally between these measures, it won't correct it.

Since Dell (or Asus or LG or Benq) does not use "selected panels" from LG (IPS) in their assembly process, this kind of thing "could" happen (it happens). Maybe NEC's  Spectraprofiler or Eizo's Color navigation need only 10 measures per ramp balck to color beause of the very low nonlinearities in panel response.... but for "B" grade panels more patches are needed in black to color ramps.

Dell (nor Benq) nor Xrite care about that. DCCS (custom version of i1Profiler) won't be fixed for this.

Perform all 4 steps (ALL FOUR, disable all color enhancements in your grapghics card driver, nvidias are very very bad to work wih photography, nvidia is A DECADE BEHIND AMD/ATI in solutions for image treatment). Check uniformity in grey ramp, if it is VERY noticeable, return your dell for refund.

Also consider to sell that useless nvidia and get a gamer AMD (if you game or want OpenCL computation power) or an AMD Firepro (in order to enjoy the marvelousness of a 10bit workflow where color management rounding errors do not exist)

Hi
I was searching the net as had the exact same problem and came across this post, also using an i1 Display Pro to calibrate. I'm a web developer and noticed that some greys, at the light end, are too warm to be considered a real grey. Bootstrap styled websites tend to use a shading of around RGB 249,249,249 for some elements, this is where it seems to go wrong, and darker greys seem okay.

To answer your questions, it doesn't matter if you use HDMI or Display port, or which USB port you connect the i1 to, the results will all be the same.

Note I have two Dell monitors, a U2415 and an older U2412M, both have been calibrated within a few minutes of each other, connected to the same PC, and the U2412M doesn't have the grey issue, but the new U2415 does. Having both monitors makes it easy to spot the problem by dragging across to straddle both screens the light grey colour, with the U2412M looking perfectly light grey, and the U2415 not looking grey on these very light shades, they are far too warm.

I can use the default ICC Dell colour profile, this makes the grey at 249,249,249 appear truer, but shifts the problem to greys a bit darker. All the while the U2412M is rendering grey consistently across all shades. I've re-calibrated several times and can't correct the issue with very light shades of grey.

See http://getbootstrap/css, and I know the shaded code example areas from working on other PCs and monitors is a light grey, but it doesn’t look light grey on the U2415, it’s almost a cream colour.

It sounds like you have the same issue with their more expensive and larger panel, perhaps this is a firmware bug which is common to these monitors.

The i1 Display has been used on several displays on different computers, and I've checked these and greys all look good, it is just my U2415 that has a problem.

As the i1 Display (and others like it) only measure samples of each colour, it can't measure every one of the 16million colours a display can show, it has to assume that the display is linear enough that it can skip and just sample a tiny subset of colours, so for grey it might jump in 16's so that the last sample before white is 239, 239, 239, you can see these samples on the screen of course as it measures. If the monitor has a wonky bump in its linear rendering of grey and that bump is say sitting in the middle of 239 to 255, then perhaps that is what we are seeing. So this could be firmware or calibration bug affecting these newer models.

Regards

Phil

Hi

Thank you for the response.

-wrong calibration program. Use DispcalGUI with WLED xrite spectral correction.

I will try that program, however the program calibrates fine, with all things being the same (same graphics card, same computer) with a Dell U2412M, plus an older monitor before that.

-a bad GPU for calibration. If you do not have an AMD GPU, and that GPU actually handles the physical output of your computer (for laptops and ultrabooks, no problem with desktop computer dedicated GPU) you'll get banding. That banding causes red to green oscillations in grey ramp due to poor quality LUTs in GPU, they have limited bitdepth. Unless you get an AMD GPU and use a program that loads properly ICM VCGT to GPU LUT, you won't get rid of this.

It is an Intel PC using Intel HD 4000 graphics, both monitors connect to the same graphics cards, only the U2415 experiences this problem.

-browser like Firefox in full color management mode. U2415 has a gamut a little larget than sRGB. Your GPU cakibration can be not exactly with a sRGB gamma, HENCE, Firefox color manages "RAW" RGB values to be rendered in sRGB colorspace. You have only 8bit per channel to do calculations (a browser has to be FAST, it os not GMIP nor Photoshop), hence some banding will arise due to color management.

The issue with incorrect greys is present everywhere, not just in Firefox or a browser, I will try calibration to sRGB colorspace to see if that helps.

How do you know this isn't related to the original post, they never came back with a solution, if they ever found one?  If I had posted a new thread I'd get someone telling me it's already been asked and not to duplicate threads, I find someone with the exact same symptoms and I then get told it's not related.  I couldn't win.

Regards

Phil

Because the different performance of the two monitors is caused by two different native gamma responses, each one have their particular calibration  (aka GPU LUT correction). The bigger the correction, the bigger the grey ramp oscilation. YOUR ISSUES AREC CAUSED BY POOR LUT PERFORMER GPY (non AMD/ATI, non Quadro card) plus (once you get a proper GPU for calibration) a bad program (unles you use DiscalGUI)

Unless you do THAT, you wont get rid of your issues since they are no related to monitor, but for GPU hardware unable to calibrate EVERY monitor.

The original thread issue was caused by asumptions made by DCCS which are false: they took to few patches no measure each channel gamp ramp, hence if there were gamma anomalies between that intervals not related to measured points (start and end of interval) they won't corrected. THIS IS NOT RELATED to GPU, since LUT is loaded insde monitor, a linear LUT (input=output) is loaded in GPU LUT so no GPU correction is applied (unless user screws something because its lack on knowledge of what he is doing)

DispaclGU takes (could take) an iterative process which in its final grey ramp calibration stage will measure 96/256 greys. Unless you have a very very deffective monitor (or TV connceted to an HTPC), 96 greyramp point will get a proper calibration (256 entries of 16bit fro VCGT tag of ICM profile)... so DCCS statement related to your cause:

-is false, because is related to the program you used, which CANNOT be DCCS sinec you do not have hardware calibration

-is false, since if even using PROPERLY DispcalGUI you get banding (green-red oscilation on grey ramp) these scilations are caused by your poor performer GPU (iGPU or nvidia, same cr4p are both of them) because they have 8bit LUT entries (bad rounding errors). AMD/ATI cards have 12bit per entry (10bit the eldest) plus temporal dithering to output to an 8bit HDMI/DVI/DP output

Open this URL in a non color managed browser, Internet Explorer if possible:
[View:www.lagom.nl/.../gradient.php:550:0]
Do you see vertical green-red oscilatiosn on grey, her you have the answer WHICH IS NOT RELATED TO THIS THREAD, or DCCS but because of your GPU, mostly.

Hi

I have since resolved this issue of light greys being far to warm.

I tried the DispCal GUI but it gave the exact same result, just took a lot longer, although does seem very thorough it did nothing to help.

The problem appears to be due to the Color settings on the monitor.  I needed to change the RGB settings to get to 6500K, and needed more blue in my case.  As the color setting menu defaults to 100 for each of Red, Green and Blue, the only way to get more Blue was to knock down the Red and the Green, this left Blue at the 100 mark, and Green and Red at 96 and 97. 

Playing about with settings after another calibration and seeing the same troublesome grey, I discovered, with the warm grey shade on the monitor, that knocking down the Blue channel a couple of notches suddenly saw the warm grey become true grey, which is odd considering that reducing blue is actually warming the display (and whites did become a little warmer) and it should have been the opposite effect!  So after some trial and error, getting 6500K by dropping Blue down to 96, then red and green down to 93 and 92, and another calibration, greys are now grey all the way up through a grey gradient, matching my other monitor.

I suspect with Blue at 100 in the color menu, and after calibration, the Blue pixels in the panel were in effect full on before they saw 255, i.e. they hit their ceiling too early, so grey up to around 245, 245, 245 was okay, but as it went higher, I was getting no more blue, so grey become warmer and dirty looking.  If you have a correctly calibrated monitor (as I do now) and create a colour swatch as 249, 249, 245, that gives you the sort of grey I was looking at!

I would say to the OP, have you adjusted the RGB color values to get your white adjustment before calibrating, if yes, is Blue still at 100 or very close to the top?  If that is the case, adjust the color menu to something like 94, 94, 94, then adjust from there, and make sure Blue doesn't go above 96 or 97, I guess the same applies for Red and green as well.  We basically need some headroom left.  Alternatively, I expect if you select the preset 6500K color temperature (which is calibrated with that headroom), and you calibrate, you'll find the grey problem gone as well.  In my case the preset 6500K was more like 6300K hence using the RGB color settings.

Regards

Phil

No... you are not even close to know what are you talking about, so you write the kind of nonsenses we have been reading last days.

1) RGB gain controls fixes white, then grey neutrality and brightess are fixed in GPU LUT. Sice you have that garbage of GPU so many levels loist results on banding. With an AMD GPU that wont happen even correcting white in GPU LUT (even losing levels)... so you were experiencing the hardware limitation I was trying to explain to you but... it seemed to complicated fo you... or even didn't read.

2) He has AN HARDWARE CALIBRATABLE MONITOR....so CAL1 and CAL2 HAVE all this stuff writen in the LUT3D, plus gamut limitation to arbitrary RGB primaries in gamut. All is done in DCCS. So your last post is a nonsense or a joke. You do not have even access to RGB gain controls. Custom color mode, with its 6 axis CMYRGB, RG gain, offset, hue... try to guess to waht is translated by monitor electronics... (jsut a clue... internal LUT)

Try to know what you are talking about before posting that kind of "magic" explanations which actually are what have been explain to you.

Hi

Well garbage or not I fixed it, nothing you suggested worked, although I do thank you for you input, it just didn't help at all.  I've checked for banding and don't see any, so perfectly happy to have light greys back to being grey.

Nothing wrong with my graphics card as I explained, the second Dell monitor had no issue.  It was simply the Blue channel being full on in the menu which meant it hit a ceiling (clipped) before green and red, this manifested itself in problems display neutral greys above around RGB values 249.  

I can now enjoy my new two monitor setup, I thank you for your input, my explanation you may think rubbish, but it solved the issue completely, I'm happy, and hopefully this will help others.

Regards

Phil

Your issues were caused because what I explain to you, it's your problema if you do not believe it. If it was solved was due to easier LUT calibration, PERIOD, so please educate yourselft about calibration before posting nonsenses.

And as a final note, id tou actually performed an GPU calibration, I mean, calibrate greys for desired brightness (gamma) and neutrality (same "color" as your desired white, YOU HAVE BANDING with an iGPU.
The opposite if fALSE, al lie, a way to lead people to missunderstandings and make them WRONG hardware for calibrate displays. Is as easy as this.

So before posting nonsense facts educate yourseft a little about this matters, first about what this thread was about, LUT3D calibration which has NOTHING to do with your issues, second about the actual causes of your issues (GPU LUT with 8bit per entry).