What's in the AdobeRGB profile? (Q related to Samsung XL20/XL30 displays)

Discussion in 'Digital Darkroom' started by alex_iwonttell, Jan 9, 2010.

  1. Hello,
    Has anyone dissected the AdobeRGB profile, or knows what it actually contains?
    The reason I'm asking: I have the Samsung XL20 display, which covers some 113% of AdobeRGB, so I usually just work in this color space. XL20 offers the option to calibrate 4 profiles and save them directly into the display: one profile covers the whole gamut of the display (for massochists who love overly deep reds), the others are sRGB, Adobe RGB and 'emulation'. In emulation mode one can store any color space with a smaller gamut (say, of a printer or another display). It's also possible to specify the parameters manually (green, red, blue, white and black point, gamma, luminosity) - see screenshot so you know how it works.
    I figured I just save the AdobeRGB profile in both the dedicated AdobeRGB mode and emulation mode, with different luminosity. One can't do that directly, because trying to emulate AdobeRGB just stores the profile in the AdobeRGB mode automatically. So instead, I open the AdobeRGB ICC profile in the setup interface of the display, tweak the settings a bit so it's no longer recognized as AdobeRGB, then set the points back so it's the same as AdobeRGB. Then I save the profile for emulation.
    After this, switching between AdobeRGB and emulation modes on my display should give me exactly the same picture - but it doesn't. The emulation mode clearly has more definition in dark areas.
    This leads me to a conclusion, that the AdobeRGB ICC profile contains some more information than what can be entered manually. Can anyone confirm that? What exactly is different? My display's software recognized the AdobeRGB ICC most likely by the Adobe's copyright. Is there any way to obtain and exact copy of Adobe's ICC profile but without the copyright (only for my own purposes, to use it for calibration of the emulation mode of my display)?
    Of course it's also possible that the display or software behaves differently with different profile, but that doesn't make much sense. Also, I tried recalibrating the display (in AdobeRGB mode) with X-Rite software and there was no change, so apparently the display's interpretation of AdobeRGB is correct.
    Comments?
     
  2. All matrix profiles are very simple and describe a white point, the three chromaticity values for RGB and a TRC (gamma) setting. You can see all the values by simply selecting it in the Photoshop Color Settings and clicking on “Edit” where a dialog will appear with all the values above.
    That said, there’s about zero reason to mess with any of the settings (after which, its not Adobe RGB any more). Nor does the working space have to have anything to do with the display, display calibration etc. The display and the editing of images is totally independent and has been since Photoshop 5. Calibrate and profile your display and move on.
     
  3. Well, it does have have something to do with my display... Some Eizo displays use the same philosophy. Having the display take care of most stuff is quite a good idea actually. Want to see how it's going to look on a different display or in sRGB color space? Just press a button.
    But I'd like to know whether the AdobeRGB ICC profile is any different than what I understand and what you describe (a simple profile with just a few points defined). I guess I'll have to contact Samsung to know.
     
  4. Well, it does have have something to do with my display...​
    No, it doesn’t. If it did, if your display produce exactly Adobe RGB (1998) (which is a synthetic, solely mathematical color space) you wouldn’t have to messing around with Colorimeters, software and building profiles. Since Photoshop 5, the display conditions and the working space have been divorced. You calibrate and profile your display, a profile defining that condition exists, allowing you to work in any color space (RGB or CMYK), Display Using Monitor Compensation is used to produce the correct preview of that color space upon your display using that display profile.
    Many applications are not ICC aware. They look awful with a wide gamut display. So your unit, the Eizo and my NEC allow one to switch to an sRGB behavior (not really, its a simulation and unless you have something like the HP DreamColor, you can’t alter the actual chromaticity of the display). If you see butt ugly previews in say a non ICC aware web browser because you’re dealing with a wide gamut unit, you set it to simulate sRGB and it doesn’t look awful. But its not color managed at this point and in a way, it doesn’t matter because in ICC aware applications, as long as the document has an embedded profile and you have a profile that defines your display, you’ll get proper previews.
    Best to start here:
    The Role of working spaces in Adobe applicaitons
     
  5. Andrew, I think I know what color spaces are. Let me repeat what some displays do: they let themselves be calibrated to a particular colorspace/profile and save that setting *directly* into the display - so the display acts like a true AdobeRGB device (or some other emulated device). That means you don't need to specify a monitor profile, or just use AdobeRGB ICC profile as a descriptor of the display. Of course it's only a preview and doesn't change the actual color space you work in, I didn't say that. But that preview is calibrated.
    It may sound weird to industry veterans, but it works - it's basically the same like changing output profiles in the application, only it's done inside the monitor (including using emulation of other color spaces). Perhaps not 'mathematically' correct (which it can't be anyway, since are eyes are analogue), it works very well.
    The discussion is just academical anyway. Color spaces are not perfect things, they just try to override imperfections in devices.
     
  6. Let me repeat what some displays do: they let themselves be calibrated to a particular colorspace/profile and save that setting *directly* into the display - so the display acts like a true AdobeRGB device (or some other emulated device).​
    I don’t believe that’s at all correct. The Adobe RGB (and all RGB working spaces) have very defined reference media (the ambient light around this theoretical space) and there’s simply no way, especially with a device yo claim is 113% of Adobe RGB (a spec itself that’s incorrect at least based just that spec) can produce Adobe RGB (1998) anymore than a device that’s 110% of sRGB can produce Adobe RGB (1998).
    That means you don't need to specify a monitor profilE.​
    That’s ridiculous and incorrect. Without a display profile, all ICC aware applications can’t function correctly. In fact, in the absence of a good, ICC profile built of your display, the OS will make an assumption otherwise and use that generic profile for all previews. Until you understand Display Using Monitor Compensation in ICC aware application and the role of both display and document profile, you’ll never move forward in understanding the process.
    The discussion is just academical anyway.​
    No, it over because you asked a question based on incorrect assumptions and when corrected, keep ignoring the facts of the color science. So at this point, do whatever you wish to make your display or color management system appear to you to do what you wish.
    Color spaces are not perfect things they just try to override imperfections in devices.​
    That’s a nonsensical statement. You’d be betters served by reading at the very least, the URL provided above.
     
  7. Andrew, you still assume I don't understand the difference between the working color space and monitor profiles. Your assumption is based on my simplistic sounding question.
    So again. My question was, what's actually included in the AdobeRGB1998.icc file? I didn't ask about color spaces, theoretical definitions in the white papers or differences between the color spaces for work and viewing. I really just want to know what specifications are included in THAT ONE FILE.
    I'd check myself but there aren't any free icc file editors. I've read there is one for Linux, but it can't open that AdobeRGB1998.icc file because - get this - it's not according to spec! Does my question still sound so nonsenical?
    Again, I understand the differences and the usual calibration workflow. The tech I speak about has nothing to do with usual monitors and their 'modes', although it sounds like it.
     
  8. Andrew, you still assume I don't understand the difference between the working color space and monitor profiles.​
    Did you read the article above? If so, I think you might want to retract that statement.
    I told you what’s in the Adobe RGB (1998) profile. It is both a working space and a color space. You’d best serve yourself and others here if instead of typing, you did a little reading first. I have plenty of other PDFs you can use to educate yourself but start at the beginning with the URL I posted above.
     
  9. Well, tell me what's so wrong about my scenario:
    There's a display with wider gamut than AdobeRGB. It can be calibrated so it shows the exact specs of AdobeRGB (excl. outside conditions, like lighting). That means, you can just use AdobeRGB1998.icc as the monitor profile (or none, if the app assumes the monitor profile is the same as the working space, like some do). The display handles the rest, incl. clipping the colors not contained in the AdobeRGB specs. And - again, it's calibrated.
    What's so wrong with that? What would you with this? What profiles would you assign to the display? Would you have the need to calibrate with another tool just because it's more traditional, even if it gives the same results? Or what?
    And yes, I've actually read your paper and didn't learn anything new. My problem is practical, not theoretical, while you still assume I'm an ignorant idiot.
     
  10. BTW if you're curious why I use the ARGB mode of the display instead of its whole gamut, that's because it looks fugly and unnatural (reds too deep in comparison with other colors). So I just use the gamut clipped to ARGB, calibrate and use colormanaged apps.
     
  11. That means, you can just use AdobeRGB1998.icc as the monitor profile (or none, if the app assumes the monitor profile is the same as the working space, like some do).​
    That’s ridiculous and unnecessary.
    My problem is practical, not theoretical, while you still assume I'm an ignorant idiot.​
    Its hardly practical and the more you write, the more I tend to agree with your second statement above.
     
  12. Whatever. One thing I hate are "industry professionals" who can't comprehend simple questions outside their narrow scope, like it's obviously your case. See you in ten years when color aware devices are standard. For now, thanks for repeating obvious things I didn't ask 10 times over. Why couldn't you just spend ten seconds thinking about my questions instead of 10 minutes typing the same thing is beyond me, just don't spend anymore with either please.
    If anyone else is willing to help with my question even if they aren't Adobe betatesters or whatever, or know how internals of Samsung displays work, I'll be glad.
     
  13. Can someone please make up some rubbish to make Alex happy....
     
  14. You really are a dick.
     
  15. If there's a moderator on this forum, I'd like to have this topic removed please. It doesn't help anyone, especially not me, because of this smartass.
     
  16. You really are a dick.
    If there's a moderator on this forum, I'd like to have this topic removed please. It doesn't help anyone, especially not me, because of this smartass.​
    You’ve got to be kidding! Yes, the moderator should ban you from the forums for such comments!
     
  17. I apologize for the dick part, I'm an Eastern European and I lose temper easily. But seriously, if you can't reply to a question because the practice described is totally different than what you're used to, you should either try to interpret it correctly or just ignore it. Or, if you have useful information to add or correct, you can do just that instead of posting completely unrelated generic stuff nobody asked for.
    My loss of temper is just because of my disappointment we can't communicate correctly. Either because of me not using the correct technical terms, or you not trying to understand some particular problem you're not aware of. Whatever.
     
  18. Apology accepted. But I see no reason to go any further.
    Your main issue stems from these statements:
    L20 offers the option to calibrate 4 profiles and save them directly into the display: one profile covers the whole gamut of the display (for massochists who love overly deep reds).​
    There’s nothing in terms of the differing color spaces that should warrant this comment. In fact, if you look at sRGB and Adobe RGB (1998), the big differences are the green chromaticity.
    In emulation mode one can store any color space with a smaller gamut (say, of a printer or another display).​
    Emulation on this and any other LCD display expect the HP Dream Color is not color managed (not calibrated and profiled). The entire idea behind this “mode” as I said above is to provide a somewhat less than ugly color appearance of images shown outside ICC aware applications on a wide gamut display. IOW, with proper color management and ICC aware applications, nothing should look ugly unless the RGB numbers really define an ugly color appearance. Outside such applications like most web browsers (aside from the recent release of Firefox and Safari), you can view pretty ugly appearing images because the browser assumes sRGB and your display is providing Adobe RGB (1998) or greater. The “emulation” mode is one that says “we don’t expect fully color management previews but lets not make them look butt-ugly”.
    Again, before I leave this entire mess, there’s zero reason in ICC aware applications to do anything other than profile the display. It doesn’t have to be anything like Adobe RGB (1998) and probably is not for the reasons I’ve already defined. Outside ICC aware applications, just set the emulation to sRGB and hope for the best. Outside ICC aware applications, the previews are simply not accurate or correct. The degree they are incorrect can be accounted for somewhat by acting like a dumb sRGB device that has no clue about the display or document profile. You’re asking the application to send the RGB numbers directly to the display (don’t use, because you can’t use, Display Using Monitor Compensation). This is how Photoshop 4 and earlier and all non ICC aware applications behave. They have zero knowledge of the display or its profile, nor the document and its profile. It makes the viewing of content that’s not color management aware look less awful on a wide gamut display. That’s why the simulation exists. Your display simply cannot alter its native chromaticity! The HP can. No other system today provides this.
    Or to put it another way, lets say I wave a magic wand and now all web and non ICC aware content is now assumed to be not sRGB but Adobe RGB (1998). This is something that may result years from now when virtually everyone is using a wide gamut display. If you view sRGB images on the web (or other non ICC aware applications), it will look real ugly. All Adobe RGB (1998) content while not being properly color managed will look OK. If such a scenario took place, you’d need to set your sRGB or super wide gamut display emulation to Adobe RGB (1998) otherwise everything would look wrong. All Adobe RGB (1998) content would look reasonably well.
    You’ve got a display that says its 113% of Adobe RGB (1998) (mostly marketing speak but the point is, its a wide gamut display, not a so called “sRGB” gamut display). When you view stuff outside ICC aware applications that assume an sRGB behavior, it looks awful. You can’t calibrate to sRGB. Again, unless you’ve got a DreamColor, you’re just simulating, without a true ICC profile and Display Using Monitor Compensation, sRGB. Stuff looks less awful, but its not necessarily correct. You might be able to test this by setting your display for this simulation, opening an sRGB document and compare that to the same document in Photoshop or a similar ICC aware app. Problem is, in the simulation for sRGB, no profile is properly being used for Photoshop (you’d need two displays to see the two previews side by side) They should look much closer than the preview without the simulation but I bet you they don’t match exactly.
    Bottom line. Calibrate and profile your display and work in ICC aware applications taking advantage of the wide gamut properties. If you have to view images outside of such applications and want the so called sRGB, no color management documents to look less poor, invoke the sRGB simulations. There’s zero reason you need to be messing around with the Adobe RGB (1998) profiles or mucking around with the display since whatever you do, the previews are not color managed.
    Or get yourself a DreamColor.
     
  19. Andrew, I completely understand what you're saying. But please read further.
    As far as I know, my display does the same what you're describing when you talk about the DreamColor. Actually Samsung was the first to bring this stuff to mass market. I've reviewed one of their displays back in 2004 I think, which could do something similar. But that wasn't a wide-gamut display, only an expensive consumer display with fancy design and features.
    When I speak about profiles stored in the monitor, I don't mean those warm/cold/movie/photos profiles found in most monitors. I speak about profiles which are calibrated to a specific color space (using a colorimeter, just like normal calibration) and stored in the monitor.
    So, if I calibrate and store an ARGB profile, I don't need to specify a monitor profile in applications or OS. I can just use the ARGB ICC profile. The monitor handles the rest, including clipping colors outside of the specified ARGB gamut and using the stored profile to show a calibrated picture.
    The display does the same as what an application or OS would do. As for the emulation mode, it should work like soft proofing - again, calibrated. It's ingenious.
    Is it technically correct? I'm not sure, that's why I asked the original question. But it indeed seems to work that way - when I try to recalibrate the display using X-Rite software (using ARGB as reference), there's no difference.
    Either way, this is the reason I wanted to know more about the internals of the AdobeRGB1998.icc file... Or the way Samsung XL20/30 actually work.
    I hope you're still here and understand what I mean.
     
  20. As far as I know, my display does the same what you're describing when you talk about the DreamColor. Actually Samsung was the first to bring this stuff to mass market.​
    Nope, it was NEC (see http://www.ppmag.com/reviews/200602_neclcddisplay.pdf) at least in terms of a LED display.
    And no, the DreamColor is quite a different beast from the Samsung.
    So, if I calibrate and store an ARGB profile, I don't need to specify a monitor profile in applications or OS.​
    It would be useful if you’d tell us why you think this is at all necessary or useful.
    I'm not sure, that's why I asked the original question.​
    I think before anyone goes further, you have to figure that one out!
    Either way, this is the reason I wanted to know more about the internals of the AdobeRGB1998.icc file​
    You got that answer in my first post.
     
  21. True, NEC was first with such a display. These Samsung units are basically the same thing, LED and all, with enhancements lying in the profiles. Samsung basically combined this technology with their multiple-profiles capabilities.
    In your NEC review, you write: "The LCD2180WG-LED does have a hardware switch to set the display behavior to mimic sRGB, but using it invalidates the current calibration and the asso- ciated ICC profile. I would like to have the ability to calibrate and profile for this limited color gamut as well as Adobe RGB (1998), and be able to switch on-the-fly via software."
    The Samsung does exactly that - calibrated profiles for other color spaces - the monitor's full gamut, ARGB, sRGB and emulation. It may not be the same as HP, I don't know the underlying technology, but Samsung made a big deal out of it back in the day, so perhaps it's something similar. I think it does change some characteristics in the hardware, perhaps different circuitry treats different profiles... I don't know, but it works very well, except of the weird problem which lead me to my original post.
    "So, if I calibrate and store an ARGB profile, I don't need to specify a monitor profile in applications or OS." - I didn't phrase that in a fortunate manner. I meant the profile which is created for example when calibrating a display, including gamma curves and all. That information is now directly in the display for multiple profiles, and ARGB is used by the OS/apps as a reference point for displaying. The 'useful' part is, you don't need to switch profiles when switching the display built-in profiles. Does that make sense?
    Since you have experience with the similar NEC unit, how would you treat this display? Considering you'd want to use the built-in profiles e.g. for soft proofing or for different lighting conditions?
     
  22. So, Alex. Can I persuade you to talk about the monitor outside of the question about ICC profiles, etc. I nearly bought that monitor to replace my CRT. The LED back light and greatly enlarged gamut of possible color was intriguing.
    How long have you had it? Has the monitor improved your ability to make an accurate print? Has the monitor improved your ability to edit your images? How is the monitor for other uses, such as general surfing of the Web, such as viewing sites like Photo.net?
    Any information you can share on its use would be helpful. Thank you.
    David Ralph
     
  23. David,
    I can't comment on printing much, since there aren't places around here which can do decent calibrated prints and provide a profile for everyone to download (unless it costs a real fortune). If you can make such prints, having a good wide-gamut display like this one surely helps.
    For image editing it's excellent for the price it's currently available. I haven't directly compared it with other LCD displays, but it has replaced my old iiYama CRT. The gamut is crazy, that's why I keep it switched to ARGB or sRGB modes, the reds which can be displayed are too bloody. When I see it with my gf's laptop display side-by-side... The laptop looks like a 50-year old washed out black-and-white print in comparison. It can show darks and whites at once - the only values I can't recognize by sight are from 253,253,253 up. For everyone who considers this display size and price, it can surely replace any CRT. Very good for the eyes too, no blinking (even some cheap LCDs blink if you work with them for too long).
    The internal calibration is extremely convenient. Now, when even web browsers are becoming color-management-aware, it's a nice thing to have such a display. I still need to find out how the emulation mode works. If you don't trust it (like Andrew here), you can always calibrate it by the usual means. Also, calibrator included.
    I like the ability to use 'perfect' AdobeRGB and sRGB. Unlike some displays, which have some colors better behaved than the others, this actually looks natural, so I just use ARGB and switch to sRGB when I view unprofiled pictures. You won't want to go back.
    I've had this monitor for a year and after a short while, I bought another one for my other place too (and was a verge to getting a third one for a dual-display environment).
    A nice review is here, where the display's performance closely matches those with 3x the price: http://www.trustedreviews.com/monitors/review/2007/09/07/Samsung-SyncMaster-XL20/p1. Bottom line - for the current price, go for it!
     
  24. I would like to have the ability to calibrate and profile for this limited color gamut as well as Adobe RGB (1998), and be able to switch on-the-fly via software.
    So, if I calibrate and store an ARGB profile, I don't need to specify a monitor profile in applications or OS.​
    Why? What possible benefit would this produce and, since the system always uses some profile for previews in ICC aware applications, what possible benefit would there by to attempting to trick it (meaning now, all ICC aware applications could be wrong, just as if you calibrated and profiled the display, then trashed that profile and introduced on that isn’t describing the display conditions).
    ICC aware applications always reach for some display profile. There’s no, no color management by not having a display profile option. You can’t turn Display Using Monitor Compensation off, certainly in Photoshop.
    In the case of the NEC LED, when you use the sRGB simulation, its not producing correct previews since the system is still looking at the original profile. But the idea is you’re using this outside ICC aware applications because the wide gamut profile description and display conditions make sRGB like images look ugly. So the simulation is a less bad preview but incorrect. If you’re browsing the web (in a dumb non ICC aware browser), its not ugly but its still not correct compared to viewing in an ICC aware browser with an sRGB like display, an Adobe RGB like display or a 113% of Adobe RGB like display.
    Since you have experience with the similar NEC unit, how would you treat this display?​
    I told you above. I follow my own advise. I calibrate the display to target aim points based on the print viewing booth next to the display and I move on. My web browsers are both ICC aware. If I want an sRGB behaving display. I hook one up (actually I have a dual display system, the other is a NEC 2490).
     
  25. Andrew, I keep telling you the same thing over and over again!
    Just look how my Windows is set up.
    This display is a color management aware DEVICE, so it will convert the incoming image information to display the correct colors! The calibrated color profile is stored in the display, not externally . And it will do that with both sRGB and AdobeRGB - both calibrated! It's not the same as the NEC!
    The point of this is, the display can provide multiple independently calibrated profiles , while the OS and color aware application still use only one (AdobeRGB1998.icc) for display reference . The display will convert the color information by itself.
    "If I want an sRGB behaving display. I hook one up." With this display, you have both calibrated displays in one unit. That's the whole point of this system.
    00VTBt-208687584.jpg
     
  26. Also for reference, here's a measurement from the X-Rite software, when the display is internally calibrated to AdobeRGB (here it needs a bit of recalibration after a few hours).
     
  27. Attachment got lost.
     
  28. Can't upload the attachment at this time. Never mind. The curve in the measurement is a straight line and the measured gamut is 100% AdobeRGB, that's the point.
     
  29. Anyway. It makes no difference which of the color profiles in the above screenshot is assigned to the monitor. The display takes care of the colors and the result is calibrated, unless there's a custom gamma curve loader running or something.
     
  30. I don’t do windows (historically none or broken color management). But that’s besides the point. Again, what possible benefit is there to your idea that “storing” the profile in the panel is at all useful or necessary?
    The point of this is, the display can provide multiple independently calibrated profiles , while the OS and color aware application still use only one (AdobeRGB1998.icc) for display reference . The display will convert the color information by itself.​
    For ICC aware apps so can my NEC by simply picking the user settings for calibration (which has to affect the panel to alter its behavior as specified by the user) and automatically switching the profile in the system. Doesn’t your software allow you to do this? Its moot where the profile is stored (and its moot that you need to somehow know or mess with Adobe RGB (1998)). You have one setting that’s say 300:1 contrast, 150cd/m2 and D65 with a TRC gamma of 2.2. A profile is built of that behavior of the display. You now want a setting of 200:1 contrast, 120cd/m2 and D55 with a TRC gamma of 2.2. To “switch” you have to update the calibration in the panel (or LUT, mine does nothing to LUTs), and then switch to the correct profile built of that calibration. Neither may be Adobe RGB (1998) or close to it as each has differing calibration that I specify for specific soft proofing needs. Again, there’s zero reason to be thinking that “well I want Adobe RGB (1998)”. What you want is a calibrated behavior from the device (defined by white point, TRC gamma, cd/m2 and if the display can control both white and black, contrast ratio). They may be a mile from the theoretical specifications of Adobe RGB (1998) (where do you see the specifications for blacklight luminance and what if its a mile from the print viewing conditions? You absolutely do not want that behavior!).
    "If I want an sRGB behaving display. I hook one up." With this display, you have both calibrated displays in one unit. That's the whole point of this system.​
    The NEC WG displays can do a very basic "sRGB" emulation but it is fixed and not perfectly accurate. As I've said, its solely for non ICC aware applications. Basically a unit like an Eizo takes the incoming RGB video data, de-gamma's it (so it is "linear"), then puts it through a matrix that maps RGB into whatever RGB colorspace you want (as long as it is within the gamut of the display itself), then it re-applies the gamma correction before being sent to the display panel. That is basically what all of the 14bit, 16 bit stuff that Eizo goes on about is for. That high bit processing kicks in when you are using the color gamut emulation function. When you are using the normal gamut mode that’s bypassed.

    The argument for doing it in the display is getting less and less since the preview application is responsible for converting the legacy image so it can be correctly viewed on the display. That argument for needing the application to handle it holds not only for wide gamut displays but *all* displays - just you *really* need to do it on WG displays, but you can kinda get away without on normal gamut displays. Again, that’s due to the huge legacy of sRGB like (not exactly sRGB) content.

    And who wants to be switching the display's gamut setting depending on what legacy file they are viewing?

    As I’ve said over and over, feature is good for some applications that don't support color management (like some video editing apps, and most Linux applications where a specific gamut is needed), but nowadays almost all viewer apps support color management.
    You are either concerned about proper previews meaning ICC aware viewers or you’re viewing outside such products and all bets are off. Calibrate and profile the display to the target calibration based on the final output viewed by the display (the cd/m2 target and the fact LCD’s native’s intensity is so damn high, we get these millions of posts about “my prints are too dark”). That target doesn’t have to be anything like Adobe RGB (1998) and as I’ve said from post #2, you have no reason to be mucking inside this specification, certainly not for a simulation mode that’s entire existence is to allow you to view sRGB like images, outside an ICC aware application on a WG unit. You’re digging yourself a tunnel down a big rabbit hole you don’t need to. Calibrate, profile and move on.
     
  31. I see now we are talking. Just to explain why I'm doing this.
    To “switch” you have to update the calibration in the panel (or LUT, mine does nothing to LUTs), and then switch to the correct profile built of that calibration.​
    With Samsung, you do the calibrations beforehand. 4 profiles = 4 calibrations. For each profile there's a separate calibration, where the display limits its gamut, does the calibration and stores the result. To the application world, it can act as a single display (say, AdobeRGB or antyhing else), but in reality it's like having 4 calibrated displays with different gamuts.
    Again, there’s zero reason to be thinking that “well I want Adobe RGB (1998)”. What you want is a calibrated behavior from the device​
    Yes that's true. I've just picked AdobeRGB as my default gamut/color space/reference, because the default gamut of the monitor makes everything non-calibrated (like the Windows desktop) look fugly. So this is just arbitrary. You don't have this problem on a Mac even on a wide-gamut monitor. Of course the display still has a 'custom' mode with full gamut, which can be calibrated and stored the old-fashion way.
    If I'll want to see how the image should look on a different device, that's what the emulation mode is for (which is calibrated beforehand to the specifics of that device). But I have concerns about the quality of that emulation - that lead me to my first post.
    So I'm just trying to learn the specifics of my display and use the advantages. I have an idea how to confirm the quality of the emulation; will try it later.
     
  32. With Samsung, you do the calibrations beforehand. 4 profiles = 4 calibrations.​
    That's the same with NEC and every product I've ever used. Alter calibration and the current profile is invalid. With the NEC, I don’t know what it “physically” does in terms of the profile when you switch calibration targets. I don’t “see” four profiles for four calibrations, it might store them somewhere but its moot. You switch calibration targets, the associated and correct profile is updated in the system. And that’s good because when you go into say Photoshop’s color settings, select RGB working space and then examine the list, the correct profile is listed as it should be (Monitor:XXX where XXX is the profile the system, Photoshop and all other ICC aware app’s are using). I want that profile recognized by the OS and not stored (if it is) in the panel.
    I've just picked AdobeRGB as my default gamut/color space/reference, because the default gamut of the monitor makes everything non-calibrated (like the Windows desktop) look fugly. So this is just arbitrary. You don't have this problem on a Mac even on a wide-gamut monitor.​
    There’s nothing different here on the Mac. The basic “flaw” in your thinking starting from post #1 is you somehow have to deal with the specifics of Adobe RGB (1998). You don’t. Its a synthetic color space built with simple math used for editing images (its a working space, a Quasi-Device Independent editing space). The display is an output device.
    The reason I sent you to the URL is to understand the role of working spaces and how they are completely divorced from the display. That’s the big deal about color management in Photoshop 5.
    Do this. Calibrate and profile your display and if you have utility to examine its gamut plot (or more), you’ll see its not Adobe RGB (1998). It doesn’t have to be Adobe RGB (1998). Its “Alex’s LED display RGB” set to some setting, producing a profile that defines that setting. It might be somewhat close to Adobe RGB (1998) it might not, doesn’t matter. As I said above, if you could exactly hit Adobe RGB (1998) with all the various parameters that define it (down to ambient light, proper or idealized backlight intensity), you wouldn’t need that colorimeter or that software package. Your display would exactly produce Adobe RGB (1998) and the Adobe RGB (1998) profile would or could be used as the display profile. Clearly Samsung is providing you software and tools because your display is unique and ever changing. Why else do you conduct a calibration every month or so? Why build a new profile for each calibration if its indeed exactly Adobe RGB (1998) even after calibration (answer, its not).
    You’re viewing content in something close to say sRGB on the web (again, like your display, this content may or may not be in sRGB and the vast majority of images on the web are untagged and not color managed). When you view this content on an sRGB like display, outside ICC aware applications, they look OK. They may not match the same sRGB content in Photoshop but that is because Photoshop is an ICC aware application. The numbers are being properly previewed using your custom display profile, the embedded doc profile (or assumption) and the Display Using Monitor Compensation architecture.
    Now in this web browser, you view this sRGB data on a WG display without color management. It looks ugly as you describe. What’s the solution? View the numbers tagged with the proper profile in an ICC aware app. Doesn’t matter if its in sRGB, Adobe RGB (1998), ProPhoto RGB or Epson Luster RGB. They all preview correctly. You don’t have to be setting the display for anything other than a proper calibration for proofing. OR setup a simulation of sRGB which isn’t accurate but again, like viewing on that sRGB like display, less ugly (but not correct nor matching Photoshop). Where does having an Adobe RGB (1998) internal display profile do anything useful?
    So I'm just trying to learn the specifics of my display and use the advantages. I have an idea how to confirm the quality of the emulation; will try it later.​
    Concentrate on ideal calibration settings, build a profile and move on. IF you find yourself viewing ugly content outside ICC aware applications, if you wish, set up the sRGB simulation so it looks less ugly, understanding the previews are not necessarily correct but instead less ugly (because in simulation mode, you’re not getting full color management from either the display or the app). Its all rather simple actually.
     
  33. FWIW, this is what started the flames and again, I stand behind the statement based on a whole lot of text above:
    That said, there’s about zero reason to mess with any of the settings (after which, its not Adobe RGB any more). Nor does the working space have to have anything to do with the display, display calibration etc. The display and the editing of images is totally independent and has been since Photoshop 5. Calibrate and profile your display and move on.
     
  34. “Alex’s LED display RGB”​
    LOL I love this. Yes, that's the term I should have used instead of overusing the 'AdobeRGB emulation/mode/profile' description. If I've used that, we might have understood each other from the beginning.
    You see, I should have worded my original question like this:
    I have a display with a dedicated, calibrated "Alex LED RGB" mode (by the way based on data/gamut from the AdobeRGB1998.icc file, but of course little having common with the AdobeRGB specifications). My display also has an emulation mode and I'm trying to mimic that "Alex LED RGB mode" in emulation... But I run into a visible bug with gamma. WTF? Is it something special that's contained in the AdobeRGB1998.icc file?​
    Well... I hope that clears it up! Of course then we run into the "why would I want to do that" argument, which has nothing to do with this.
    Nevertheless, I think I've identified the problem. When I switch the display to 'Alex LED RGB emulation' (instead of the dedicated Alex LED RGB mode) and try to recalibrate, the resulting gamma curve has a visible dip in the dark area... Meaning the display is obviously screwing up the emulation by showing the dark areas brighter and not the other way around. Will contact Samsung about that, maybe there's a reason for this.
    So Andrew... Thanks and have a nice day :)
     
  35. My display also has an emulation mode and I'm trying to mimic that "Alex LED RGB mode" in emulation...​
    We’re not there yet. You have Alex RGB mode, meaning the display calibrated to your target calibration needs with a profile. That’s going to work just fine with all ICC aware apps. Why then would you wish to use it in emulation? The emulation is to move to a greatly different behavior (sRGB or something along those lines) because you have a WG display viewing images not anywhere in that color space outside of ICC aware applications.
    When I switch the display to 'Alex LED RGB emulation' (instead of the dedicated Alex LED RGB mode) and try to recalibrate, the resulting gamma curve has a visible dip in the dark area... Meaning the display is obviously screwing up the emulation by showing the dark areas brighter and not the other way around.​
    I suspect because you want sRGB emulation. Again, read the above points about Adobe RGB (1998) outside ICC aware apps that expect something like sRGB. Emulation as the term suggests is for this issue. IF you had a super wide gamut unit, everything in non ICC aware app’s produce something close to Adobe RGB (1998), you’d want an Adobe RGB (1998) emulation. That’s not the case today. You want to emulate sRGB when working outside ICC aware apps and use Alex RGB (with that “13%” greater gamut) in ICC aware apps. If not, why did you buy a WG display?
     
  36. No, I'm not trying to emulate sRGB... The display has a special mode for that. And even when calibrated, I understand it's not perfect.
    I was really just trying to get two modes of the display (1. Alex LED RGB, which Samsung calls "AdobeRGB", and 2. "emulation"), to display the same image. It doesn't, even after calibration based on the same ICC profile. The geek in me wanted to know why.
    It's like when you have two cameras of the same type and with the same shutter speed, same ISO-rated sensitivity, same aperture and same settings provide a vastly different resulting exposure. One of those cameras/modes is obviously faulty - or both - and I wanted to know which one. BTW my other XL20 behaves the same, so it's clearly a technical issue.
    Why do I want to do that? The emulation mode, in theory, should provide the same resulting image as the reference profile with a smaller gamut (printer, different monitor...). That's definitely a nice toy to have - like soft proofing. So I wanted to test how it works and move from there.
    The AdobeRGB (Alex LED RGB) mode definitely works well - when I preview the before/after calibration in X-Rite software, it is completely the same, or with differences common when recalibrating. The gamut in that mode is equal to the AdobeRGB gamut as specified in it's ICC profile (or your paper), thus allowing to use AdobeRGB1998.icc as a display profile. I know how twisted it sounds, but that's the way the display was designed. I wanted to get the same results from the emulation mode.

    Why don't I just use the full gamut at least for color-aware apps instead of arbitrarily limiting myself to Alex LED RGB? Why don't I use normal ways for soft proofing? Because there aren't any places where I can print accurately anyway, so even the Alex LED RGB mode is already almost more than what I can take advantage of. It's better than just taking chances in sRGB mode. Using the full gamut could easily screw up my attempts for a nice print, if I'm not careful.
    I might re-evaluate when I finally get a decent printer. For now, I have a geeky toy to play with...
     
  37. No, I'm not trying to emulate sRGB...​
    And yet, that’s where its somewhat useful.
    I was really just trying to get two modes of the display (1. Alex LED RGB, which Samsung calls "AdobeRGB", and 2. "emulation"), to display the same image. It doesn't, even after calibration based on the same ICC profile. The geek in me wanted to know why.​
    Why would they call your custom display color space Adobe RGB (1998)?
    The emulation mode, in theory, should provide the same resulting image as the reference profile with a smaller gamut (printer, different monitor...). That's definitely a nice toy to have - like soft proofing. So I wanted to test how it works and move from there.​
    Again, without understanding what Samsung has told you, I don’t know why you’d do this other than emulate something that’s a far cry from the native wide gamut behavior in a calibration you desire. What are they telling you this is supposed to be used for? If its the same (which seems odd), what’s the point?
    The AdobeRGB (Alex LED RGB) mode definitely works well.​
    And I’d expect that to be the case.
    when I preview the before/after calibration in X-Rite software, it is completely the same, or with differences common when recalibrating.​
    This is Match software, where after the process there’s a woman’s face and you toggle before and after? Depending on the display technology, I can see why you’d see no difference. The adjustments are in the panel, not a LUT on the graphic card so I’m not sure the on and off (before and after) would do anything anyway.
    The gamut in that mode is equal to the AdobeRGB gamut as specified in it's ICC profile (or your paper), thus allowing to use AdobeRGB1998.icc as a display profile​
    But we are told this device exceeds the gamut of Adobe RGB (1998) by 13% so how can that be? Why would anyone want to clip the gamut by 13% or any amount (other than emulating sRGB as an example)? There’s some fundamental process here that doesn’t make sense to me.
    Because there aren't any places where I can print accurately anyway, so even the Alex LED RGB mode is already almost more than what I can take advantage of.​
    Well you’re kind of painting your files into a corner. The weak link here is the display gamut. Lots of images and output devices exceed Adobe RGB (1998). I only work from Raw to ProPhoto. There are lots of colors that potentially I can’t see on my WG NEC but I can see them when I print them to my HDR inkset on the Epson. And in your case, we’re talking 13%. And if you view the very simple triangular shape of your display profile, plot that next to a far more complex plot of a printer profile, you’ll see lots of areas where they overlap (display gamut might be a tad larger in one area of color space, printer gamut is a lot larger in other areas).
    I might re-evaluate when I finally get a decent printer. For now, I have a geeky toy to play with...​
    Just be careful here because if you clip images in a working space (editing space) to Adobe RGB (1998) they can’t come back unless you re-render from Raw or rescan using a larger input color space. I think you are concentrating far too much on the weak link in terms of gamut, the display. Its one thing to want to view sRGB like images outside ICC aware apps. But since around 1998, the separation of display color space and editing and output color space has been routine. Prior to Photoshop 5, your display color space, its gamut limitations and all the idiosyncrasies of this device were your editing space. That’s not a limitation any more.
     
  38. The emulation mode, in theory, should provide the same resulting image as the reference profile with a smaller gamut (printer, different monitor...). That's definitely a nice toy to have - like soft proofing.​
    I kind of suspect you can do this in Photoshop by actually soft proofing the image, potentially with more precision (since ACE uses 21 bits to do this). And with a printer profile, you can now toggle the rendering intents to see the differences. Matrix profiles only have a colorimetric table (until we get V4 ICC working space profiles). Making a larger gamut display a smaller gamut in Photoshop is easy (going the other way, not possible).
    If you open a wider gamut image in Adobe RGB (1998), duplicate it and convert the dupe to sRGB, the two should appear nearly identical.
     
  39. OMG! This truly is the thread from hell.
    Though I often get into similar heated discussions with Andrew (i.e. don't even mention using Mac's eyeball calibrator to assess the quality of third party LCD's lest yea suffer "digitaldoggedness") I have to agree with what he's saying here. I'll try to fill in the logical holes not communicated, hopefully without turning it into a longwinded essay. Wish me luck.
    Computers are dumb machines. Anything you want out of a computer has to be told to it in the language of math. Color spaces are 3D maps mathematically describing every plot to every position for every color that is encompassed within this map. The amount of various colors available to preview is limited by your display whose color model must also be described by a 3D map.
    A simple synthetic matrix color space is all that's needed to edit color for smooth transitions when editing on a display that's not so smooth and whose mathematically described 3D model by physical measurements has too many bumps in its road map (the core of this discussion). You can't precisely map color within the mathematical world of a computer if you don't tell it exactly where to go. Using a synthetic matrix color space to tell (lie) to the computer that your display is a synthetic device when it's not will introduce imprecise output whether on the display and/or to a printer.
    You need to think of computer color from the perspective of a cartographer. Even though some maps look quite similar viewed close up their roads to get from point A to point B never are.
     
  40. I only work from Raw to ProPhoto. There are lots of colors that potentially I can’t see on my WG NEC but I can see them when I print them to my HDR inkset on the Epson.​
    I envy you. If I want to get a half-decent print, I need to travel 50 kms to get to a place where at least they know what RAW is. And they don't provide color profiles either. The last place which did, closed down after being in business for around a year. Some places can dig out profiles from the times when they set up the printer for the first time, several years ago. Quite worthless.
    As the result, I often prepare multiple versions of one photo (with tweaks to be slightly 'off' from what I see on my display) so I get at least one that looks quite right. Because if I don't get one, I need to travel back, make edits and then go make the print again. In such a scenario, it's not very wise to risk too saturated colors. I rather have reddish orange showing up as reddish orange, instead of deep red showing as purple in print. The latter has happened to me more than once, teaching me the deep reds my display can show apparently aren't easy to print.
    Again, if you have your hi-quality printer on your desk, calibrated, that's a different beast.
    Just be careful here because if you clip images in a working space (editing space) to Adobe RGB (1998) they can’t come back unless you re-render from Raw or rescan using a larger input color space.​
    I think I'll be more then happy to do that when I can print properly. Actually I usually don't do much edits after developing the RAW - mostly color tweaks, so at least I'll be able to get those right.
    Why would they call your custom display color space Adobe RGB (1998)?​
    I suspected they just call it that way to have a cool-sounding product. When I got the display, I didn't believe Samsung made anything special to the AdobeRGB mode. Now I see it's actually heavily optimized.
    But we are told this device exceeds the gamut of Adobe RGB (1998) by 13% so how can that be? Why would anyone want to clip the gamut by 13% or any amount (other than emulating sRGB as an example)? There’s some fundamental process here that doesn’t make sense to me.​
    The display's native gamut is 113% AdobeRGB, exceeding it mostly in reds. In AdobeRGB mode ("Alex LED RGB") it is supposed to be identical to the AdobeRGB gamut. As to "why?", see what I said about printing.
    This is Match software, where after the process there’s a woman’s face and you toggle before and after? Depending on the display technology, I can see why you’d see no difference. The adjustments are in the panel, not a LUT on the graphic card so I’m not sure the on and off (before and after) would do anything anyway.​
    Eye-One Match 3, yep. The preview feature shows the state before and after calibration (that woman face). Since the state 'before calibration' is already calibrated internally in the panel (using the same colorimeter), it doesn't need further calibration in Eye-One Match and thus no custom profile.
    I kind of suspect you can do this in Photoshop by actually soft proofing the image, potentially with more precision (since ACE uses 21 bits to do this).​
    Most likely. But not everyone has Photoshop (I don't) and not all applications have this option. So it's a handy feature... If it works well, which I'm not so sure.
     
  41. Hi Tim,
    I'm not disagreeing with Andrew are you either. You see, he apparently can use the theoretical advantages of wide gamut and stuff like that in practice. In my case it causes more problems than it solves. Fortunately I have a display which can arbitrarily limit its own possibilities, thus reducing some headaches. Does that sound like and oxymoron? I can imagine it does to you or Andrew; that's why I'm taking this time to explain why I do this stuff :)
     
  42. Missed this:
    Using a synthetic matrix color space to tell ( lie) to the computer that your display is a synthetic device when it's not will introduce imprecise output whether on the display and/or to a printer.​
    True in theory, in practice this Samsung display takes that synthetic data and transforms it to its own colorspace and resulting image. I think the driver is doing part of the job, that's because there are no difference between which monitor profile is selected. Perhaps Photoshop would override this behavior, I can't know.
     
  43. I think the driver is doing part of the job, that's because there are no difference between which monitor profile is selected.​
    If you're on a Windows system changing the profile either within the display (if that's what you're talking about) or in the system using Window's video driver/color management interface panel will most certainly not show a difference because Window's can't do a live update of the LUTs (Look Up Tables) which are the internal curves embedded in a profile made by a measuring device like a colormeter that knocks the kinks out of non-smooth characteristics of a device which is the actual state of your display.
    You have to reboot to get the updated LUT's but that wouldn't take effect anyway since you're selecting a synthetic profile and these don't have LUT's to load into the video card.
    Again, Alex, you need to think of all of this like a cartographer (map maker). It's about precision. You can't readily see the precision using your eyes because it's being described to the computer using math.
     
  44. I meant those profiles in the OS... Or apps which can show the changes on the fly. I think I used it in Bibble or something, not sure right now. Frankly, my head is close to exploding, I can't think anymore (it's just midnight) and I still have to get some office work done. Had a hard day :) Will verify some things tomorrow.
     
  45. Also keep in mind (only after you get some sleep of course) a custom profile unlike a simple matrix working color space has two parts to its profile, the LUT mentioned above and the same XYZ matrix numbers as in the working color space profile. Even if you set a working space as your system profile, Photoshop will use it just as it would with a custom profile's to control the appearance of hue and saturation of color managed images. It's just that the custom profile's matrices were formulated in combination with the LUTs built from a measured response of your display. Working space profiles have no known physical measurement embedded in them because they have no LUTs.
    Combine this with the fact that luminance is tied to hue and saturation.
    Your display can have various luminance levels and contrast ratios just by adjusting the backlight. What was AdobeRGB working space's backlight set at when it was created? No one knows because obviously it doesn't have a backlight to adjust. It's not a real display.
    When your display is measured and forced into a target of 2.2 gamma, neutralized 6500K and 100-120 cd/m2 brightness by the final LUTs that load in the video card, the matrices recorded in the profile Photoshop uses are matched and mated to the LUTs that affect your display globally. The matrices don't globally change the appearance of a display.
    Photoshop only adjusts previews in images tagged with a given gamma based working space profile using the simple 2.2 gamma curve written into the custom profile along with the XY numbers forming the matrix to accurately display hue/sat. The LUTs that are mated to the custom profile's matrix need to be loaded for Photoshop to display hue/sat accurately. Remember luminance which includes contrast is tied to hue/sat.
    Working spaces don't have any luminance describers that match up to LUTs because they don't have LUTs.
     
  46. This is to illustrate what I'm talking about...
    Below on the left is a screenshot of Colorsync Utility showing the makeup of my i1Display custom profile of my iMac that includes a LUT. In this case it's just showing the blue channel. Its color gamut size and shape is very close to sRGB and I could load it as my system profile but, it's not an exact description of the condition of my display with warts and all. The wrinkly blue channel with its bent down hook in the highlights shows how off my iMac is from being completely neutral down the entire tonal scale, how off it is from 2.2 gamma and 6500K.
    When sRGB working space is loaded as my display profile, reds turn slightly orange and skintones dull with a noticeable green cast in color managed images viewed in Photoshop. That's the price paid for being imprecise in describing the real condition of my display.
    The Mac loads the RGB LUTs live without a reboot so I can see this hue/sat shift in Photoshop when selecting the custom i1Display profile and the sRGB working space profile as my system profile.
    00VTZu-208887584.jpg
     
  47. OK so, if you guys are still following...
    Tim, all that is true. However you've missed the point of this display storing the calibration data internally. That includes LUTs.
    In the usual calibration environment, a color-aware app works in the color space of choice. To view the image on the display, the app and the operating system use the display's profile (LUTs and all). The end result are some sort of raw data feed into the display. Sorry if I didn't use the correct technical terms, but I hope we get the point.
    With XL20, it's different. If you travel back to my original post, there's a screenshot of the Samsung calilbration/setup software with the AdobeRGB1998.icc profile loaded. Those specs seen in that screenshot are what the display is calibrated to when running in AdobeRGB mode (now renamed Alex LED RGB). Notice the target gamut - white triangle - is smaller than the display's native gamut - pink triangle. The display is then calibrated so the imprefections are ironed out (like with normal calibration process with Eye-One software for example) and the profile is stored internally. In other words, it is trying to be as perfect AdobeRGB device as a display can be. Who cares how exactly it is done. This is how it is designed.
    Of course we run into the problem how to introduce such a display to the system and applications? As Andrew pointed out a bazillion times, a display profile is necessary, but which one? If you think about it, the synthetic AdobeRGB1998.icc file is an obvious choice, because the LUTs are already stored in the display.
    Why would you want to have the LUTs stored in the display instead of the operating system? I don't know about Mac but on Windows, to actually see the LUT's you need to run an application on startup (Adobe Gamma, X-Rite's tool etc.). That can be easily invalidated by any software which changes the gamma, like a video game or movie player. On Windows Vista, as little as a system prompt is needed to trash your whole calibration. Having it in the display is a a very practical thing.
     
  48. I'll try to upload the screenshot I've tried before. This is the display calibrated to AdobeRGB mode (Alex LED RGB), when measured in X-Rite One-One Match 3. See that LUT? No bumbs like a usual display would show. That's because it's already calibrated internally.
    However in this particular case recalibration might come handy (some time passed since calibration). If I'd save the profile at this point in Eye-One Match, I'd treat this display like a normal display with externally stored LUTs. Than you would have at least two places where the display LUTs are saved. So I'll rather fire up Samsung software, recalibrate there and save internally. If I do want to use an external tool, like this Eye-One Match, I can switch to the Custom mode of the display, which is not calibrated internally, and have a usual display.
    You can also see the measured gamut, which matches the AdobeRGB gamut. Yes I have my doubts too (seems a tad smaller in greens), I have to check with Samsung how is that taken care of; LUTs are easy, with gamut I'm not so sure. I think the panel tries to compensate by overblowing the greens, as green is the first color to get out of whack when the display is not calibrated for some time. As I said; the result is a custom display with its own artificial color space anyway.
     
  49. See that LUT?​
    I'm referring to the LUT in your first Samsung screenshot which is not a LUT but the TRC's=(Tone Response Curves-one for each RGB). They act as a simplified 2.2 gamma curve that's embedded in all profiles whether custom or synthetic and used by color management apps as a reference to maintain constant gamma regardless of the device the image is viewed on.
    The LUT as I said above globally changes the look of the display if it happens to not match exactly target gamma of 2.2 and 6500K neutrality. The LUT smooths the kinks, warts, blemishes, odd tints that exists in an analog device such as a display. AdobeRGB working spaces doesn't have these kinks and blemishes and the Samsung software can't make your display fit to this because it knows nothing about those kinks and blemishes.
    Why would you want to have the LUTs stored in the display instead of the operating system?​
    Alex, I have to admit I'm not familiar with the Samsung software in your screenshot. I couldn't tell if that software was operating within the display's LUTs or video card's which is why I asked...
    ...changing the profile either within the display (if that's what you're talking about) or in the system using Window's video driver/color management interface panel...​
    If you think that software with its limited interface is precise enough in telling you what's happening under the hood then use it, but I wouldn't because it's not showing me a 3D model comparison between a display's true response which has to be measured and the AdobeRGB synthetic color space which is based on NTSC specs with an error in calculation in the green colorant created and left in the final build by Adobe engineers. There's no real display that has the same exact 3D color model as AdobeRGB.
    The software does say "Emulation". That's not precise enough for me especially when I paid good money for an X-rite colormeter package that tells me exactly the condition of my display and knows all its color idiosyncrasies and corrects for them accordingly.
    I still think you don't understand the physics behind calibration that Andrew and I tried to communicate and why it's necessary to measure and not rely on "Emulation" software. If you feel confident it's telling you the truth about the condition of your display then use it.
     
  50. I reread this thread and I have to get something clear about your intent behind this topic.
    Do you think the Samsung software is useful or are you just showing it because it looks like a clever piece of software? You know a Gee Wiz kind of toy? I'm not sure.
     
  51. However you've missed the point of this display storing the calibration data internally. That includes LUTs.​
    Mmh, that's interesting. Are you saying that the LUTs that smooth out the kinks of your display as measured and built by the Xrite colormeter package can be uploaded to the Samsung internal LUTs?
    How can you prove this? How do you know which LUT loader is being used since you can't detach the display from the video card? The only way you could prove this is happening is to disconnect the display from the computer and see if a grayramp still looks the same on screen as it did when connected to the computer. I don't see how this can be done.
     
  52. The discussion thread below on the XL20 sheds some light on what Natural Color Expert is capable of doing:
    http://www.freelists.org/post/argyllcms/i1-Display-bundled-with-Samsung-XL20,13
    Since it's an LED backlit display there is an option to purchase an i1 Display calibration package with Samsung's name stamped on the colormeter. There's still a question of whether the internal LUTs can be accessed, replaced or modified due to some firmware concerns mentioned in that thread.
    I don't believe custom LUTs created and embedded by third party calibration packages can be uploaded to the XL20's 10 bit internal LUT chip.
    The pdf on that model doesn't indicate that can happen as well.
     
  53. Since it's an LED backlit display there is an option to purchase an i1 Display calibration package with Samsung's name stamped on the colormeter.​
    The big question being, does it have modified filter matrices for the WG unit vs. what appears to be an “off the shelf“ (sRGB expecting matrices) of Alex’s EyeOne display? NEC has such an option.
     
  54. Haven't read the whole response yet (will in a minute), but you've asked - how do I know if the LUTs are really stored in the display? Well. With Samsung software (NCE) you first calibrate the display and store the resulting profile in the usual fashion. Then, the display asks 'Do you want to upload this profile to the panel?' Click yes, the display blinks, changes colors and there it is.
    No you can't upload third party profiles into the display (only emulate them). Perhaps one could swap the just-created profile with another one before it's uploaded to the panel, I haven't tried that. As I said, after I calibrate the display this way, I can measure the result with Eye-One and the display measures 'perfect', i.e. no recalibration needed. So yes, obviously the LUTs are stored in the panel.
    Anyways, as I mention, the LUT profiles are also stored as normal files. I'd like to see what's inside as well. Can you please check the curves and show me screenshots? I have no way to preview the LUTs in these files, since I don't have Photohsop or any other app which can do this. But I'd like to see how the internals of the panel work. Here are those profiles: http://ulozto.net/3625496/xl20colorprofiles.zip Thanks in advance! I do want to know what the display actually does.
     
  55. With Samsung software (NCE) you first calibrate the display and store the resulting profile in the usual fashion. Then, the display asks 'Do you want to upload this profile to the panel?' Click yes, the display blinks, changes colors and there it is.​
    Ah. I wonder if this is just a poor choice of words on their. It sounds like its asking you to upload the LUTs into the panel, something that might be necessary under Windows (not Mac). Tim?
     
  56. I'm referring to the LUT in your first Samsung screenshot which is not a LUT but the TRC's=(Tone Response Curves-one for each RGB).​
    No, that's just specifying the target you want to achieve. Same like with any other calibration software, but you can also specify the desired gamut.
    If you think that software with its limited interface is precise enough in telling you what's happening under the hood then use it, but I wouldn't because it's not showing me a 3D model comparison between a display's true response​
    I'd like to see that too. But it seems to work with the same precision like the Eye-One Match. After all, the color measurement device is the same and the calibration process looks exactly the same. Want me to record a video? :)
    I still think you don't understand the physics behind calibration that Andrew and I tried to communicate and why it's necessary to measure and not rely on "Emulation" software.​
    I hope you don't think so anymore. I understand the workflow - after all, at first I've used this display just as a normal display without all this things. I'm trying to find out how reliable these Samsung specialties are.
    I reread this thread​
    Wow! isn't your head hurting? :)
    Do you think the Samsung software is useful or are you just showing it because it looks like a clever piece of software? You know a Gee Wiz kind of toy? I'm not sure.​
    Well it wasn't really my intention to show all the specifics. It's just that Andrew thought I'm so uninformed, it forced me to explain all this. I think this Samsung software is damn useful, but I want to know how exactly reliable it is. Samsung is not exactly Eizo, Gretag Macbeth or some of the other companies you trust by default. Also this emulation mode seems to have an obvious bug. You know, I'm an IT journalist/reviewer and I can get really curious, especially about my own equipment.
    The only way you could prove this is happening is to disconnect the display from the computer and see if a grayramp still looks the same on screen as it did when connected to the computer. I don't see how this can be done.​
    I measured the result with X-Rite Eye-One Match (see the attachment in my response from Jan 12, 2010; 06:12 a.m). It seems to work as advertised. Also, I have two different computers with two XL20 units with two calibrators. That's the best I can do for now, but I hope one of you can show me what's inside those profiles I've uploaded/posted in my previous reply.
     
  57. Ah. I wonder if this is just a poor choice of words on their. It sounds like its asking you to upload the LUTs into the panel, something that might be necessary under Windows (not Mac). Tim?​
    No, far from it. On Windows, for example when you calibrate using Eye-One Match, the resulting profile needs to be loaded on startup with a special application (anyone remembers Adobe Gamma and all the confusion it caused?). This panel's hardware is basically replacing that application. Some Samsung displays have been doing this for a few years and believe me - when I've encountered it for the first time, I was as confused as you are.
     
  58. But now that you mention it Andrew, I wonder how this display works under Mac.
     
  59. If you want, you can check the XL20 manual here: http://downloadcenter.samsung.com/content/UM/200701/20070118101944203_BN59-00577A-01Eng.pdf but beware, it may confuse you some more. Also it doesn't mention all the characteristics of the emulated profile.
    Anyway, quoting from the manual:

    Using the Emulation tab, you can create a new profile that contains your selected settings, apply an existing profile, or apply a stereotype of a system profile. Different from a profile created from the Calibration tab, you can also change the R, G, and B color gamma. Of course, you cannot use a gamma value beyond the range of gamma values that can be expressed by your Edison monitor.​
    By 'stereotype of a system profile', they mean sRGB or AdobeRGB. Not sure what they mean by Edison monitor.
    This is how's the emulation process is described:
    To emulate using a new set of settings, select your values.


    Brightness : The target brightness: Select your personal maximum brightness.
    Adjust Black Level : Select a black level.


    Color Coordinate : White Point : Directly enter a coordinate, or move the white slider on the blue bar to the required value position.

    Gamma :Select Gamma values for the R, G, and B components.

    To create a monitor profile using a system profile or an existing profile created, select a profile from the Profile list and click the Start button. If your profile is not in the list, click the Load Profile button and select it.
    When selecting a profile for which you already performed emulation, you can skip step 2 in the Calibration section and apply the settings directly by clicking the Apply Now button. However, downloading the settings to your monitor is not possible because it is assumed that downloading them was already performed at the early stage of execution. Therefore, if you want to download the settings to your monitor, you should not click the Apply Now button and you should proceed with step 2 described in the Calibration section.​
    It should be pretty self-explanatory. Notice how they always repeat 'save the profile to the monitor'.
    Then, the calibration takes place. After that,
    You will be asked whether to save the emulated settings to your monitor. To save, click the OK button. It will take some time to save these settings. The emulated settings saved in the monitor are saved as the emulation mode which constitutes a part of the color mode of your monitor. From now on, your monitor will be displayed with the saved settings whenever emulation mode is selected.​
    Should be clear enough, I think.
     
  60. No, far from it. On Windows, for example when you calibrate using Eye-One Match, the resulting profile needs to be loaded on startup with a special application (anyone remembers Adobe Gamma and all the confusion it caused?).​
    Understood but it *appears* this might be the same basic functionality. Do we know for a fact that an ICC profile data is loaded into the panel? Could this just be poor nomenclature for “load the LUT data” somewhere? It appears to be behaving like the gamma loader no?
    It should be pretty self-explanatory. Notice how they always repeat 'save the profile to the monitor'.
    Then, the calibration takes place. After that.​
    Seems totally backwards and not possible. How can you load a profile THEN calibrate? Why? In every product I’ve ever worked with, you calibrate (or not if you keep a totally native state), then profile that behavior. I suspect you need to have a conversation with the company.
     
  61. Seems totally backwards and not possible. How can you load a profile THEN calibrate? Why? In every product I’ve ever worked with, you calibrate (or not if you keep a totally native state), then profile that behavior. I suspect you need to have a conversation with the company.​
    Um, it probably sounds unclear because it describes the actual steps in the software, not the theory. Also the wording is not very fortunate.
    What you can actually do, is:
    1) select a profile for emulation (.icc or .icm file). The software reads the *desired* parameters from this file - gamut (RGB coordinates), gamma, white point, black level, luminance. You can also specify all this manually. Specifying gamut is the unusual thing.
    2) the display is set to conform to these specifications and measured
    3) the calibration profile (what's the correct term for this? I mean the LUTs) is created and uploaded to the display, so the image it shows is conformant with the data specified in step 1.
    Should be more clear.
    Understood but it *appears* this might be the same basic functionality. Do we know for a fact that an ICC profile data is loaded into the panel? Could this just be poor nomenclature for “load the LUT data” somewhere? It appears to be behaving like the gamma loader no?​
    For now, let's go with the presumption the LUTs are stored in the panel. Now that you ask, I can try to unload the display driver and Samsung software and try to profile with Eye-One Match. We'll see.
    Did you get those profiles I uploaded in my post from 02:00 p.m.? I'd like to know what curves are inside them.
     
  62. Hmm this is interesting. I've got an utility which tells me whether a custom gamma curve (LUT) is loaded or not. Under Windows, any application can change the gamma curve globally. Apps that do that are e.g. video games, video players, and also Eye-One Match software, when it loads a calibrated display profile. Only one curve can be loaded at any time however. That means that when you run a game on a calibrated display, you lose the whole calibration until you load that custom gamma curve again (which usually means a system reboot).
    After calibrating my display with Samsung software, no custom gamma curve is loaded. That was my confirmation for the claim the LUTs are stored in the display.
    Now I've tried changing gamma in my nVidia graphics card drivers. Although the display has obviously changed, the utility still indicates no custom gamma curve. Hm. So the video card drivers can override Windows' gamma behavior too.
    Okay. Both nVidia and Samsung are trying to get rid of dependency on Windows' gamma curves. That's a good thing, but also a bad thing, because we can't debug such a behavior independently on each other. Now we've got three places where the gamma curve is changed when viewing the image: Windows gamma (tied to the display's ICC profile), nVidia gamma and Samsung calibration.
    We can unload the Windows gamma curve (Eye-One Match does that before calibration on purpose), but who knows what curves are nVidia and Samsung actually applying? Where does nVidia store that gamma curve?
    Sure, if I screw up the gamma in nVidia drivers, I can get them back using Eye-One or Samsung calibration. But that's hardly systematic.
    Samsung, on the other hand, provides a calibrated display, but without a special profile for applications to sense what kind of a display it actually is. Again not systematic (plus it started this whole debate).
    I see why you guys take the short way out: use a Mac, calibrate and move on. Obviously all this exists just to override Windows' color unawareness.
     
  63. Illustration screenshot attached. Display gamma curves are modified, the display has changed, yet the gamma sensing utility can't detect it. Who knows what Samsung is actually doing...
    00VU7o-209239584.jpg
     
  64. From the Natural Color Expert section of the pdf manual with my comments in italics:
    A Color Calibrator (huh?) is provided which allows the color characteristics of your monitor to be calibrated according to your needs. With Natural Color Expert, you can create an icc profile for your monitor only. If required, you can also create an icc profile with some changed monitor characteristics. Save this icc profile to your PC and as the color mode (huh? Is this a LUT?) of your monitor also, which allows you to view colors in a defined color space just by changing the color mode of your monitor. (all this without measuring anything, not a good idea) You do not need to use a separate CMS program such as Photoshop. (You'll have to test this by viewing in a non-color managed app the standard PDI color target whose RGB numbers are written in AdobeRGB. The skintones better not look magenta, dull brown or saturated orangy tan)
    NCE seems to do the same thing as the video driver software like Nvidia offers within the Window's system which basically allows the user to manipulate contrast, RGB curves and other tonal settings through the video card which is also similar to what you can do with an eyeball calibrator.
    What's not clear is if Samsung's NCE software is manipulating the video card to come up with the adjustments shown in the interface and then upload those adjustment settings to the display's internal 10 bit LUTs which is what I'm wondering the mode change is about. Wonder what options are available in the display's OSD menu and if it can conflict with the NCE software. It's not clear what the software is manipulating when you adjust those sliders, the video card or the display's LUTs.
    Alex, all four of those profiles have no embedded LUTs which is understandable since it was created on a Windows system so I can't show them to you. LUTs are embedded in the form of 'vgct' tag within the profile on Mac systems as shown in my previous Colorsync Utility screenshot.
    I examined all four in Colorsync Utility and compared XL20-adRGB to AdobeRGB98 and their XL20-sRGB to original sRGB and all their numbers are slightly off from the real Photoshop working spaces. The 3D gamuts between XL20-adRGB and XL20-emu are a tad different but not enough to show a difference. I understand these four profiles were not generated by measuring the display with your X-rite calibration colormeter and software, correct?
    I'm providing screenshots showing the differences between the two Adobe profiles showing the differences in the amount of line tag descriptors embedded in each one. The other screenshot shows a 3D gamut model comparison between your XL20-Calibration and XL20-adRGB. Your display's gamut is way bigger than AdobeRGB. Not sure if that's good or bad.
    00VUCH-209285684.jpg
     
  65. Here's the 3D gamut model comparison:
    00VUCL-209287684.jpg
     
  66. Thanks Tim.
    Hm. The gamuts in that 3D chart should be roughly the same, since that XL20_adrgb has an arbitrarily limited gamut (what we call Alex LED RGB). Also you're saying the differences between XL20_adrgb and XL20_emu should be invisible? That's what I intended (the target specs are the same), but the monitor shows a visible difference between them. This all probably means that my XL_xxx profiles are just synthetic and don't really describe the current state of the display. Probably only Samsung software can interpret them correctly. What do you think? Can you show how XL20_calib looks like? That one should contain the full monitor gamut.
    These are the profiles created by NCE after measuring the display using the colorimeter. After they are created, NCE offers the option to upload them to the panel. Or perhaps not *them*, but the differences between them and the current state of the panel. Does that make sense? I've recorded a video of this whole process but conversion to SWF went kaboom, so I'll post it tomorrow if you're interested. The measurement process is however similar to that of X-Rite Eye-One software (altering color triangles on a blank screen).
    Your questions:
    color mode (huh? Is this a LUT?)
    I thought they refer to the gamut, but now I'm not sure.
    (all this without measuring anything, not a good idea)
    No, the display is measured by NCE and colorimeter and these ICC profiles are the result.
    (You'll have to test this by viewing in a non-color managed app the standard PDI color target whose RGB numbers are written in AdobeRGB. The skintones better not look magenta, dull brown or saturated orangy tan)
    I don't have the PDI one, I've got this one: http://img693.imageshack.us/img693/3165/mmcntechnicaltarget07fo.jpg (not sure if the embedded profile survived upload to imageshack). It looks good, plus it also looks exactly the same in unmanaged XnView and colormanaged Firefox. That should be the case, right?
    What's not clear is if Samsung's NCE software is manipulating the video card to come up with the adjustments shown in the interface and then upload those adjustment settings to the display's internal 10 bit LUTs which is what I'm wondering the mode change is about. Wonder what options are available in the display's OSD menu and if it can conflict with the NCE software. It's not clear what the software is manipulating when you adjust those sliders, the video card or the display's LUTs.​
    It's manipulating the panel directly, not telling the video card or Windows. So it's not using the same ways like e.g. Eye-One Match is using, but the result what I see should be identical at the end. And it is, too.
    The OSD is inactive in these special modes. I can only enter the desired data within NCE (as seen on the screenshot in my original post) and the display is then calibrated to those specs.
     
  67. I read further through the rest of the NCE section of the pdf. I think I figured out what it's doing, maybe.
    There is mention of Settings in the Calibration Tab that are performed before measuring with Samsung's colormeter that include setting brightness, color temp appearance and gamma adjustments set by sliders in the software interface which later get downloaded to the monitor which suggests NCE is communicating with the display much like you'ld do if adjusting settings through the OSD menu except you're doing it through software. Still not sure if the video card is being utilized when adjusting these settings.
    The colormeter then just measures the response and builds a simple matrix based profile without building and loading a LUT adjustment like EyeOneMatch does. The Settings downloaded to the monitor as it says in the pdf acts as the LUTs which isn't a very precise way to put a display into a standardized state. If the display ages you're left with those three slider adjustments to bring it back into specs.
    Keep in mind these pdf instructions are based on using Samsungs colormeter, not X-rite's. X-rite builds a LUT that requires a LUT loader which X-rite furnishes upon startup. Samsung utilizes user adjusted Settings as its LUTs I'm assuming because there is no mention of creation or downloading of LUTs in the entire pdf.
    Regardless of this it doesn't seem wise to use X-rite LUTs in combination with NCE since it requires the brightness, color temp and gamma to be set with its sliders and downloaded to the monitor.
    NCE is like calibrating and profiling using a combination eyeball calibrator slider adjustments and measuring the response with the colormeter which explains why the tag descriptor numbers in the XL20 profiles examined in Colorsync Utility were so different from the original AdobeRGB98 working space.
    And I'm still not sure or clear about this. It's the best I could derive from that pdf. The person who wrote it should be fired IMO.
     
  68. BTW don't confuse yourself with those profile names. I believe your 3D chart shows XL20_adrgb, not XL20_calib like the caption suggest, right? But that's fine, because this is very informative. Can you also show the XL20_calib? Thanks.
     
  69. I think you'll find answers to your questions in my previous post (we've typed simultaneously). Just to be clear: those settings in NCE are only the target you want to achieve. In calibration mode, that's birghtness, black level and gamma. In the 3 emulation modes, you can also specify the gamut and white point. It's NOT like an OSD - you can't even access the OSD in these modes. The display is then measured and calibrated to achieve those specs. The end result is the same as if it'd be calibrated using X-Rite software. BTW the colorimeter is made by X-Rite and is most likely the same thing as Eye-One Display 2. I'm actually using X-Rite drivers, not Samsung's.
    Anyway, good night. Need to get up in 3 hours.
     
  70. Can you show how XL20_calib looks like?​
    The white frame in the screenshot of the 3D gamut model is the XL20_calib profile. As you can see its huge compared to the smaller color model within the white frame which is the XL20_adRGB. I got mixed up in calling it AdobeRGB98 which wouldn't make any difference anyway since their gamut models were identical. They are simple matrix profiles without a LUT.
    You can download the PDI color target by scrolling down close to the bottom of this page:
    http://www.gballard.net/psd/srgbforwww.html
    That target you supplied doesn't have enough memory colors in the form of real objects. The skintone of the model looks a peachy beige neither over or under saturated with just the right tinge of yellowish light brown to look normal, healthy and balanced.
    These are the profiles created by NCE after measuring the display using the colorimeter. After they are created, NCE offers the option to upload them to the panel.​
    According to the instructions in that pdf the only thing that gets uploaded to the display are the slider adjustments applied in the Calibration Tab where Brightness, Color Temp and gamma are set. This is where the wording gets confusing because they call the whole process a calibration and combine it with the creation of the profile which is matrix based. That matrix based profile can't be uploaded to the display or else Photoshop can't access it to accurately preview color managed images unless a copy of the matrix profile is also uploaded to the display.
    It's not clear. I think a calibration session is what's uploaded to the display and is tagged to the matrix profile to let you know when the next calibration is due.
     
  71. This is what gets uploaded. There's no indication or proof NCE is building LUTs to load in the display.
    Brightness :
    The target brightness: Select your personal maximum brightness.
    Adjust Black Level : Select a black level.
    Color Coordinate :
    White Point : Directly enter a coordinate, or move the white slider
    on the blue bar to the required value position.
    Gamma : Select Gamma values for the R, G, and B components.
    Does the screen change with each adjustment to these sliders?
    BTW don't adjust each RGB gamma or else EyeOneMatch software will have a lot of kinks to sort out when it builds the required LUT to ensure neutrality down the entire tonal scale.
     
  72. Forgot to take a leak before going to bed. Had to check back while passing by the PC too.
    Tim, that PDF is confusing you because you aren't seeing the entire process. I'll show the video tomorrow. They also use shitty selection of words. By 'settings' they mean 'calibration target' (what you want to achieve). 'Calibration' is one of the tabs in the NCE interface. The other is 'emulation', which allows to emulate sRGB, AdobeRGB and a 3rd ICC profile of choice. Thus, the display has 5 modes: Custom (ignore it), sRGB, AdobeRGB, 'Emulation' (that 3rd profile) and Calibration.
    Also what you're not seeing in the PDF is an actual measurement taking place (the same thing you're seeing when calibrating using Eye-One Match software). It'll be clear when you see the video of the process.
    I got mixed up in calling it AdobeRGB98 which wouldn't make any difference anyway since their gamut models were identical. They are simple matrix profiles without a LUT.​
    Ah, OK. So you agree it makes little difference whether I use AdobeRGB.icc or my XL20_adrgb.icc for whatever purpose I need the specified gamut? Back to this later.
    Night...
    P.S.
    Does the screen change with each adjustment to these sliders?
    No. That's just the target settings you want to achieve.
     
  73. Use AdobeRGB if your work is going to press (litho-offset). Go with sRGB if your work is going to a Walgreens photo lab, or a web site.
     
  74. The Uniformity section confirms and corrects my misunderstanding what's going on here.
    The colormeter (and you must use it to affect change in the Uniformity TAB) measures the response in various selected areas of the screen that aren't uniform and makes adjustments I'm assuming to the 10 bit LUTs. I don't think you can select different areas of the backlight to do this.
    This is the wording that confused me:
    When exiting the program (NCE) after saving the changed settings as the XXX(TAB) mode, the color mode of the monitor will change (does this mean the appearance or just the mode?) to the just saved current XXX(TAB) mode.​
    Since you said the screen doesn't change when making these settings with the sliders within the NCE interface also tells me the display's internal LUTs are being manipulated during the measuring process, just not shown.
    I'm trying to determine if the NCE software allows internal LUT manipulation similar to NEC's SpectraView software on its brand of displays. I guess it is. The manual just isn't being clear about it.
     
  75. Yep the display can calibrate it's own uniformity as well. Fancy. I think this is done by changing the LED backlight rather than the LCD panel. LED backlight needs to be cut into groups and those groups need to be controlled individually. Samsung is probably taking this to the next level. I think it works too well to be just another LUT thing, but it's just a guess.
    Anyway. That display documentation you have is rubbish; fortunately I found the NCE help file is much more helpful, so here's a PDF copy: http://ulozto.net/3636702/nce-help.pdf
    I'm also working on the video so you can see what the calibration looks like because it may not be obvious unless you see it. I want you to know how this display works before I start picking your brain :)
     
  76. OMG. I am teaching photograhy forums. I show some of my large prints at each. I get ooohs and aaahs every time I bring them out. I have a Samsung 22 inch monitor. It matches my printer very closely because I have a colorimeter and I have significantly reduced the factory brightness and contrast on the Samsung. I use AdobeRGB not Samsung RGB. I don't understand one damned word of this debate and I don't care to. In my blessedly ingnorant state I get professionally done salable prints. I want nothing more. I have been on the practical side of this business for a number of years. It doesn't make sense to me to make this so complicated. BTW way what I know about color managment came from Andrew Rodney and he makes sense to me and I can understand what he writes.
     
  77. At this point, what I do understand is, there’s no way I’d buy this Samsung based on the convoluted process and instructions they are passing onto their customers. Color management is tough enough so KISS is my motto.
     
  78. Andrew, I would not buy another Samsung because the controls on the bottom of the monitor are extremely difficult to use, I don't think much of their software for controlling anything but the actual image and how it calibrates is fine IMO. Once I get a good calibration I don't screw with it unless I have to. I don't understand their instructions either. I just ignore them.
     
  79. What are you two talking about now? I'm not Samsung salesman and I'm not forcing you to buy this display. But we'll see what you'll both have to say when this method is ironed out and industry standard.
    I use AdobeRGB not Samsung RGB. I don't understand one damned word of this debate and I don't care to. In my blessedly ingnorant state I get professionally done salable prints.​
    Cool. Be that way. I'm a journalist and I like to know how new things work. And you know what, explaining complicated technology to common people is much easier than explaining simple things to people like you and Andrew.
    Color management is tough enough so KISS is my motto.​
    Be it. Your problem is, you won't accept any other ways to achieve the same result. The funny part is, this whole thing is supposed to make color management EASIER and at the same time make WG displays usable for normal work outside color managed stuff. EASIER. No, it's not there yet. Have digital cameras been so easy to use 10 years ago?
    But again. I'm not forcing anyone to do anything. I'm trying to understand what needs to be ironed out and so I've asked a series of questions and the only thing I got from you is ridicule, while the whole problem is you don't understand what am I trying to debug.
    You have the means to help me debug. like those 3D charts Tim provided. That's what I call helpful! Not your off topic comments.
     
  80. we'll see what you'll both have to say when this method is ironed out and industry standard.​
    I’m not holding my breath and neither should you. You ping us if and when this happens.
    You have the means to help me debug.​
    Oh no, I’m done here.
    I'm a journalist and I like to know how new things work.​
    As a journalist, you should contact Samsung first, PhotoNet way down the line. When you get the answers that “satisfy” you and when this product is a “standard” (whatever that’s supposed to mean), you know where to get hold of us.
     
  81. Don't worry. When WG displays are standard, all the producers will have to come up with ways to make them usable for other stuff than just photo imaging. Everyone understands that, except people who think everyone else necessarily has an A2 printer sitting on their desks.
    you should contact Samsung first, PhotoNet way down the line​
    Yeah. But not because Samsung ever gets technical with journalists, but because Photo.net has ever bitching members like you who post irrelevant unrelated stuff instead of actually wanting to help. What I don't understand is why you wasted all this time of yours and mine when you never wanted to understand the problem in the first place. Seriously no idea. Never mind. As you say, move along.
     
  82. What I don't understand is why you wasted all this time of yours...​
    Me either, especially after being called a Dick!
     
  83. Andrew, you are a seriously patient fellow.
    Alex, if there were any justice you'd be paying Andrew (and Tim) for their time and expertise rather than ignoring answers that are outside your preconceived idea of what is right.
     
  84. Alex. I apologize if I offended you. I was just being direct in expressing my opinion. Serously, I have difficulty understanding the relavence of this whole discussion. One of my best friends was a a longtime newspaper editor. He constantly emphasized the importance of simple direct statements. In my former life I was head of a major technical research and system development effort that has US and worldwide impact. You probably use today what we worked on ten years ago. Your technical writing (maybe it's the language barrier) leaves a lot of things unanswered to be kind. Good scientific exposition reduces itself to the simplest way of expressing your ideas. To gain support you have to write to people who don't understand this as well as you do. This whole discussion was so disjointed that it is hard for anyone to follow. Let me ask you some questions about what you are writing about. First, what is the problem that the technology you are talking about tries to solve? After you state the problem and its impact then discuss some means to solve it with some examples to support your contentions. Then draw some conclusions, and finally make some recommendations. If you don't conform to a a broad structure that logically goes to supportable conclusions you won't be understood. You also make some assumptions about where technology is headed without well supporting those assumptions. If you had led us down a logical track and maybe this discussion wouldn't have gotten so personal. You are a journalist. Andrew is more of a scientist. I have led scientific and technical programs that require much more than journalistic depth to get at real answers. BTW Andrew Rodney has published on the subject of color managment and is recognized for his work in this area. I sincerely think he deserves a modicum of respoct. Again, Alex I did not mean to add to the hostility that was already attendant here and I apologize. I have supervised many scientists and technologists to try achieve some significant advance in the field in which I worked. Even though we disagreed among us we remained civil. So, with the utmost respect for what you are trying to say, I say I don't fully understand the issue and it's impact upon photographic technoloy nor do I understand the basic benefits of what you are espousing. You can credit that to my inabiltiy to understand what you are saying but I have read a lot of scientific papers that I have understood. Best of Luck to you.
     
  85. First, what is the problem that the technology you are talking about tries to solve? After you state the problem and its impact then discuss some means to solve it with some examples to support your contentions. Then draw some conclusions, and finally make some recommendations. If you don't conform to a a broad structure that logically goes to supportable conclusions you won't be understood.​
    Thanks for the concise rundown on this. I'll propose a conclusion as I did on my last post. Is the NCE software and Samsung display hardware operating the same as NEC's Spectraview calibration software and hardware? To be more clear can the Samsung package manipulate the internal LUTs using its NCE software the same as NEC's calibration/display package?
    Was hoping Andrew would chime in on this because he's more familiar with NEC displays and Spectraview software than I am. I can't afford to buy either the Samsung and NEC and its calibration packages so I'm trying to understand if both do the same thing during calibration and profiling.
    There's too many holes in Samsung's explanation and wording to know absolutely sure what it's doing.
    Alex, can you do the PDI color target experiment where you force the XL20 into emulating (visually) the AdobeRGB space using the NCE software, load the XL20_emu matrix profile in the Window's system, then view the PDI file in both Photoshop and a non-color managed app and see if they are identical in appearance? If they are identical then the XL20 package is doing what it advertises.
    And oh, I just remembered, disable ALL system based LUT loader so no curves are downloaded to the
    video card upon start up.
     
  86. Also, if after disabling all system based LUT loaders, if the NCE software chokes, crashes or offers a warning box stating it needs them, this would indicate it's NOT manipulating the XL20's internal LUTs at all.
     
  87. OKAY EXPLANATIONS FIRST
    First, what is the problem that the technology you are talking about tries to solve?​
    First let's define what is Samsung trying to solve. Two things:
    1) have the calibration data (LUTs and other related things) stored directly in the display. So for example Windows can't screw it up like it often does with a usual profiling process. Eizo, NEC and HP are doing tihs as well.
    2) have five displays with different characteristics (incl. gamut) in one. These are:
    a) the natural state of this display, but not calibrated or externally calibrated (this is what I used at first, until I begun to use the following);
    b) the natural state of this display, but internally calibrated. If you want to edit pictures while taking advantage of the full display gamut.
    c) three emulated displays: sRGB as according to specs, AdobeRGB as according to specs and another emulated display. This allows you to view images in the specified color space even without color-aware applications.
    You see, if someone is editing a picture of a red rose on a normal-gamut display and saves the image without a specified color space, and I view it on my display in AdobeRGB mode, what I see is probably close to what they'd like the others to see (at least closer if I's use a sRGB display). At least that's my experience; you can't really tell what the others intended when you only have an untagged image. If they'd want other to see some EXACT colors, they'd tag the image as e.g. AdobeRGB, but then they might not see what they are actually editing, if they work outside their own display gamut.
    It's just... Well, it's trying to help people who don't necessarily use color-aware apps. But don't forget all these modes are calibrated!
    Also it allows to use the display for normal work (like spreadsheets) without having to see the eye-bleeding red color all the time, but still taking advantage of the most (i.e. the specified part) of the display's gamut. I.e. you're not forced to use sRGB emulation for everything else outside of Photoshop. You can use AdobeRGB mode or the other emulation mode. The AdobeRGB mode actually looks very good.
    Okay now, what I am trying to do is to find out how reliable these modes are if one wants them to use for color-critical work. I'm also trying to find out how is Samsung achieving their goal. Because after having two these panels for a year, I do believe it's an ingenious idea.
    I admit I might have explain all this sooner, but I didn't expect what roar are my questions going to cause. It's a series of misunderstandings. When I write 'I work in AdobeRGB mode', Andrew thought I don't even know what I am saying, much less doing. I totally respect Andrew for his work and I'm too sorry we didn't understand each other. You see the funny part is, if someone who hardly knows color management read my first question, they'd probably understand :) I'm used to working with people who don't know much. Work deformation...
    Now that you know what's it about, does my original question make sense?
    Is the NCE software and Samsung display hardware operating the same as NEC's Spectraview calibration software and hardware?​
    I'm not familiar with NEC software. I believe that is doing the same as NCE's Calibration tab and the Samsung Calibration mode. I'm rather trying to debug the emulation mode. Here's a video of what it does: http://ulozto.net/3644980/nce-emulation.swf - the actual process differs a bit (mostly in length) depending on what it is doing - calibrating or emulating.
     
  88. load the XL20_emu matrix profile in the Window's system, then view the PDI file in both Photoshop and a non-color managed app and see if they are identical in appearance? If they are identical then the XL20 package is doing what it advertises.​
    100% identical in unmanaged XnView and colormanaged Pixel Studio Pro (basically a PS clone). It also makes no difference if I use AdobeRGB1998.icc, XL20_adrgb or XL20_emu as my display profile (I can switch them in PPP as well). That's where my claim "I cam just use AdobeRGB1998.icc as my display profile and the display handles the rest" came from.
    Also, if after disabling all system based LUT loaders, if the NCE software chokes, crashes or offers a warning box stating it needs them, this would indicate it's NOT manipulating the XL20's internal LUTs at all.​
    Nothing like that (there's no NCE process even running). Windows LUTs are disabled.
    Also the funny part. On that web site (http://www.gballard.net/psd/srgbforwww.html), there are some roll-over tests.
    Here's what I see when rolling over my mouse over them:
    First test - no diff between AdobeRGB and sRGB. Some difference (in the drop shadow under the yellow logo in right top corner of the test image) when clicking for AppleRGB. I think they forgot to include the drop shadow :)
    Second test - no diff.
    Third test - untagged sRGB is oversaturated. If I now switch my display to sRGB mode, it's the othey way around (but none looks especially right). Most untagged images actually look good in my AdobeRGB mode, because of ways most people edit images.
    Fourth test - untagged AppleRGB is darker (duh). But I can emulate my display to AppleRGB as well if I'd want to :)
     
  89. More. Here: http://www.gballard.net/photoshop/srgb_wide_gamut.html on the bottom, there's another rollover test for untagged and tagged AdobeRGB. The author writes: "Untagged Adobe RGB will historically desaturate and washout on un-managed devices, but the wide-gamut monitors are displaying Untagged aRGB fairly normal — hence people are calling them Adobe RGB monitors."
    Well, I see no difference when rolling over my mouse over there, both in colormanaged SeaMonkey and unmanaged Internet Explorer. So I guess I can say I'm using an AdobeRGB display (when running in AdobeRGB mode).
    Sure, I still need to switch to sRGB emulation in some cases, like if I want to edit *my* pictures for the web and I'm using an unmanaged app for editing.
    But other WG displays don't offer AdobeRGB emulation - for now, so they are unusable for anything except colormanaged applications. They are unusable for the Windows desktop, games, spreadsheets, movies, everything except Photoshop and co. But display producers have to find a way to make the displays usable! Samsung have obviously found one. LG didn't and their first cheap wide gamut display sucked in sales, because the image looks sh*te.
    If other display producers follow Samsung and offer AdobeRGB emulation (even uncalibrated), AdobeRGB can become a new web standard, replacing and killing sRGB. Okay, it might not be the true Adobe RGB (1998), it might be called Alex LED RGB (hardly, heh) or WebRGB or whatever. But I don't think Adobe will complain if everyone calls it 'AdobeRGB mode', like Samsung does.
    If you've missed my explanation, please see my Jan 14, 2010; 05:31 a.m. post!
     
  90. But other WG displays don't offer AdobeRGB emulation - for now, so they are unusable for anything except colormanaged applications. They are unusable for the Windows desktop, games, spreadsheets, movies, everything except Photoshop and co. But display producers have to find a way to make the displays usable! Samsung have obviously found one. LG didn't and their first cheap wide gamut display sucked in sales, because the image looks sh*te.
    If other display producers follow Samsung and offer AdobeRGB emulation (even uncalibrated), AdobeRGB can become a new web standard, replacing and killing sRGB. Okay, it might not be the true Adobe RGB (1998), it might be called Alex LED RGB (hardly, heh) or WebRGB or whatever. But I don't think Adobe will complain if everyone calls it 'AdobeRGB mode', like Samsung does.
    Alex the above is your problem statement that IMO should have started your first post. This then tells people what you are talking about and what you are trying to prove. That should then lead you to a cogent simple description of what Samsung has done. Please elimiinate the acronyms if you want us unwashed to understand what you are talking about. I am beginning to understand that it is monitor emulation of color space with untagged images. You also need to supply definitions of things like LUT. I am afraid you really got into a lot of detail without really explaining your premise first and leaving the unsophisticated like me behind. I have worked with a lot of technical people trying to make them well understood to people who supply money for projects. If sophisticated technical concepts are not well understood by those who supply the money they don't supply it. It is not clear to me whether Samsung has really solved your stated problem. I am not debating your facts only trying to understand what you are talking about by having it explained in a clear logical, structured order as is done in most scientific presentation and reducing the ideas presented into plain and simple language that can be understood beyond the arcane vernacular only understood by a few. After all you are on a forum read by a lot of people who may benefit from what you are describing who don't speak the language of display science. Good luck to you.
     
  91. Alex, that link to the video didn't allow me to download it. "No Free Slot" is what I got.
    Could you take a photo with your digital camera showing both PDI images viewed in each app in a zoom size similar to what's shown on the gballard site and post it here. You can edit it to look as you see it on your display. What I'm trying to nail down is if the display is not only showing this color target identically in two apps but also correctly.
    The reason I emphasize this is because I've been studying and researching as an observer (I've never gotten paid for this) on how well third party applications, meaning anyone other than Apple, Adobe and pro level display calibration packages, implement color management and accurate color reproduction according to strict API-(Application Programming Interface) integration within different OS's and ICC standard encoding since 1998 when I first got into digital imaging research. And I can tell you from my experience that PDI target has gotten screwed up in so many odd ways (especially on Windows systems) due to improper, misinterpreted or corrupt encoding that I just want to be sure by ruling that out.
    The fact as you claim that a display manufacturer such as Samsung can nail the color response characteristics of any given synthetic working space just by having its software manipulate the internal LUTs (Look Up Table) makes me a little suspicious.
    I thank you for posting this topic and the work you've put into it because this is new information if not interesting.
     
  92. Well I got the video to download and play. Doesn't tell me very much. Can't see the effects of the calibration, just the process.
    If you don't mind I'ld still prefer to see the digital camera shot mentioned in my last post.
    Thanks.
     
  93. Alex the above is your problem statement that IMO should have started your first post. This then tells people what you are talking about and what you are trying to prove.​
    Point taken :) I really suck at asking questions. I'm rather used to explain what shutter speed is to people that think 1/500s is bigger and better than 1/125s because they see '500' indicated on their camera.
    With such an unusual question, I really didn't know how to formulate it, so I rather oversimplified (since very long initial post often get ignored). Anyway, since I'm trying to debug this display, I'll be thankful for any input you may have.
    Could you take a photo with your digital camera showing both PDI images viewed in each app in a zoom size similar to what's shown on the gballard site and post it here.​
    I tried but it didn't work well. My Nikon D200 + 50/1.8 produced too much moire even at fairly soft f/1.8. My Sigma DP1 (Foven sensor) doesn't suffer from moire, but it's colors are so much out of whack it I couldn't get the photo to show close to what I see on the display in reality (I only use B&W with the Sigma for a reason).
    Instead, I recalibrated the display using NCE, then immediately measured the result with X-Rite Eye-One Match 3. It measures almost perfect. When I click Before/After (meaning before and after Eye-One calibration), I can see just a tiny bit of difference - I believe meaningless in practice and quite likely to be caused by the calibrator itself. That's the best proof I can give you for now. See the attached screenshot. I was suspicious about Samsung as well, but it seems to work.
    You can download the resulting X-Rite profile here: http://dl1.upnito.sk/download.php?dwToken=dd06634e9899c33d0e9199edd05f80b7 (on that web site you need to wait, click, wait and click). It's much larger so it should contain the LUTs as well. There's also a .txt file, which shows how Windows would see that LUT, and a spreadsheet with chart from that data. You can check and tell me if you think it's precise enough? I'd also like to know how close is this measured gamut to AdobeRGB. I appreciate your help with this debugging.
    I'll now try to load Samsung drivers instead of X-Rite's to see if it changes something.
     
  94. Next session consisting of calibrating with Samsung driver and reading with X-Rite driver shows a similar result, but the curves are around 50% different from the previous measurement. Some parts of the curve show more inaccuracy, some show less.
    Thoughts? Perhaps using the same calibrator for both Samsung calibration and X-Rite measurement skews the results a bit? Hardly, I guess. How much inaccuracy can the calibrator itself provide to the measurement?
    I might add I've done this test before and when visually inspecting the result (Before/After button) I saw no difference, but I also might have not see so well.
     
  95. Hm, I think I figured out the whole problem, including my original question. Now that I've watched both NCE and Eye-One Match doing their calibration job, I've realized that E-OM uses many more colors than NCE. NCE basically only repeats some blue, green, red and white squares. There are some differences in shades, but far less than all the yellows and pinks E-OM is throwing at the display. So the accuracy obviously has got to be much lower. It is definitely 'good enough' for me (no chance I would notice and even watching the test target changing in E-OM doesn't give clue which version is more right), but surely not very accurate. I guess our suspicion about Samsung was valid after all.
    This is stupid from them. Samsung was supplying the Huey calibrator at first, which can't even calibrate gamma. Then they replaced it with Eye-One Display 2, but obviously forgot to adequately upgrade their software. I've just installed the latest version... They did upgrade the software somewhat, since the previous versions showed an image of the Huey calibrator, but not enough. I guess this is the most obvious question I'll need to ask Samsung.
    Still I gotta wonder how come the display is quite accurate as it is in its natural state...
     
  96. Funny :) This is how the display measures when the AdobeRGB mode is reset to factory state. Yeah, there certainly is internal calibration, just not as precise as it might have been.
    So I think what they write about 'coarse calibration' on this forum (http://www.freelists.org/post/argyllcms/i1-Display-bundled-with-Samsung-XL20,13 ) is actually caused by NCE software being too simple. Case closed I guess.
    I'll try to reach someone who knows this stuff at Samsung. Wish me luck. Fortunately these very panels, including my units, are manufactured about 30 kms from my place, so I may drop them a visit, perhaps they know who I might need to contact, heh.
    And thanks to everyone, especially Tim, for your input.
     
  97. When I click Before/After (meaning before and after Eye-One calibration), I can see just a tiny bit of difference​
    Remember when you see the screen change, you are seeing the LUTs loading which is the calibration. The actual ICC profile file is the recorded measurement in the form of a matrix formula that color managed apps use to maintain accurate color in tagged images.
    Since you said you turned off the LUT loader in the OS system, the slight change to the screen you are seeing is the LUTs embedded in the display assuming you are viewing this before and after change in NCE.
    If calibrating and profiling using X-rite EyeOneMatch on Windows I'm also assuming it needs the LUT loader active. Not sure on that, but the fact you are seeing very slight changes to the screen in NCE matches up with the straight line LUTs shown in the first EyeOne Match screenshot. It's telling you the XL20 can easily be brought into specs without a lot of LUT bending. That's the sign of a high quality display or very precise LUT manipulation from NCE.
    Since you can't take a photo of the duplicate views of the PDI target, I'll just describe the colors to verify correctness. The left ear on the baby on the far right should be pinkish compared to somewhat tan forehead with noticeable amount of yellow but not much (a light orangish brown as a better description). The dark parts of the hair should be brown with a tinge of red. The eyes should be a gray blue, not intense blue. The red patch in the X-rite color chart should not have an orange or magenta tint. It should look hot red and not glowing. The PhotoDisc logo at the top should look sort of matte finish warm yellow (desaturated school bus yellow) with a tinge of red and not intensely saturated.
    I'll have to download the X-rite profile later. The page isn't loading for some reason. Besides I don't think it's going to tell us anymore than we've already figured out. The preview of PDI target color managed through the display and in a color managed app is all that matters anyway. To me I think the NCE is doing a good job, but it's hard to be sure without actually seeing color targets on the display.
     
  98. Since you said you turned off the LUT loader in the OS system, the slight change to the screen you are seeing is the LUTs embedded in the display assuming you are viewing this before and after change in NCE.​
    I'm not sure I'm following you. Windows doesn't have a LUT loader of its own, it only knows when a custom curve is active (and doesn't and can't do anything about it). I can only unload it using an utility I have here, where I confirmed ther's no LUT loaded. When I click before/after in X-Rite E-OM, it is loading or resetting the LUT using X-Rite's own LUT loader. That little piece of software needs to be run on Windows startup for the LUTs to actually take place. I believe that LUT is also embedded in the created ICC profile itself.
    But you've came to the same conclusion I have: X-Rite E-OM is telling me that after NCE's calibration, the display needs a little bit more of calibration.
    What you describe about the PDI picture is exactly what I see. The way you described the other image I've posted a few days ago (the girl with a desperate need of a tan) is also what I see.
    Yeah, I think Samsung calibration is 'good enough'. But I'd still like to see how the display's gamut as measured by X-Rite conform to AdobeRGB gamut. So I'd like to ask you to check that profile when you have time.
    For now I'm attaching the LUT curves from the X-Rite measurement as seen by Windows and brought into a larger graph. Still looks quite straight, though not completely perfect. Would you have the need to calibrate a display which measures like this?
     
  99. Mirror for that profile: http://ulozto.net/3653234/x-rite-profile.zip hope this works.
     
  100. Alex, I have to use the VIP download. Why don't you just use YouSendit.com to host your file? It's free with a simple interface without all the script laden clutter.
    Other than that those are acceptable LUT curves in that x-rite curve blowup. In fact they're better than a lot of LCD display's I've calibrated. A CRT I used to have delivered a flatter and more even LUT but that's because CRT phosphor's spectral color characteristics were well defined for colormeters to measure. LCD's have two hardware elements that emit separate spectral responses colormeters now have to deal with which include the CCFL/LED backlight and the separate RGB phosphor colorants that form the individual pixels on screen.
    This Colorsync User's list discussion:
    http://lists.apple.com/archives/colorsync-users/2009/Nov/msg00175.html
    ...by a color scientist testing precision and repeatability on a wide range of colormeters on the market is where I got this information from. Note the comment that perceptively on screen after calibration all colormeters worked well except differed in how well they could calibrate the shadows. As usual with color scientists and most color guru's none produce photos showing what they are talking about. I have no idea what Delta E difference looks like so most of that color science stuff doesn't translate to real world usability. It really isn't that hard to take a photo of a display and edit it to look correct. I do it all the time and get quite accurate results as others I've seen do posting on the same subject.
    To give you an idea the precision required in a LUT just create a B&W gradient in your imaging app, open the curves tool and try to duplicate those curves for each RGB channel and see how it affects neutrality throughout the entire gradient. I guarantee it's going to get real ugly fraught with multi-colored banding.
    About the LUT loader I wasn't referring to Window's own loader, but X-rite's as you've indicated. I'm not quite sure how and when NCE and X-rite implement this within their software. I was reading the FAQ over at Integrated Color's Discussion forum, a company that makes Color Eyes Display calibration software and the owner indicates Window's must load the LUT for the video card at startup or the LUT for the display for calibration packages that manipulate the displays internal LUT through DDC (Display Data Channel) communicated through the display cable connection. Not sure if NCE is utilizing this function at startup or doing it live within its software and maybe X-rite as well but for the video card without a restart.
    Just know this you can't have both video and the display's internal LUTs loading at the same time and I'm not sure if its possible to know if and when this is happening.
     
  101. Here's some photos of user's displays having uniformity issues with the new iMacs:
    http://forums.macrumors.com/showthread.php?t=824940
    Looks like they're being pretty picky, but I'm glad my 2004 iMac isn't as bad.
    Hope your XL20's white balance is more neutral.
     
  102. Just a note about the AdobeRGB space Emulation profile built using NCE, that software is having to build additional color tables separate from the simple LUT curves that control neutrality similar to what's shown in the X-rite png demo.
    AdobeRGB emulation requires the hue and saturation color levels be reduced by comparison to the XL20's native gamut which is much larger and more saturated. If this software isn't doing this then it would be impossible to get identical previews of the PDI color target in both non-color managed and color managed apps. With this in mind then the NCE software has the ability to adjust the hue and saturation of the display internally to fit the AdobeRGB color response. Just want that clarified.
     
  103. Sending you the file via yousendit. I usually just use my FTP, but don't want to do that for a public forum. Well I could just send it directly to your mail, but... Ah well :) The important thing is you'll get the file and can check how similar is the measured gamut to AdobeRGB. It should be quite close.
    With this in mind then the NCE software has the ability to adjust the hue and saturation of the display internally to fit the AdobeRGB color response. Just want that clarified.​
    Yes. Actually now when I think about it, it's possible that NCE is doing just that (limiting the gamut), but only very little in terms of calibration. Hm. That would make kind of sense. I have to calibrate the display using the Calibrate tab and then measure the result, perhaps that will give a better result.
    From the whole NCE package, I get the feeling the engineers had something much bigger in mind when designing it, but didn't get the whole job done. For example, after calibration, NCE saves an .icc profile as a file. Why is that even necessary if it always contains the same information? There is a missing link somewhere.
    And OK, I'll give it one more try with that photo picture of the display, but you know that's not very accurate, right?
     
  104. Not sure if NCE is utilizing this function at startup or doing it live within its software and maybe X-rite as well but for the video card without a restart.​
    I don't think that explanation you got is completely correct. Windows doesn't need to be restarted for the LUT to take place. The LUT (usually referred as a gamma curve btw, which is not precise) can be set or reset at any time. The problem is, most tools that apply the LUT only run on startup, so it seems like it can only be done at startup. Not really; when you run a videogame, it applies its own LUT and resets it on exit. If the game crashes, Windows gets stuck in any LUT the game has set. One of the problem is, applications don't read the current LUT when they change it, and don't bother changing it back; they just reset it. So if your display was calibrated using a custom LUT, that is now lost. A usual resolution is restarting Windows, but it's not technically the only way to do that.
    The point is, X-Rite is using the standard Windows API for setting the LUT (just as games etc.). Whatever is Samsung (or nVidia) doing, they are using a completely different method, so Windows, X-Rite or any other application can't see or affect it in any way. So basically when I click before/after in X-Rite E-OM, it's applying its own LUT on top of Samsung's LUT. Is that what you meant?
     
  105. LUT for the video card at startup or the LUT for the display for calibration packages that manipulate the displays internal LUT through DDC (Display Data Channel) communicated through the display cable connection.​
    Oh, wait. Now I got it. No, I don't think that's the case here either. When I reset the display to factory state, it gets a very strong magenta tint (as I saw yesterday). When I calibrate it using NCE, it becomes neutral. When I exit Windows and boot Linux on the same machine, it stays neutral. So most likely the changes are stored directly in the display and not drivers.
    Samsung is using DDC as well, but apparently just for permanent storage in the display.
     
  106. OK I figured out how to make a good photo of my display, so here it is. The two screens are XnView and Firefox. I've confirmed FF is color managed, because at first it showed a slightly different image - since I forgot to switch color profiles yesterday (the X-Rite one was still active). Here the images are matched, the slight difference you may see is because of the different resizing algorithm used by Xn and FF. They look identical at 100% size. On this photo I tried to match the right one (FF) to the original.
    So when I see this photo and the original PDI target from at least 2 m distance, they look fairly the same except:
    - the white space is clearly pink (this disappears when I watch it for a minute, as the eyes adapt)
    - the Asian lady and the rightmost baby still have a slight orange tint to them. The two middle babies also bu much less so, hardly noticeable (picking hairs here).
    - I use neutral gray as my Windows skin, but here it's blueish/purplish
    - the yellow PhotoDisc logo is cooler and less saturated
    The color checkers look fairly normal, the blue tones are a tiny bit darker. Black is also a little washed out. I didn't try to match the rest of the image.
    You should see the same I guess?
     
  107. Photo.net gave me an error, here's the photo: http://img3.imageshack.us/img3/5195/cam8103.jpg
    BTW it's made in AdobeRGB colorspace saved and tagged as AdobeRGB... Hopefully (I'm using an alpha version of RaWTherapee, so I hope it works).
     
  108. They look identical but they look incorrect. There's way too much brownish yellow in the skintones and the red patch is way too orange. The baby on the far right is suppose to have a pinkish peachy tan with a bit of noticeable yellow.
    I take it you didn't edit the image. However, this is the typical digital camera rendering when shooting a display without editing especially if it has an LED backlight like the Samsung's which is often quite blue and spikey in its spectrum that throws off the camera's AutoWB.
    Thanks for the effort.
    Below is a shot of my iMac of the two babies on the right. The file is in AdobeRGB so it needs to viewed in a color managed app. It shows how peachy pink with a tinge of yellow should be in the tan baby. To test accuracy I get identical CMYK numbers as in the original.
    00VVl7-210291584.jpg
     
  109. I've edited the image, but the orange 'tan' is something I couldn't eliminate (see my previous post). But in the middle two people I see it as negligible. When you mention the red patch looking orange... Um, i don't see a difference, but the RGB color values agree with you, it should show a little orange. I just don't see it, so it's either my eyes or the display not showing the difference. The latter might be a limitation of the display when it's trying to emulate a smaller gamut - which is happening mostly in reds. Some precision might be lost in there.
    Your photo actually seems to be a closer match to what i see, but the other way around - a little cooler. I'll try again in calibration mode.
     
  110. No, I see it now. I've swapped the AdobeRGB1998.icc profile for X-Rite's profile from yesterday as my display profile and compared the image in colormanaged FF. Your image is a much better fit and mine shows the symptoms you describe, including the orangeish patch.
    So I guess when I calibrate my display with X-Rite SW, it's a good match to your calibrated display. Unless there's a difference between our calibrators at place, I guess Samsung calibration is less accurate. It doesn't show much with the X-Rite's test image, but the difference is more visible on this PDI image.
    It's a bummer Samsung didn't go all the way, considering they use technologies for multiple profiles in lower displays. Thanks for your help with the troubleshooting Tim!
     
  111. This is not an easy image to edit for accurate reproduction as well as getting a match on a commercial presses. That's why it was created. I've seen that orangey, peachy tan baby image printed on retail packaging for baby accessories and it was a dark, dull brown color, way off.
    So are you saying the NCE emulation mode and supplied system profile aren't as accurate as the X-rite?
    Below is a screenshot converted to AdobeRGB off my system of all three versions-the original, mine I posted above and yours.
    00VVvN-210409584.jpg
     
  112. This seem to be the case. Check this attachment - this is how Calibration mode measures. The difference in the darks is very obvious by naked eye (also in Emulation mode, which fired up my original question).
    It's obvious that AdobeRGB mode is controlled by different circuitry than Calibration/Emulation mode. That's nice, but if Samsung would want them to look the same (apart from gamut), their calibration process should catch this difference.
    But, perhaps the AdobeRGB mode is somehow specially tweaked to conform to Adobe's specifications or something. Not sure. I can imagine it helps when visualizing what the print will look like. Why do you think they would be doing it?
    Also, to be fair, I've repeated the X-Rite measurement 3 times and the error of measurement seems to be quite high, almost as high as the measured difference. Also, when I click the before/after button, the difference in the PDI image is mostly in the magenta/green tint (just as indicated). But if you ask me, neither looks especially right. The green added by X-Rite makes the skin look a tad unhealthy, while the Samsung red tint looks like from a Canon digicam. The ideal is somewhere in between, but I'd pick the Samsung if I had to. I often have lunch at Vietnamese place and the Asian lady is closer to those guys.
    But without having this picture printed as reference, or having you or someone with similar expertize look at it, I can't really know. It can be just the way my eyes adapt to the LED backlight. It reminds me about the story with the Hahnemuhle Baryta FB paper - http://www.luminous-landscape.com/reviews/printers/tale.shtml
     
  113. That attachment.
     
  114. Color temp/cast perception whether it's in the paper or light is not going to have that noticeable difference compared to wrong looking hue/saturation levels which requires selective color table manipulation by mathematical calculation to micro adjust the electronics of a video card and A/D converter affecting voltage going to a display. Every pixel on the screen has a voltage rating.
    Believe me this is more complicated than rocket science as confirmed by an actual rocket scientist. Albert Einstein would be drooling and scratching his head if he ever saw the math going on under the hood of a color managed system and how it affects electrons and colorants.
    A lot of folks I've noticed who think they have a color cast in their image OR display in reality have an over saturation and/or contrast issue. This happens to me quite often when I edit skin tone and pastel colors to match or look correct. For instance the first edit I did to the digital photo of my display posted above was to back off the saturation and contrast where most of the hue/saturation corrected itself. HSL and custom DNG profile took care of the rest. Color temp correction only affected slight increases of perceived yellow.
    These threads below show photos of an Epson 2200 print of the PDI target . The print is from a $10,000 Sony Artisan calibrated professional digital imaging workstation owned and operated by Peter Figen, a professional photographer and fellow color management enthusiast I've had many conversations with in the past 8 years I've been posting on the web.
    http://www.photo.net/digital-darkroom-forum/00UbBI
    http://www.photo.net/digital-darkroom-forum/00Rr6Y
    http://www.photo.net/digital-darkroom-forum/00ULME
    http://www.photo.net/digital-darkroom-forum/00UT1m
    You'll see the skin tone of that baby slightly change but never go way off due to adaptation affecting saturation adjustment. The images look neutral but are really just over saturated caused by adaptation that's making them look different.
     
  115. Another issue to keep in mind as Andrew mentioned at the start of this thread is that X-rite's colormeter may not have the right filters that can accurately measure the XL20's LED backlight and its affect on pixel colorants. That Colorsync List discussion I linked to earlier states this by comparing these consumer grade calibration packages like the one's used here with a very expensive Lab grade spectrophotometer.
    Also the software/driver like NCE may or may not be tuned/written to formulate accurate calculations with any colormeter whether X-rite's or Samsung branded due to quality control issues. It's all speculation at this point.
     
  116. Those LUTs you posted don't tell us anything about how Samsung's software or hardware is manipulating circuitry within the display's internal electronics to control hue and saturation. Those curve LUTs only control neutrality within all 255 gray levels and slight tweaks to color temp shown in the curve offsets in the upper right corner of the graph.
    The only way to test if AdobeRGB Emulation Mode is actually forcing the XL20 to reproduce the exact color response of AdobeRGB is to view an actual AdobeRGB written image according to ICC transform function calculation after a conversion and have it look correct. If it's off in color then the NCE software is not accurately controlling the internal electronics of the display no matter what LUT curve you derive from the X-rite or NCE.
    The pinkish white point is an indication that the filters on the colormeter may not be tuned to the color response tolerances it's designed for as I said previously.
     
  117. As for color accuracy, tomorrow I'll borrow a laptop and calibrate the display to see how they match. Or at least I'll be able to describe what the differences are. The laptop display is crappy, but at least it can give me some idea.
    About that pink tint you see in the photo. When I look at the photo for long enough, that tint completely disappears and matches a white background of the display. Do you experience the same? If not, it might have something to do with the LED backlight and how the eyes react to it. Sometimes I think there's a pink shade to the white, but I only notice I watch a table of gray shades (like the bottom of the PDI image). It's very subtle, only in the white (not grays) and it disappears after a while.
    One thing is certain - my D200 sees the white of my display as neutral when the WB is set to around 4600 K. You describe the photo as warm (after my edits to match it to reality), so that fits. Yet, when the ambient light is 6500 (as measured by both the calibrator and the camera, and in actual daylight around noon), the display is completely neutral, even when I compare it to actual white and grey things I have here, including a 'grey card' cleaning cloth.
    This can have something to do with the LED backlight. Some people who replaced their incandescent light bulbs with LED lights sometimes say that other people look like corpses under that light - even when those LEDs are actually 6500 K.
    So I speculate this is something which can be controlled by our brain thanks to the specifics of the LEDs. It can be something opposite to the effect of ultrasound: we can't really hear 25 kHz sound, but we definitely notice when such sound is turned off.
    I understand calibrators need to be tweaked to specifics of LED (perhaps this is the reason?), but actually Samsung claims this calibrator IS tweaked (I haven't really believed it), so I don't know. I've read some of that discussion about calibrators, but the actual effect of LED backlight isn't really very well described AFAIK. I'll do some searching.
    I'll also try to get my hands on some actual printed color target. I know a distant photo lab that actually makes them for their clients; perhaps they can send me one.
     
  118. About that pink tint you see in the photo. When I look at the photo for long enough, that tint completely disappears and matches a white background of the display. Do you experience the same?​
    No, I don't. I see pink as I should because the RGB numbers say it's pink. I get a reading of 240,220,240RGB in the white part in question. True white on any display should look white even when you photograph it and give RGB numbers that are close to being identical. I get RGB numbers in my shot of my 6500K display of around 221,226,229 a bit on the blue side as it should since camera's AutoWB reference 5000K internally which is warmer compared to 6500K, slightly cooler but certainly not pink.
    The fact you said your shot of your display looks pink but blends in to the white of your display is a red flag that your calibration is way off because 255,255,255RGB white will at least appear white even to a digital camera due to its AutoWB as it does with my DSLR taking shots of my iMac.
    Kelvin number readouts measured with colormeters and digital cameras can't be trusted to represent a specific look because there can be several hues that can be considered a coolish 6500K and warmish 5000K and so on. If your eyes first sees pink then its pink.
    This is why I think your colorimeter's filters aren't matched to the wide gamut response and LED backlight. Discussions on the web including the Colorsync Mailing List discussion linked above points to this as the cause. Or something is getting real corrupt in the building of the profile and how it integrates with the OS or within the internal electronics of the display.
     
  119. I guess. That's probably the reason why I subjectively see the people in the PDI target as slightly reddish (but still look quite natural), while X-Rite software overcompensates and makes them look unhealthy green.
    And the reason why only 100% white sometimes feels pink, may be that it's only when the LCD pixels are completely transparent, some red pollution from the RGB LEDs can come through to become actually identifiable. The calibrator probably sees too much of it and while NCE isn't bothered much, X-Rite cuts down on the red too much. All in all I agre with your conclusions. It seems to be a limitation of the display, but should be handled better.
    Good. This all gives me enough data so I can formulate the questions for Samsung. I'll try to reach someone in the local factory who can perhaps contact me with some engineers far away... Maybe.
     
  120. That's probably the reason why I subjectively see the people in the PDI target as slightly reddish​
    I can make the PDI target reddish in Photoshop on my system by assigning the XL_20 calib profile to the image. The pink of the white of the display is going to cause your eyes to see the skin as green due to adaptation.
    00VWJq-210685584.jpg
     
  121. That's like me switching to Calibration mode in a non-color managed app. No, there's not that much tint in AdobeRGB mode, but it's the same hue. Most likely the gamut in AdobeRGB mode differs from AdobeRGB98 specs enough so it's a little visible.
    You're probably right about the pink white = green people effect. Also, this inspired me to test how other modes of my display compare when calibrated with NCE and E-OM (in colormanaged app, with the appropriate display profile set). I've created these profiles a few hours ago, so...
    Calibration - NCE looks completely neutral with the people looking good. X-Rite profile gives a slight purple tint to the darker areas, visible in the grey table and the people. Also some obvious clilpping in dark areas thanks to the applied LUT.
    Custom - NCE doesn't apply, X-Rite profile adds pink tint to some shades of grey and green to the other. It's all over the place.
    Both these profiles are utilizing the full monitor gamut and it seems X-Rite is doing shitty job calibrating them. AdobeRGB mode obviously behaves completely different to other modes. It's probably buggy with those flaws (mostly pinkish white, maybe imperfect gamut) and X-Rite is trying to compensate? I'm checking this now in complete darkness, if that changes anything.
     
  122. By the way. Tim, again I thank you for all your help and I don't want to bother you anymore, unless you're curious about the outcome of this :)
     
  123. I think you already explained the outcome of this. I'm all turned around as to what actually works but I think what you're saying is the NCE seems to give better results than the X-rite software probably on account that NCE is programmed to handle the wide gamut of the XL20.
    Still not clear if you are using two different colormeters, one from Samsung and the other that came with the X-rite package.
    But if you want to keep us posted on the final results or to summarize, we're all ears and eyes here.
     
  124. I'm using just one calibrator - the Samsung branded one. I've just downloaded the X-Rite software.
    I've used the calibrator (my 2nd unit actually) to profile other displays and it seems to be doing the job fine.
     

Share This Page