UP2414Q + XPS13 Questions

For 10bit you need a nvidia Quadro or AMD Firepro conected through DisplayPort. For laptops/ultrabook the same applies BUT miniDP output must be driven directly by Quadro/Firepro (not by intel graphics card)... so is very likely it will work on many "laptop workstation" models (Dell Precision??) unless it is stated explicitly.

Bonus Answer: The ONLY way to internally calibrate these Dell display is with i1Displaypro (best quality)/i1Pro/i1Pro2. Other kinds of devices and solutions will calibrate the monitor in graphics card LUTs, not monitor's LUT. When calibrating any kind of monitor in graphics card LUTs you will get banding and uneven gradients (which is a pain in widegamut models) UNLESS you have an AMD graphics card (12bit LUTs), or maybe a Quadro, but they should drive digital output (HDMI/miniDP) directly... which does not apply to 99% of laptops (outputs driven by intel GPU so it uses intel LUTs, discreted graphics used only to do calculations). Desktop computers with gamer AMD cards, Firepro AMD or nvidia Quadro (not NVS Quadros) cards will be able to calibrate any kind of monitor in cards' LUTs without banding.

So it seems that you do need a i1DisplayPro.

If you would clarify one point, you mention that I need a Quadro or Firepro graphics board to take advantage of a 10-bit color display, but go onto say that it's very likely laptop workstations will work. So, does this mean conclusively that my XPS 13 9343 with Intel 5500 GPU will not support 10-bit color under any circumstance, even if I have an i1DisplayPro?

Exactly. 10bit support in Photoshop requires Quadro or firepro OpenGL 10bit output support, Displayport and "10bit input" monitor (regardless of actual monitor panel bitdepth).

This is not a Dell's limitation, it's the same for Asus, NEC, Eizo...

Maybe you could bet deep color in DirectX without a Quadro or Firepro, but Photoshop and software like that uses OpenGL so... useless.

When I said than even some laptop workstations with Quadros won't give you 10bit workflow, I meant that DisplayProt output must be driven by Quadro/Firepro. Most laptop workstations (Lenovo for example) use Quadros to do calculations but HDMI/Displayprot outputs and laptop's screen are driven by integrated intel GPU and its LUTs. So no 10bit in that cases.

The cheapest way to get 10bit is a desktop computer with nvidia K620 or Firepro w2100. If laptop is needed, ask manufacturer for actual 10bit support (Dell and its Precisions, for example)

K420 = Yes, it' seems that has 30bit support. From PNY:

"Plus, an all-new display engine drives up to four displays with DisplayPort 1.2 support for ultra-high resolutions like 3840 x 2160 at 60 Hz with 30-bit color. - See more at: www.pny.com/nvidia_quadro_k420;

I didn't know that a lower verison of K620 existed. Thanks for the info

AMD 6850, of course if will show no banding if you calibrate in GPU AdobeRGB,sRGB or Custom OSD modes. Just make sure that you use 3rd party LUT loaders like ArgyllCMS rather than windows loader.

But I think you missunderstood me, I use ArgyllCMS/DispcalGUI to "reprofile" (just measure and store monitor behavior in an ICC v2 profle), I do not  "recalibrate" (calibrate in LUT3D and then "fix" white and grey ramp). I do that because DCCS matrix profile asumes that after calibration you get a true perfect neutral grey and DCCS table profile is not compatible with Firefox and otehr browsers. The same is true for i1Profiler (DCCS is actually a modified i1Profiler)

The wider the gamut, the more noticeable GPU LUT calibration banding is. UP2414Q factory sRGB OSD mode wll show less banding with an intel HD5500 if calibrated in GPU, than factory AdobeRGB osd mode calibrated with HD5500. Think of an elastic grid, same number on knots/nodes (RGB 8bit values). The more you expand it (bigger  gamut), the further nodes are from each other. If the distance between "grey neighbours" is higher it's harder to correct them, to neutralize grey ramp to a given white color.

Same applies to DCCS, it has "no banding" due to monitor's internal high bitdepth LUT but it could show some "smooth" grey colorations towards green-red on native or AdobeRGB  DCCS gamut emulations which are less likely to appear on sRGB-like DCCS gamut emulations. I do not know if these issues are due to to few native gamma ramp measurement in DCCS (IMHO, here is DCCS fault) or they are an issue due small number of nodes in LUT3D (dell and other manufacturers only tell us LUT3D bitdepth/node, not actual node density, how many nodes, nodes=9x9x9? each node  bitdepth=14bit x 3). If the later is true (low density LUT3D cube), then sRGB or AdobeRGB factory calibrated modes should show same behaviour (grey oscilations) which is not what I see in my unit. Hence, IMHO it's DCCS fault but Dell & Xrite seem that will never fix it.

These grey oscillations, if preset on your system after DCCS calibration could be easily fixed with your AMD 6850. Just calibrate again in ArgyllCMS/DipscalGUI (GPU LUTs). After that, no "smooth color tints" in gradients, all grey & smooth, but make sure GPU LUT is loaded with ArgyllCMS, not by Windows OS LUT loader.

Yes, you can use RG_Phosphor.edr with another GB-LEDs. Use it with ArgyllCMS free software suite. It will translate EDRs to CCSS text file with several RGBW spectra at 1nm resolution. The last four are RGBW spectral samples for GB-LED if you wish to make a "clean version".

Actually, **ALL** GBLED  NEC spectrawiews use **THE** SAME RG_phoshor correction bundled in Basiccolor/Spectraprofiler.

You can even use this CCSS tranlated correction to "fix" Spyder4 family lack of GB-LED support, but only with ArgyllCMS. If you have a GB-LED spectraview and a Spyder4/Ellite it will never hardware callibrate it properly, it cannot measure it properly. In best case scenario it could try to apply its low resolution RGBLED correction, which is very likely to give you a pink tint in white.

Thanks a lot for your reply.  Hope to get some more advices from you.

I am looking for a light weight notebook for photography editing. I will probably consider the new MacBook or the Surface Pro 4 which will release in summer. I wish I can find a display that covers 99% sRGB gamut.  Do you have any recommendation?  What is the calibration target for the notebook with just a brightness control? D65, gamma: 2.2 / 2.4 / sRGB?

Thanks

If laptops native white point is "close" to D65 and has very good contrast (>900:1) like HP Spectre, Dell XPS, Toshibas, Macbook PRO -not air- and FullHD (or even more) IPS-like screen, I'll aim to D65 even loosing a few grey levels. It's very likely that will have 96-99% sRGB coverage. If you find one, some of these models have matte screens which are better than glossy.

Otherwise (low contrast TN-like panels), "closest daylight white" to native or even "native" white may be good solutions since if contrast is not good you could trade too much contrast just for white.

Since all have sRGB-like gamuts, closest to native "gamma shape" (2.2 or sRGB) gamma will be good choices. It is not whise to edit AdobeRGB images in sRGB screen, work your RAWs to sRGB. Anyway, Photoshop will color manage to show image's profile correct gamma.

Try to get a hardware calibratable monitor fro laptop, your chances are 99% to get an intel iGPU/non Quadro nvidia LUTs laptop.

Could you define "close" to D65 in terms of CCT range? I really have no idea. How can I decide/calculate what is the  "closest daylight white" to native?

I quoted the following from http://dispcalgui.hoech.net/

"So if you are displaying images encoded to the sRGB standard, or displaying video through the calibration, just setting the gamma curve to sRGB or REC 709 (respectively) is probably not what you want! What you probably want to do, is to set the gamma curve to about gamma 2.4, so that the contrast range is expanded appropriately, or alternatively use sRGB or REC 709 or a gamma of 2.2 but also specify the actual ambient viewing conditions via a light level in Lux, so that an appropriate contrast enhancement can be made during calibration. If your instrument is capable of measuring ambient light levels, then you can do so."

It suggests to use "2.4" or "2.2/sRGB with Lux value" for the sRGB gamma curve. Is this what I should do or should I just use 2.2/sRGB without Lux value. (I am using i1 Display Pro). I am very confused. Does the gamma curve only affect non-color managed applications? So color managed applications will always display correct gamma regardless of the gamma curve.

You said it is not good to trade too much contrast for the white. I though all displays have more contrast than they need. The required contrast for sRGB and adobe RGB is 400:1 and 288:1, so any display should be capable of that even after the calibration adjustment, right?

I don't understand why the adobe RGB has a much higher white point luminance but ambient illuminance is much lower comparing to sRGB. The display is too bright in a dim environment, isn't it?

sRGB (www.color.org/.../srgb.xalter)

White point luminance: 80 cd/m2

Black point luminance: 0.2 cd/m2

Ambient illuminance: 64 lux

adobe RGB (www.color.org/.../adobergb.xalter)

White point luminance: 160 cd/m2

Black point luminance: 0.5557 cd/m2

Ambient illuminance: 32 lux

Would you follow this standard when you calibrate you monitor?

Is there any benefit to use this updated sRGB v2 'black scaled' profile as color space?

www.color.org/srgbprofiles.xalter

Many Thanks

Gamma 2.2/srgb/L* and such are defined as curve "shape" to color management engines.

So no matter what is your actual contrast or cd/m2 at usual values (forget extreme values), a black (0,0,0) to white (255,255,255) linear gradient extended over an image will retain "curve's shape". I mean how fast greys darken near black, how fast greys light up to white.

That  "shape" is what PS/GIMP and others color manage against your monitor's profile "gamma shape" (TRC, tone response curves).

Keep in mind that you monitor black has more than 0cd/m2 luminance, finite contrast, and ICM profile TRC could take your black as "true ideal black" in order to simplify color management. Some kind of idealized gamma response. This approach is the most common and wisest.

DCCS, ArgyllCMS, Basiccolor... tries to calbrate your monitor to mimic target gamma in that way (shape, black to white grey disposition in a linear gradient) and usually is not a good advice (and DCCS cannot do that) to enable "ambient light measurements" because it's very likely that ICM profile TRC will store "actual shape" (the deformed one adjusted to your ambient) instead of the "nominal shape" (2.2, sRGB, L*) so all color managemet it's very likely that will go wrong.

So, if you want sRGB gamma shape (dark greys lighter than 2.2, light greys darker than 2.2) you should use "sRGB gamma" no matter what dispcalGUI says. DCCS sRGB gamma works the way I told you, and sRGB factory calibrated modes from these dells come with sRGB gamma "shape" so a gradient black to white looks "the same" in all brightness settings.

In non color managed enviroments, ambient light aware calibration is useful (video).

Do gradients, you'll see "curve shape" in blacks to greys to whites dispositions on that gradient. Let's asume that your IPS laptop has a near 2.2 gamma nad your OS has a profile linked to your laptops screen storing that 2.2 gamma info. Open PS, new 1024px wide image. Make a right to left black to white gradient in sRGB. New image, adobeRGB profile (2.2), make same gradient. New image, eciRGBv2 (L*), make gradient. ProPhotoRGB (1.8), make gradient...

Compare them. Compare how fast greys darken, hot fast greys become white in each of them.

That is what your "gamma target" will aim to. Once your monitor "non color managed" response is calibrated to one of these shapes, and that shape is stored in ICM profile's TRC, Photoshop will deform RGB numeric values in an image in order that your monitor response matches image profile's TRC. So choice what you like most (2.2, sRGB). Usually 2.2 target is more useful than sRGB (more "blackish" dark greys) in a computer (most video players are not color managed, and no game is color managed).

There are other gamma targets (medical imaging) that aim for actual luminance levels (cd/m2) of each grey in order to mimic inverse of human eye response, but they need a very very good bightness screen uniformity and it's very unlikely you'll need for photo use.

L* gamma for photo tries to model that inverse eye response behaviour in a "white point luminance independent setting". L* does not care of your cd/m2 at white, just shape, just how fast grey light up across a gradient (low gamma, ligher dark greys near black, hight gamma darker light greys near white).

Forget CCT, CCT is not a valid mean to point to an actual white in a monitor, you need another coordinate to point if that bluish to yellowish CCT  white is moved towards green, pink or remains white.

So just run DispcalGUI and run "uncalibrated screen report" on laptop screen. It will plot 4 correlated temperatures and their distances to "daylight whites" (all of them from cool to warm) or to "blackbody whites" measured in deltaE2000 (how far towards green-red). The lower the delta to daylight the better.