- This topic has 50 replies, 4 voices, and was last updated 4 years, 2 months ago by
.
-
Posts
-
So, if I make a CCSS correction with i1Studio with the monitor OSD set to “Custom Color” (full gamut) and then calibration in “Photo” setting with 6500K white balance using i1DisplayPro and with monitor set to “Color Space” >> “Adobe RGB”, that is the closest I can get to the perfect AdobeRGB calibration without Windows and without hardware calibration using DUCCS, right? I just need to be sure to use correction file with 3nm steps.
Do you have any advice for me? I appreciate your help.
Calibrite Display Pro HL on Amazon
Disclosure: As an Amazon Associate I earn from qualifying purchases.I can imagine works where you need that monitor itself is almost a match for sRGB, or Rec709 g2.2/2.4 or P3 D65… but I cannot imagine situations where you’ll need AdobeRGB exact match.
But IDNK how bad behaved is your display in Standard/COlorTemp/CustomColor mode. sRGB/AdobeRGB factory modes are usually OK on grey neutrality… so if they are good and Custom Color mode grey is very bad… maybe it is useful to do what you want.
Otherwise it’s easier to go to Custom Color mode, set whatever white you want and use Full Native Gamut. Photoshop will have no problem showing proper colors in AdobeRGB images, maybe banding due to Photoshop “color engine” to display truncation…but this is not caused 100% by calibration but due to TRC mismatch. If you worry about this a GPU that causes no banding and a matrix profile + single curve, even powerlaw TRC will minimize those issues at the expense of accuracy. Validate such “idealized” profile with DisplayCAL, and if it’s OK you can use it.My girlfriend professionally designs vector graphics in sRGB and I edit my pictures shoot in Adobe RGB from time to time. We do not use Photoshop as it is not available for Linux. The biggest upside (or downside depending on the perspective) is that colour profiles in Ubuntu are system-wide, therefore good calibration works in every single native application from file manager, web browser, office suite up to photo-edit software and vector software, but must be quite universal.
We found that calibration in “Custom Color” (full gamut) OSD mode often results in oversaturated colours in many applications and I suspect calibration in “Color Space” >> “sRGB” OSD mode will limit colours visible on my pictures shoot in Adobe RGB. That is why I was looking for the best possible accuracy, but gamut limited to the Adobe RGB.
Also UP2718Q is a “high-end low cost monitor” or “low-end high cost monitor” however stupid it sounds. It is way worse than truly professional EIZO monitors, but at the same time it is better than probably 98-99% of monitors used by other people. We, especially my girlfriend, do not want to see colours or shades which are beyond Adobe RGB/sRGB as most people would not be able to see them correctly, even after calibration. Do I think right?
I understand that you recomend to try calibration in “Custom Color” (full gamut) OSD mode and bring RGB gains manually and see what happens, right?
Charlesss, I’m interested in your experience because as I said a bit ago I just stared running the older siblings of the UP2718Q
According to a pair of UP251Ds which I have measured backwards and forwards, albeit with a dated instrument, Dell’s certified Adobe RGB and sRGB modes deliver what they claim. You can use them out of box, and find proper TRCs, excellent gray tracking, exact gamut limits, and color temp within a couple hundred K of standard.
If they don’t look right return the unit because it should look great! Based on Dells cert and what I’ve confirmed by measurement, it should be dead on. For example, display EDID primarys match measured within a hairs width. Same for Adobe RGB and sRGB gamuts matching spec.
And uniformity should be acceptable w/ UC off. And improve with it on. Caveat below.
UC loses have your contrast ratio, which probably doesn’t matter, and provides a uniformity benefit. Buuut…
On UP2516D, UC and any color reference mode except Standard mode are mutually exclusive. UC On (“Calibrated”) appears to be “Standard” mode, but while UB implies Standard, Standard does not imply UC.
Check in the menu by setting Adobe RGB and looking at UC in Display menu. If it’s grayed out, then you can have one or other not both.
Vincent hopefully will correct / clarify the following thoughts
Re Linux, “system wide” calibration makes sense–that is, what gets load into the vid card gamma table (VCGT), but the actual handling of the profile has to be done on a program-by-program basis. So if the program is capable, like Gimp, then good to go with gamut limiting in a WGG mode. In other progs you will have over saturation per your larger display space in Custom Color mode, Normal mode (which is full gamut), Adobe RGB and DCI reference modes.
But good news, Dell Adobe and sRGB reference modes probably work great out of the box. So just use them. The trick will be flipping the SW profile back and forth with the Dell setting, Look for a utility to help, or make one yourself… Welcome to Libre! Haha, I’m sorry… This is stuff is always more convoluted than you hope.
Custom Color is useful to dial white and black in before profiling if you are using color managed apps and you want a one setting fits all. For example, I measured and tweaked this pair in Custom Color mode, profiled with an sRGB TRC and everything else as measured, and now can run Adobe on Mac with an alignment that does everything from Adobe, DCI, sRGC, in one alignment. Great! And for programs that fall through the cracks, at least the TRC is gracefully sRGB.
As to DUCCS, I see no advantages because I’m getting the full 8-bits of linear response without clipping or contour (banding). So what can a DUCCS alignment offer??
Verification comes in well within 1 dE2000.
Question for you, why do you use both an i1Display and i1Pro? And why worry about correction for i1Display, I thought that instrument was supposed to be able to handle this tech??
UP2516D is GB-r panel BTW, as is UP2718Q if you can believe this site:
https://www.displayspecifications.com/en/model/f4aec13
As pertaining to minor tracking errors in green due to spectral bump, or a squiggle on a uv plot, seems like no concern given your work… But I would like to hear more
At a practical level, your display is 99% of a high-cost pro unit. In my case, it’s 99% of the color for 1/4 the bucks.
Yours
/wire
My girlfriend professionally designs vector graphics in sRGB and I edit my pictures shoot in Adobe RGB from time to time. We do not use Photoshop as it is not available for Linux. The biggest upside (or downside depending on the perspective) is that colour profiles in Ubuntu are system-wide, therefore good calibration works in every single native application from file manager, web browser, office suite up to photo-edit software and vector software, but must be quite universal.
We found that calibration in “Custom Color” (full gamut) OSD mode often results in oversaturated colours in many applications and I suspect calibration in “Color Space” >> “sRGB” OSD mode will limit colours visible on my pictures shoot in Adobe RGB. That is why I was looking for the best possible accuracy, but gamut limited to the Adobe RGB.
Then they are not color managed and that particular windows manager for ubuntu is not applying any color management system wide. It is “Publishing” display profile system wide and color managed apps “ask” OS about default display profile, in the same way as windows. If app does not “ask”, no color management => oversaturated images “supposed” to be rendered most likely as sRGB.
In that situation if DUCCS preconditioning is not possible (like a 2nd old computer to run DUCCS, then copy ICM v2 profiles to your Linux), “safest” choice is to use 2 OSD modes and one profile for each mode.
Default mode should be sRGB with a DisplayCAL calibration at certain OSD brightness (mode Dell monitors share brightness control across OSD modes), Displaycal will correct gamma and white using GPU.
Then when YOU want to work with some color managed editor (otherwise it’s piontless) switch OSD mode to Custom Color or factory AdobeRGB at same OSD brightness settings (it’s shared across OSD modes, IDNK about THAT particular model) and use another ICC porfile made with DisplayCAL for THAT osd mode.
When your gf want to use co,mputer for general purpose or non color managed vector editor, switch back to sRGB mode AND change default profile to the one made for sRGB OSD mode.Display OSD mode and profile made from that OSD mode (with GPU calibration or without GPU cal like un DUCCS) should match. This way color managed apps work as intended.
If you can use an older computer to run DUCCS approach is the same CAL1 for your gf (and label that profile with a significative name like CAL1_sRGB.icm), CAL2 at native gamut or AdobeRGB emulation and another significative name for profile. Calibration is stored in CAL1/2 lut-matrix-lut, in monitor, you need to copy ICM profiles to your Linux OS and “publish” them (make one default) in the same way you do in windows or other OSes
Also UP2718Q is a “high-end low cost monitor” or “low-end high cost monitor” however stupid it sounds. It is way worse than truly professional EIZO monitors, but at the same time it is better than probably 98-99% of monitors used by other people.
I think 2nd one is a better match. You ned to understand that the bigger the gamut, the more noticeable are color uniformity errors across screen. Money you pay for a CS goes to that kind of QC.
Also TRC differences are more likely to end in banding and AFAIK there is no image editor in Linux that can overcome that like Lightroom /CaptureOne does… or photoshop with a 10bit supported card & monitor.
That’s why *I* recommend a 27″ QHD Eizo CS widegamut monitor as an starting point that suits even most pros rather than UHD 27″/32″ widegamut adventures with Benq and Dell and Asus and those kind of friends. Issues in their software are another can of worms, specially for illustrators who need pristine gradients.
CS is even cheaper than that Dell and its expected to have under 2deltaC (color stain, tints, not talking about brightness drop) uniformity across all screen with uniformity compensation off.We, especially my girlfriend, do not want to see colours or shades which are beyond Adobe RGB/sRGB as most people would not be able to see them correctly, even after calibration. Do I think right?
As said before, sRGB makes sense, to play a “safe” card and to have a sRGB (or sRGB gamma 2.2) calibration on screen & gpu regardless of app support for color management.
AdobeRGB no, you’ll need a color managed app for using it as intended. If you use Custom color OSD mode, and a profile made for it, adn you see oversaturated reds while working on an AdobeRGB image… then that app is NOT working as intened and you should not use it, or check if you configured it wrong. Keep in mind that the same situation can happen you in AdobeRGB mode with cians and greens… but maybe it’s subtle for that image and goes partially unoticed… but that app is not working OK and you should avoid it.I understand that you recomend to try calibration in “Custom Color” (full gamut) OSD mode and bring RGB gains manually and see what happens, right?
I recommend 2 GPU calibrations, one for “fixing” factory sRGB oSD mode, if it needs to be fixed. Then another at native gamut in custom OSD mode for color managed editors, and you need to select its profile before you start that color managed image editor. Once you end your work, revert to sRGB mode and set its profile as default.
Or even better, do the same with DUCCS and another old computer, copy ICM profiles, validate with DisplayCAL. If they are OK, great. If tehy are not (due to mild to severe DUCCS limitations, but benq is even worse), you can try to fix white and grey i GPU with DIsplayCAL in the same way you may do with factory AdobeRGB mode.@Wire Well, maybe you have a wonderful unit or older model was better. UP2718Q has a very wide gamut and perfectly covers sRGB, Adobe RGB and DCI-P3, but brightness uniformity is terrible. On low brightness it so poor it actually annoys even in normal office work.
Without calibration the white point is over 6500K and more like 7000K which is also very annoying: too cold for printing, too cold for softproofing and even for web publication. The worst part is that i1Studio and i1DisplayPro indicate ΔE between 2-3 before calibration which makes the reference modes almost useless without calibration.
I am looking for a Linux-compatible solution to reduce ΔE to minimum and avoid oversaturated colours.
As to DUCCS, I see no advantages because I’m getting the full 8-bits of linear response without clipping or contour (banding). So what can a DUCCS alignment offer??
Gamut emulation, GPU independent “chance” of smooth gradients, several white points at several gamut emulations.
Chance because it depends on display QC, DUCCS may have hard times trying to correct severe grey issues, while ArgyllCMS on a GPU with high bitdepth luts and dithering shoould be able to fix it.Mr Gill, ArgyllCMS creator owns a Dell UP like you two and a few weeks ago he noticed that there is a public SDK from Dell to manually upload lut-matrix-lut HW correction to screen , the same way DUCCS does to upload calibration to monitor. Dell SDK (and HP for Z27x) have versions for win/macos/Linux… and almost common data types so maybe there is a chance for you.
ArgyllCMS has a mailist, since ArgyllCMS is donation powered… you may want to show interest in that feature in mailist, ask about it, if it is possible and even push it a little with donations. Just a hint, ask there…imagine AdobeRGB custom calibration gamut emulation on Linux, worth asking and donation if it’s possibleAnother option is a CS, open Color Navigator, “click next, click next”, calibrate and profile and do you work with images.
Then they are not color managed and that particular windows manager for ubuntu is not applying any color management system wide. It is “Publishing” display profile system wide and color managed apps “ask” OS about default display profile, in the same way as windows. If app does not “ask”, no color management => oversaturated images “supposed” to be rendered most likely as sRGB.
Well, maybe I am wrong, but when I add and enable ICC profile in GNOME settings all applications change colours accordingly. I also believe my picture editing software Darktable has full colour management, but of course might be wrong.
In that situation if DUCCS preconditioning is not possible (like a 2nd old computer to run DUCCS, then copy ICM v2 profiles to your Linux), “safest” choice is to use 2 OSD modes and one profile for each mode.
Default mode should be sRGB with a DisplayCAL calibration at certain OSD brightness (mode Dell monitors share brightness control across OSD modes), Displaycal will correct gamma and white using GPU.
Then when YOU want to work with some color managed editor (otherwise it’s piontless) switch OSD mode to Custom Color or factory AdobeRGB at same OSD brightness settings (it’s shared across OSD modes, IDNK about THAT particular model) and use another ICC porfile made with DisplayCAL for THAT osd mode.
When your gf want to use co,mputer for general purpose or non color managed vector editor, switch back to sRGB mode AND change default profile to the one made for sRGB OSD mode.DUCCS is not an option unfortunately. I always calibrate for 25% brightness (approx. 120-122 cd/m²). I understand why you suggest calibration and profile for “Custom Color” OSD mode (full gamut), but why sRGB over Adobe RGB? In the sRGB colours are very washed out and I mean very. “Custom Color” OSD mode give extra-saturated (oversaturated?) colours, sRGB on the other hand gives very washed-out colours even in a web browser, unnaturally washed-out if I am correct.
Display OSD mode and profile made from that OSD mode (with GPU calibration or without GPU cal like un DUCCS) should match. This way color managed apps work as intended.
If you can use an older computer to run DUCCS approach is the same CAL1 for your gf (and label that profile with a significative name like CAL1_sRGB.icm), CAL2 at native gamut or AdobeRGB emulation and another significative name for profile. Calibration is stored in CAL1/2 lut-matrix-lut, in monitor, you need to copy ICM profiles to your Linux OS and “publish” them (make one default) in the same way you do in windows or other OSes
Also UP2718Q is a “high-end low cost monitor” or “low-end high cost monitor” however stupid it sounds. It is way worse than truly professional EIZO monitors, but at the same time it is better than probably 98-99% of monitors used by other people.
I think 2nd one is a better match. You ned to understand that the bigger the gamut, the more noticeable are color uniformity errors across screen. Money you pay for a CS goes to that kind of QC.
Also TRC differences are more likely to end in banding and AFAIK there is no image editor in Linux that can overcome that like Lightroom /CaptureOne does… or photoshop with a 10bit supported card & monitor.
That’s why *I* recommend a 27″ QHD Eizo CS widegamut monitor as an starting point that suits even most pros rather than UHD 27″/32″ widegamut adventures with Benq and Dell and Asus and those kind of friends. Issues in their software are another can of worms, specially for illustrators who need pristine gradients.
CS is even cheaper than that Dell and its expected to have under 2deltaC (color stain, tints, not talking about brightness drop) uniformity across all screen with uniformity compensation off.I know that wider gamut might actually be a problem in some cases. The issues is we actually needed 4K resolution for other work and 4K EIZO CS is very expensive and ugly as hell. It would also be waisted as we can run only in 8-bit mode.
As said before, sRGB makes sense, to play a “safe” card and to have a sRGB (or sRGB gamma 2.2) calibration on screen & gpu regardless of app support for color management.
AdobeRGB no, you’ll need a color managed app for using it as intended. If you use Custom color OSD mode, and a profile made for it, adn you see oversaturated reds while working on an AdobeRGB image… then that app is NOT working as intened and you should not use it, or check if you configured it wrong. Keep in mind that the same situation can happen you in AdobeRGB mode with cians and greens… but maybe it’s subtle for that image and goes partially unoticed… but that app is not working OK and you should avoid it.I understand that you recomend to try calibration in “Custom Color” (full gamut) OSD mode and bring RGB gains manually and see what happens, right?
I recommend 2 GPU calibrations, one for “fixing” factory sRGB oSD mode, if it needs to be fixed. Then another at native gamut in custom OSD mode for color managed editors, and you need to select its profile before you start that color managed image editor. Once you end your work, revert to sRGB mode and set its profile as default.
Or even better, do the same with DUCCS and another old computer, copy ICM profiles, validate with DisplayCAL. If they are OK, great. If tehy are not (due to mild to severe DUCCS limitations, but benq is even worse), you can try to fix white and grey i GPU with DIsplayCAL in the same way you may do with factory AdobeRGB mode.I do not understand one thing, why do you recommend sRGB calibration target and OSD mode over Adobe RGB target and OSD mode? Colours in sRGB OSD mode look very washed-out, unnaturally and incorrectly washed-out if I am correct.
As to DUCCS, I see no advantages because I’m getting the full 8-bits of linear response without clipping or contour (banding). So what can a DUCCS alignment offer??
Gamut emulation, GPU independent “chance” of smooth gradients, several white points at several gamut emulations.
Chance because it depends on display QC, DUCCS may have hard times trying to correct severe grey issues, while ArgyllCMS on a GPU with high bitdepth luts and dithering shoould be able to fix it.Mr Gill, ArgyllCMS creator owns a Dell UP like you two and a few weeks ago he noticed that there is a public SDK from Dell to manually upload lut-matrix-lut HW correction to screen , the same way DUCCS does to upload calibration to monitor. Dell SDK (and HP for Z27x) have versions for win/macos/Linux… and almost common data types so maybe there is a chance for you.
ArgyllCMS has a mailist, since ArgyllCMS is donation powered… you may want to show interest in that feature in mailist, ask about it, if it is possible and even push it a little with donations. Just a hint, ask there…imagine AdobeRGB custom calibration gamut emulation on Linux, worth asking and donation if it’s possibleAnother option is a CS, open Color Navigator, “click next, click next”, calibrate and profile and do you work with images.
That a great piece of advice, it would be really worth to add full Dell monitors support into AgryllCMS. Love or hate, they are quite popular and if we can extract a bit more from them, that would be great.
You keep recommending EIZO CS. Does it make any sense to buy them to run them in 8-bit mode only (Intel iGPU limitation) and are their built-in colorimeters supported in DisplayCAL?
I do not understand one thing, why do you recommend sRGB calibration target and OSD mode over Adobe RGB target and OSD mode? Colours in sRGB OSD mode look very washed-out, unnaturally and incorrectly washed-out if I am correct.
I do recommend full native gamut over AdobeRGB because its more versatile. eciRGBv2 images can be encoded and shown in 8bit. There are flowers, printable, outside AdobeRGB.
Also I said that if app color look oversaturated in red when in full native gamut (Custom OSD mode) then.. that app is not color managed. It’s mosly like a smoking gun, that apps at most forces you to have an screen matching some default colorspace, that is not a color managed app.
sRGB mode is an ADDITIONAL failsafe configuration to use in those poorly color managed applitaions (or wrongly configured by user, which IDNK). It’s like Windows.
You may notice cnahed when you apply profile because grey calibration in GPU changes things; white & gamma… but that alone is not color management. Usually it’s just grey calibration + ICC “publising for those app who ask”
As to DUCCS, I see no advantages because I’m getting the full 8-bits of linear response without clipping or contour (banding). So what can a DUCCS alignment offer??
Gamut emulation, GPU independent “chance” of smooth gradients, several white points at several gamut emulations.
Chance because it depends on display QC, DUCCS may have hard times trying to correct severe grey issues, while ArgyllCMS on a GPU with high bitdepth luts and dithering shoould be able to fix it.Mr Gill, ArgyllCMS creator owns a Dell UP like you two and a few weeks ago he noticed that there is a public SDK from Dell to manually upload lut-matrix-lut HW correction to screen , the same way DUCCS does to upload calibration to monitor. Dell SDK (and HP for Z27x) have versions for win/macos/Linux… and almost common data types so maybe there is a chance for you.
ArgyllCMS has a mailist, since ArgyllCMS is donation powered… you may want to show interest in that feature in mailist, ask about it, if it is possible and even push it a little with donations. Just a hint, ask there…imagine AdobeRGB custom calibration gamut emulation on Linux, worth asking and donation if it’s possibleAnother option is a CS, open Color Navigator, “click next, click next”, calibrate and profile and do you work with images.
That a great piece of advice, it would be really worth to add full Dell monitors support into AgryllCMS. Love or hate, they are quite popular and if we can extract a bit more from them, that would be great.
You keep recommending EIZO CS. Does it make any sense to buy them to run them in 8-bit mode only (Intel iGPU limitation) and are their built-in colorimeters supported in DisplayCAL?
-uniformity, a superb advance.
-GPU independent bandless grey calibration
-CS do not have colorimeter, you need to buy an i1DisplayPro. CGs have internal colorimeter (limited in 2 newer and cheaper models) and they are not supported by ArgyllCMS AFAIK, but they can use i1d3. CGs intenal colorimeter are just for fast preconditioning from a extremely good starting point: a few measures are enough to correct gamma if grey is OK. For a extremely accurate caracterization (“make a table profile”) that can be used for a LUT3D creation, an i1d3 or better device is mandatory… just for speed: imagine time cost going from 10 steps per cube side to 17So I would choose a reliable & uniform 27″ QHD at 1000 euro rather than a 1500 euro 27″ UHD dell, or even worse, a benq (at east Dell uses “close” spectral corrections to their display actual backlights). YMMV.
IDNK how much would cost CS2740, 27″ UHD?… I think should be between 1500-2000 euro. 0% to 33% extra worth paying for <2dC across screen. Again, YMMV…
PS. If grey is extremely good… you can “cheat”: use ideal profiles with idealized TRC. If monitor is good enough to track it, displayCAL will validate it OK and colro managed apps will “simplify” things so even at 8bit the chances of color managed induced banding are less likely to happen.
If your Dell UP after caibration behaves that good… you can do it too, just choose a simple profile type like matrix + SINGLE curve (shared TRC even if they are not equal upon measurement but “mostly equal” visually)Re bad uniformity… Does enabling UC make brightness uniformity acceptable?
The effect of display uniformity on happiness with a unit is very subjective, but one thing everyone knows is when something just looks bad, e.g., blotchy. You get a bletch reaction, like the iPhone 11 Pro camera lens that make some people go ick!
If it’s quite bad, do engage Dell support to complain. It might not help you but they need to hear reasonable feedback and get it into their process. And if you are fair toned with them, they are helpful with returns covering shipping. Take a photo that conveys represents what you see and include it with your comments of being disatisfied.
I took a chance at this price-point knowing that Dell has a probably-deserved reputation problem with panel QC not because they suck at it, but because the nature of the tech is difficult to control and they tend towards lowering prices (happily to me). I’ve owned 4 pairs of Dell LCD over 16 years and none have been perfect, but they’ve all been very good and useful. And all still work. But I’ve read of plenty of real problems with some models. Same is true for every other tech Take tube TVs in 80s and 90s. The tech was well understood, but things could and did go went wrong with several variables of design / unit QC. I feel pain about designs that are just bad even though they tried really hard. I went through 3 generations of Panasonic home theater projectors before I had to admit that company can not build a device that lasts. Now they no longer sell consumer stuff. Got a Sony in 2016 and this thing has gorgeous color and is looks perfect after 4 years, just changed the lamp after 4000 hours as it started to flicker and like new.
I’ve measured uniformity on these two Dell UP2516D and with UC off, it drops by about 6.5 dE from center to left/right edge, with UC on about 1.8 dE for one unit, and about 3 dE for the other. Only brightness change, color looks stable. There is some very slight color gradation with UC enabled and a pure white field. For my photography, these errors matter not at all, and have zero affect on task.
Do I wish it were perfect?! Yes. But my eyes are far from perfect too, so what’re-ya-gonna-do.
For the money, it reasonable to expect the display doesn’t make you go gah! OTOH, nothing’s perfect. And this is just as true of the super expensive gear.
A suggestion thats close to grinding bones and sprinkling them around your display is run the “panel conditioning” feature overnight for a couple of nites and see if that has any affect on uniformity. It’s meant to deal with dead pixels, but I have found that LCD’s generally look better as they are exercised for a while. I also use a screen saver that washes color across the face to exercise the grid. Is this just juju? IDK, but it doesn’t hurt… Do this while you are waiting for Dell Support email to say something robotically supportive. One of these UP2516D arrived with a cracked panel and they exchanged promptly with good followup so Dell support is basically helpful.
Pls forgive all my rabbling… Just having fun
/wire
Re bad uniformity… Does enabling UC make brightness uniformity acceptable?
Not on these lowcost monitors. Uniformity is improved but you loose too much: locked osd controls (in some units RGB gains, in others brightness), loose HW calibration in some (most) models, loose too much contrast for video or even “spare time activities”.
A common sRGB screen from Dell, their U series can have 1000-1200:1 @ D65, they are resasonably cheap, reasonable QC for their price and since their are sRGB uniformity (bleeding excluded) is not usually very bad and can be used out of the box.
Paying more than 1000 euro for a UP widegamut (or some Benq SW) to end at 600:1 and locked RGB gains so white has to be GPU corrected is a bad joke IMHO.
If it’s quite bad, do engage Dell support to complain. It might not help you but they need to hear reasonable feedback and get it into their process. And if you are fair toned with them, they are helpful with returns covering shipping. Take a photo that conveys represents what you see and include it with your comments of being disatisfied.
Not really, Dell support has explained their extremely poor QC requirements at factory about a few inches arround center, there was some FAQ in their support site about this a few years ago. Also not all people live in US, or they bought them through resellers.
I took a chance at this price-point knowing that Dell has a probably-deserved reputation problem with panel QC not because they suck at it, but because the nature of the tech is difficult to control and they tend towards lowering prices (happily to me).
It’s certainly not a “dell problem”. It’s a QC problem. QC problem is a price problem. Panel manufacturer like LG or Samsung can provide handpicked batchs at some price and less QC at other price.
These lesser QC if we exclude bleeding maybe good enough for common sRGB monitors, hence good reputation for U models like U2415, but in these widegamuts is magnified and red/pink or green tints on sides but close to center are common.
I do believe that panel model in a CS maybe the same as in an UP, the difference is QC.
As CS prices droped a lot since AH-IPS and GB-LED… it is now worth the risk to play low cost poor QC card.
Poor software (and DUCCS is the less worse of them) fo HW calibration is another handicap…but do as you wish.I’ll say it again, it’s not a Dell’s problem. “Dell monitors” (generic) are not bad, their U sRGB series are usually very good, or Benq PDs, excellent for their price.
Low cost widegamut are usually mediocre at best. It is a problem limited to low cost, poor QC widegamut monitors.
Hence your previous dell experience with your previous sRGB dell monitors is irrelevant on this matter, I mean it is not related with current subject. It is irrelevant too if you owned a WG CCFL like U2410 since price differences between QC equivalent with HW cal like PA241W was HUGE. Since GB-LED in 2015? that price difference went down making the saving not worth the risk IMHO
450 -> 650 euro 24″
700 euro -> 1000 euro 27″
And you gain unifomity dC, uniformity=on in all osd modes and a more reliable software.But… YMMV, not worth my money.
I’ve measured uniformity on these two Dell UP2516D and with UC off, it drops by about 6.5 dE from center to left/right edge,
no info about color, thats why DisplayCAL reports uniformity in several ways.
with UC on about 1.8 dE for one unit, and about 3 dE for the other. Only brightness change, color looks stable. There is some very slight color gradation with UC enabled and a pure white field. For my photography, these errors matter not at all, and have zero affect on task.
And you loose DUCCS so if you have a laptop… bad news.
IDNK if using Dell SDK you can modify HW LUTs white UC=on, I suppose you cannnot.
Also Dell do not care about some older GB-LED models with HW cal, they were excluded from SDK although DUCCS uses the same SDK for HW calibration on these “old” models.Vincent, what happens to a display’s usefulness when you lose 1 bit of contrast ratio? How does this manifest visibly?
Re SDK and UC I suspect you are right, and I am very curious how the make UC work
It was worth it to me to play low cost cause my work will not suffer from slight defects; and yes in states I take support more for granted. I bought direct from Dell.
* * *
Attached are some photos to give a very rough sense of what I see… Beginning with UC off and UC on
Took dispcal(1) / spotread measurements at min / max luminance in native mode:
CR holds well at 1140 to 1170
color temp 6500 to 6519, Duv 0.0056
G2.24 – G2.252nd unit pretty much same, slightly lower min lum CR at 959
These variances are as much about instrument repeatability limits as displays. But the native performance between units does visibly differ. In reference modes they look very close to the same. With custom cal they look exactly the same side by side.
At $299 US delivered, hard to complain
Photos are just a to give an idea. All the photos glitches are just iPhone artifacts
/wire
PS Why do ICpCt dEs report so poorly?
Vincent, what happens to a display’s usefulness when you lose 1 bit of contrast ratio? How does this manifest visibly?
Yes, very visible. At same white level, blacks get lifted.
For print is no problem, since when you simulate black in softproof they get lifted too to match paper. Also if your editing tools does not allow that kind of softproof, you may want to screen itself have paper-like contrast.
For all other uses it is not desireable. Even 1000:1 is to low (working in video, or “spare time activities”)
These variances are as much about instrument repeatability limits as displays. But the native performance between units does visibly differ. In reference modes they look very close to the same. With custom cal they look exactly the same side by side.
At $299 US delivered, hard to complain
I agree. It’s cheaper than old U2413 or new SW240.
AFAIK there is no 25″ equivalent with QC. NEC and Eizo go form 24″ to 27″ so we can agree that its segment market is unique and difficult to compare. In my previous posts I made 24″ and 27″ comparisons.
PS Why do ICpCt dEs report so poorly?
Look at color where that happens, ¿near black color while BPC on? Here you have your answer.
