Skip navigation
Currently Being Moderated

Monitor profile is defective

Feb 11, 2012 10:36 AM

Every single time I power up CS5 I get an error :

"The monitor profile "Samsung - Natural Color Pro 1.0 ICM" appears to be defective. Please rerun your monitor calibration software.

No other application complain about this.

 

The monitor is a Samsung SyncMaster 226CW

If I go .. Control Panels \ Color Management

I can see that there is a profile "Samsung Natural Color Pro 1.0 ICM (default) file SM226CWicm

It is the ONLY profile listed.

 

As per a previous post here ... I followed advice to delete the profile ..

 

I delete it ... I am then prompted to chose a new profile .....from the list shown there is the "Samsung Natural Color Pro 1.0 ICM"

I select that.

Then next time I start CS ... same error message again.

How do I fix this ?

 
Replies 1 2 Previous Next
  • Currently Being Moderated
    Feb 11, 2012 10:46 AM   in reply to Tafflad

    Before you delete the profile, click the Add button at the lower-left and scroll down a little ways choose the one with Name: sRGB IEC6196602.1, Filename:  sRGB Color Space Profile.icm

     

     

     

    Then delete the Samsung profile and select the sRGB profile and set it as the default.

     
    |
    Mark as:
  • Noel Carboni
    23,514 posts
    Dec 23, 2006
    Currently Being Moderated
    Feb 11, 2012 11:21 AM   in reply to ssprengel

    Actually, just make the sRGB IEC61966-2.1 profile the (default) using the OS color-management dialog and you don't need to delete anything.

     

    And yep, monitor profiles are often defective. It's a real wonder that monitor companies don't know more about this - what an embarrassment to release such a profile!

     

    -Noel

     
    |
    Mark as:
  • Noel Carboni
    23,514 posts
    Dec 23, 2006
    Currently Being Moderated
    Feb 11, 2012 11:45 AM   in reply to Tafflad

    A color profile can be quite complex.  There's a good chance Samsung either crafted the profile with software that a) didn't generate a properly formatted or populated profile or b) created a profile using the very latest new standards, which your Adobe software does not yet know fully how to deal with.

     

    It's a bit surprising that Samsung would not have tested it with Adobe Photoshop, however.  That's pretty much THE preeminent color-managed application.

     

    Would you be willing to send me a copy of the profile, or post the link from which you retrieved it?  Edit:  I've just been over to the Samsung.com web site to try to get a copy of that profile and it's maddening!  They really need to get their act together regarding how to design a web site.  I was unsuccessful.

     

    -Noel

     
    |
    Mark as:
  • Noel Carboni
    23,514 posts
    Dec 23, 2006
    Currently Being Moderated
    Feb 11, 2012 12:38 PM   in reply to Tafflad

    Thank you.  I was able to get the SM226CW.icm file out of that.  I like to collect questionable profiles to test my own software with.

     

    On examination, the Samsung profile doesn't appear corrupt, per se, though it does seem to be missing the CMM type field (it's blank; all the profiles I recall say ADBE, or APPL, or Lino, etc.).  The curves also appear to have a gamma of 1.45, which is pretty low.

     

    Perhaps Adobe is picking up on these irregularities.  Note that Photoshop gives you the opportunity to use it anyway.

     

    I'm happy that when the "use anyway" option is chosen, my own color-managed software performs identically to Adobe's, so the content is interpretable after a fashion.

     

    Since this profile seems pretty different than the standard sRGB profile provided with Windows, and if color accuracy is important to you, you might want to consider getting a measurement device which can calibrate/profile your display.  On the other hand, if consistency amongst the various apps you use on the display is most important to you, the use of the sRGB profile for the monitor may be a good way to go.

     

    Thanks again.

     

    -Noel

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 11, 2012 2:11 PM   in reply to Noel Carboni

    More likely the matrix is non-invertable or doesn't match the white point, the white point not D50, etc.

    ACE does a lot of error checking on profiles, because we've seen so many that ignore the spec.

     
    |
    Mark as:
  • Noel Carboni
    23,514 posts
    Dec 23, 2006
    Currently Being Moderated
    Feb 11, 2012 2:18 PM   in reply to Chris Cox

    Yeah, it's not too new...  This profile is dated 2007.

     

    Funny thing is Samsung knows about this - a search for the error leads one to their site, among others.

     

    -Noel

     
    |
    Mark as:
  • Noel Carboni
    23,514 posts
    Dec 23, 2006
    Currently Being Moderated
    Feb 11, 2012 3:10 PM   in reply to Tafflad

    I personally wouldn't want to use a profile that caused the software to emit an error every time.

     

    How would you rate the relative importance of these things (i.e., if you couldn't have both, which would you prefer)?

     

    • Having Photoshop display things on your monitor that are showing correct color, so that you can send an image off (e.g., to a printer or for display on the web) and have the color come out as expected.
    • Having Photoshop's display of colors match what you see in most other applications on your system.  What you see in one app more often looks like what you see in another app.

     

    Depending on your monitor, because of the current state of the art these are actually mutually exclusive.

     

    Unfortunately, the decision on how to set one's system up for color-management has to be a personal one.  Only you can know what your image editing needs are, what color spaces you prefer to work in, and how you want your graphics work with color-management to fit in with the rest of the things you do with your computer.

     

    One easy answer is to purchase a color measurement device and use it to create your own custom calibration and profile.  Yes, that costs money.

     

    -Noel

     
    |
    Mark as:
  • Noel Carboni
    23,514 posts
    Dec 23, 2006
    Currently Being Moderated
    Feb 11, 2012 3:36 PM   in reply to Tafflad

    When I tried your profile all the whites turned orange.

     

    -Noel

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 11, 2012 8:17 PM   in reply to Tafflad

    Have you actually tried to use a calibrator like i1 pro to calibrate/profile the monitor?

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 12, 2012 12:09 AM   in reply to Tafflad

    The canned monitor profile provided by Samsung may have been defective from the beginning, or the corresponding file may have become corrupted over time, or (less likely) some hardware component(s) may have broken down in the electronic circuitry of the monitor itself.  Can you use some kind of eyeball calibrator that may be available for Windows until you procure a hardware calibrator puck?

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 12, 2012 3:13 AM   in reply to Tafflad

    I googled SM226CW and came up with a model that appears to be wide gamut.

     

    In that case you shouldn't use sRGB as replacement for the Samsung profile - Adobe RGB would be a closer match.

     

    But with wide gamut monitors you buy into a different paradigm (so to speak) than the traditional one. The only sensible thing is really to use a calibrator, use color managed applications whenever possible, and ignore what you see in those that aren't (because it just won't be right no matter what).

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 12, 2012 4:01 AM   in reply to Tafflad

    The type of connection (DVI or VGA) has no bearing on this.

     

    Wide gamut means the monitor can reproduce a wider range of colors - more saturated -  than a standard gamut monitor. The difference is similar to the difference between the sRGB and Adobe RGB color spaces.

     

    Here's one way to illustrate it:

     

    wide gamut2.png

     

    A monitor profile will take this into account and remap the colors accordingly so that colors that fit into the display gamut will display correctly in either case. Colors that are outside the display gamut will be "clipped" to the gamut boundary. But if you use the wrong profile this remapping breaks down, and things will appear either over- or undersaturated.

     

    Does this make any sense?

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 12, 2012 5:18 AM   in reply to Tafflad

    As for software calibrators similar to Adobe Gamma, there is this: http://calibrize.com/

     

    But you would have to start by pinning down the RGB primaries, which these solutions won't do for you. Adobe RGB won't be far off, so if you make the profile based on that it should work reasonably well.

     

    Still, it's a halfway solution. I'd like to repeat what I wrote above for emphasis: Wide gamut monitors introduce a whole new playing field. A lot of things one could previously take for granted are no longer valid. A calibrator should really be included in the deal (which it usually is for the more high-end units).

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 12, 2012 8:49 AM   in reply to twenty_one

    This is crazy making, for me, anyway. What you are saying is that, lacking a proper calibration procedure which ignores the color space in which you work to correctly map the test color to the output of the monitor, one simply employs the color space of your choice. Well, let me tell you that even in a calibrated environment the results may still not be right if the job done is not correct.

     

    A case in point:

     

    I decided to re-calibrate my new Dell. Since I had run, several times, the Advanced procedure, using the i1colorimeter with X-Rite software and found the results consistent, I was satisfied that the procedure was correct. Prints matched beautifully, except those whose gamut range exceeded the gamut range of the monitor, which were very rare. So this time, I decided to use the Easy version. After all, my calibration points were set manually the lat time in Advanced, and Easy does not allow for resetting those.

     

    When I finished and did the before/after check, there was a noticeable shift. Hmmm. So the monitor drifts. Crap! I ran the Monitor validator. Average dE was 0.36. Good. But wait! the Average last time was almost identical, 0.4. Why am I seeing this shift?

     

    So I went back and did it again, this time using Advanced, but, like in the Easy run, I did not touch any of the manual adjustments. When I finished, again the before/after showed a shift but the shift was back to the original calibration a month ago. Average about 0.45.

     

    Well, I knew without even looking what the problem was. Easy did not hold to the tolerances that Advanced did, and when I examined the shifts from target values to actual, for each of the colors employed, I found several colors that were out significantly from their target values, some between 1 and 2. Further, these same colors were off in both Easy and Advanced, but were far closer in the Advanced run; none over dE 0.9 and the cluster around the average value was tighter.

     

    I did a substitution of last month's profile against the current. The image showed the barest shift, almost unnoticeable between the two profiles.

     

    I was much relieved.

     

    Simply put, I cannot see or agree to the notion that you can substitute a color space for the calibration profile and achieve accuracy. I t might work if the monitor had no errors at all; then you are simply limiting the output of the monitor to that of the color space, assuming the monitor will reproduce it.

     

    In short, if the colors aren't right, you are not correctly profiled. Profile the instrument with a calibration device, then set the color space in the app, if allowed, as it is in Photoshop.

     

    One final note, and that is about those colors that are significantly above the average. If you look at a specific patch, it comes with a set of L*A*B values, and the chart shows the variation from those values that constitutes the error for the corrected output.

    That error is also an average, quantifying the individual L*A*B error as an average. However, it is possible to have a consistent dE value for that color, yet the actual color be very different, if the numbers are different. From this I conclude that what I see is the relationship between the colors themselves having as much to do with what the eye sees as the error (too red etc) as the dE value itself. Applying a color space will not fix that.

     
    |
    Mark as:
  • Noel Carboni
    23,514 posts
    Dec 23, 2006
    Currently Being Moderated
    Feb 12, 2012 10:35 AM   in reply to Hudechrome

     

    In short, if the colors aren't right, you are not correctly profiled.

     

    Therein lies the problem, Lawrence.

     

    Not only do tolerances figure here - for one person, it might be good enough that "red" is subtantially reddish, while for another measurements in excess of 0.4 with an accurate colorimeter are unacceptable.  Consumers of imagery may have different tolerance for error than producers of imagery.  Thus there is no one definition of "right".  Some monitors can't even be made "right" at all positions on the same display.

     

    BUT...  Even image producers can vary in their requirements.  Simply put, not everyone has the same needs and goals

     

    For one person, having prints match Photoshop display is paramount, while for another having color-managed apps most often match non-color-managed apps may be more important.  It kind of depends on what you do and what you need.

     

    I'm not saying anyone's wrong or right here - just the opposite:  We should not be using the terms "wrong" or "right" or "correct".  Trying to determine the "best approach for the stated goals" might be a better way to discuss color-management. For that we need to get people to state goals.

     

    In the end, education is the key:  Virtually everyone who finally "groks" color-management - and it's not really that difficult to understand fully - can make the best decisions about settings and methods for them.

     

    Waiting is.

     

    -Noel

     
    |
    Mark as:
  • Pierre Courtejoie
    7,038 posts
    Jan 11, 2006
    Currently Being Moderated
    Feb 12, 2012 10:06 AM   in reply to Tafflad

    I've seen several similar threads on feedback. It seems that Windows pushes monitors profiles, and most of them are rejected by Photoshop/elements.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 12, 2012 11:30 AM   in reply to Hudechrome

    Hudechrome wrote

     

    Let's not overcomplicate this.

     

    The main point I was trying to make re calibration and wide gamut monitors, is that you don't have an immediate reference for the colors if you're inexperienced. It's like free fall, there's nothing to hold on to. All the basic stuff that with a standard gamut monitor displays more or less right, is suddenly wrong.

     

    A calibrator gives you that reference. But you need one more thing: you need to know that with that monitor only color managed material will look right. Anything that is not will simply be wrong and there's nothing to do about it. This is why these monitors aren't for everyone.

     

    So who needs to calibrate? I think it depends on what environment you're in, and what your end product is.

     

    If you're in a production chain where you rely on others and others rely on you, there's really no choice. You just have to be on the same page as everybody else; if you're not the whole production line stops dead. Wasted time and money.

     

    If you control the whole process yourself things are different. As long as the end product, like a print, comes out right, it doesn't really matter how you got there.

     

    If you work for web or other types of networking, you're somewhere in between, but mostly because it's uncontrollable anyway. When color management becomes mainstream on the web, standards inevitably tighten.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 12, 2012 11:39 AM   in reply to Tafflad

    Tafflad, in your situation I would try Calibrize, but make sure you start with Adobe RGB. Use the plain one, I've no idea where the other one comes from.

     

    And also keep in mind what I wrote above: anything that is not color managed, like Windows Explorer and Internet Explorer, will be over-saturated and wrong on that monitor. But Photoshop will display correctly.

     

    One alternative approach is to look for the monitor's sRGB preset in the OSD menus. This will effectively turn it into a standard gamut monitor, and will solve some of the more immediate problems, and will also allow you to use sRGB as the monitor profile.

     
    |
    Mark as:
  • Pierre Courtejoie
    7,038 posts
    Jan 11, 2006
    Currently Being Moderated
    Feb 12, 2012 11:46 AM   in reply to Tafflad

    Keep in mind that if Adobe Gamma was retired, it was because it could also create bad corrections, as our eyes "lie" a lot (they adapt to the surrounding conditions).

    Make sure you have a very neutral lighting in the room, have a neutral grey surrounding for your image. Some use a hood, other create black cardboard frame around your monitor.

    If you are on windows, use the old user interface (windows classic) with grey title bars. On a Mac, prefer the graphite theme.

    A Hardware calibrator analyses only the pixels they see, and are not tricked as our eyes can be.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 12, 2012 12:39 PM   in reply to twenty_one

    Actually, I'm trying to uncomplicate it by insisting on calibration first, and offering reasons why.

     

    Look, I have long experience with calibration issues, from nuclear physics audio to primary instrumentation, even electron microscopes. One  gets into trouble, no matter your needs, if the devices are not calibrated and traceable.

     

    Pierre's comment about Adobe Gamma hits it squarely. It is the reason I invested in hardware calibration in the first place. A few days pulling my hair out sent me to my supplier where I plunked down $250 (in around 2001) and never looked back.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 12, 2012 2:04 PM   in reply to Hudechrome

    Hudechrome wrote:

     

    Actually, I'm trying to uncomplicate it by insisting on calibration first, and offering reasons why.

     

    Yes, that's a good approach

     

    This should put it all in perspective: http://en.wikipedia.org/wiki/Color_constancy

     

    Looking at the monitor you essentially have no context to judge color by, and that's when color constancy kicks in: a fresh green apple will look fresh green even under a strong overall color cast. Put that image on a white piece of paper, however, and the apple is suddenly not so tempting. Color constancy is broken with the context.

     

    This is why we use calibrators. They don't suffer from color constancy. It doesn't really matter if they're not perfect, they still perform much better than the eye and brain do, even under the most favorable circumstances.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 12, 2012 2:17 PM   in reply to twenty_one

    Incidentally, this is why camera calibration is a totally different meal than monitor calibration. The mechanism works in reverse.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 12, 2012 2:21 PM   in reply to twenty_one

    Hudechrome made this point (at length): You can't use a device independent color space profile as the profile of a device (for ******* 's sake). The whole purpose of profiles is to NOT DO THAT.

    Read. Understand. See.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 12, 2012 2:32 PM   in reply to Lundberg02

    colorwiki.com        Please go there before you post again, Tafflad.

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 12, 2012 2:41 PM   in reply to Lundberg02

    Lundberg: Dont "for ******* 's sake" me.

     

    And don't "Read. Understand. See" me.

     

    I know perfectly well how this stuff works. I know perfectly well what happens when you use this type of profile for that and vice versa. It's so basic I didn't even comment on it.

     

    I'm trying to be pragmatic in a situation where the OP does not have and is at present not able to get a calibrator. I'm just trying to keep him floating in the meantime.

     
    |
    Mark as:
  • Noel Carboni
    23,514 posts
    Dec 23, 2006
    Currently Being Moderated
    Feb 12, 2012 3:29 PM   in reply to Lundberg02

    Lundberg02, please try not to make this discussion of color-management confrontational, and understand that what you might feel is wrong in your world view (and given your needs) might not be wrong for everyone.  As I said before, there is no "wrong", there are only intelligent choices. 

     

    I can easily cite reasons why using a document profile for a monitor can be useful and even preferable under some conditions, but I won't.  That's not the point.  It's unhelpful, in color-management discussions, to tell people what to go do, as though your way is the only way.  You have to understand that there is not just one right way to do things, which is why we have configuration options we can all change.

     

    Color-management discussions, for some reason, are prone to degeneration into conflict.  Let's not have that happen here, okay?  It's a complex subject, when you factor in individuals' needs, and everyone has good things to contribute.

     

    Understand that D Fosse could teach university courses on the subject.  He knows his stuff and his advice has been perfectly sound here.

     

    -Noel

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 12, 2012 4:11 PM   in reply to Noel Carboni

    I was not talking to D Fosse and he does know his stuff.  Tafflad needs to get the picture.  We go on at length here when an OP doesn't know color management basics.  Calibrate, calibrate, calibrate are the first three rules.

     
    |
    Mark as:
  • Noel Carboni
    23,514 posts
    Dec 23, 2006
    Currently Being Moderated
    Feb 12, 2012 4:20 PM   in reply to Lundberg02

    Rules?  LOL 

     

    We agree encouraging understanding is best, but throwing out advice as "rules" is a bit strong.  You don't know Tafflad's color needs and you're already talking about rules.

     

    Oh, and it does matter whose [Reply] button you push.

     

    -Noel

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 12, 2012 4:27 PM   in reply to Noel Carboni

    Yeah, I know the practical side of it, how it behaves, but I have no clue about the rocket science behind it . That's more up your alley. Luckily, the functional concepts are beautifully simple and logical. Anyway, I'm flattered.

     

    I still don't understand why color management discussions are so inflammable. They always tend to get the blood pressure up - yes, mine too, I'm no better. As witnessed above. I promise to behave

     
    |
    Mark as:
1 2 Previous Next
Actions

More Like This

  • Retrieving data ...

Bookmarked By (0)

Answers + Points = Status

  • 10 points awarded for Correct Answers
  • 5 points awarded for Helpful Answers
  • 10,000+ points
  • 1,001-10,000 points
  • 501-1,000 points
  • 5-500 points