Colormunki frustrating problems...

Discussion in 'Digital Darkroom' started by michael_kowalski, Mar 8, 2010.

  1. Somehow my post didn't go through so quickly...
    I've spent roughly 4 hours trying to calibrate the monitor using "Colormunki Photo". I've calibrated monitors before, not an expert by any means. So first I noticed is the with the white point set to D65 I got a blue-purple haze that just did not look right. Using Native white point I get more natural albeit warm colors. Comparing a neutral grey on screen to one on paper (color chip from XRite) the grey looks very warm and no where near neutral.
    I noticed through playing with luminance, that the Colormunki basically ignores the contrast setting. I can set it at 0 or at 100, and after calibration it will stay at whatever I set it to. Playing with luminance, setting the brightness setting I had trouble hitting the "green point". At this point I can go back to contrast, and upping or lowering the contrast will affect the brightness. So my next puzzling question is, if there is no set control of the contrast how does this device calibrate anything let alone my monitor? There are literally thousands of possible working contrast-brightness settings I could set my monitor to and each time I will get different colors/results/etc. I am suprised there is no optimal setting for contrast for this calibrating device. I paid for this thing I don't want to spend hours guessing what my contrast/brightness should be.
    Next there is an absolute lack of RGB calibrating. Before and after calibrating, MY RGB settings are the same, 100/100/100. Nothing changes, no options in the software to calibrate the RBG settings, nothing. I would bet the RGB settings have some role in making the colors on my screen appear warm -- the yellows are oversaturated and the reds are as well (slightly less).
    I am just really disappointed because I could have spent these 4 hours calibrating my monitor by hand to get much better results, and save a lot of money too. I will call Xrite tommorow, but have a feeling they will tell me absolutely nothing in which case I will return this piece of equipment.
  2. Comparing a neutral grey on screen to one on paper (color chip from XRite) the grey looks very warm and no where near neutral.​
    Unless your lighting is D65, that tells you nothing about the quality of the monitor profile. Similarly, unless they're the same grey, you're just grasping at straws. D65 will look warm to most users, as it's still fairly common for consumer-level displays to ship configured for an extremely blue ("white") picture. If native looks better, try working with native.
    I noticed through playing with luminance, that the Colormunki basically ignores the contrast setting.​
    Correct. You can't adjust contrast on an TFT LCD; it's a function of the backlight brightness and the design of the panel. Your "contrast" OSD adjustment just munges the incoming signal to flub some facsimile of contrast control. It shouldn't be adjusted under normal circumstances.
    Playing with luminance, setting the brightness setting I had trouble hitting the "green point".​
    The green point is more trouble than it's worth; different situations yield different luminance requirements. While getting in the neighborhood of the green zone is a good starting point, you're likely going to end up using a different number in the end. One that gets a good screen-to-print match for you.
    Next there is an absolute lack of RGB calibrating. Before and after calibrating, MY RGB settings are the same, 100/100/100.​
    They should be. See contrast above. There is no adjustment of the RGB guns because there are no RGB guns—it's a lame facsimile in software that mutilates the incoming signal. You do not want to adjust the RGB settings under normal circumstances (hence the Colormunki correctly not asking you to adjust them).
  3. Hey Colin, thanks for the responses. There were a lot of things you touched on that I didn't know about.
    Just to note on the first response, when I did calibrate with the D65 setting I actually got a very purple cast/hue to the display, as opposed to it being warm. Using the "(display's)Native" white balance gave a much nicer overall image but the temperature is warm.
    To be honest, I got the Colormunki to calibrate a monitor at work for pre-press as well as on the side, printing at home. I understand that the color calibration is relative on a small scale, calibrating the monitor at home, and printing at home, but as is this monitor does not display colors that translate into pre-press safe colors. It is very important that the monitor displays accurate unbiased colors, and it did, until unfortunately I had to reinstall Windows 7 and lost the color profile.
    In any case, if I compare a Munsell neutral 5 grey to a Munsell neutral 5 grey in Photoshop (by RGB values & fill tool), they should look the same, shouldn't they? Why would a neutral grey on my screen not look neutral? Same thing if I take a white (neutral color) piece of paper and place it over my display with a Photoshop file with white fill. The relationship between the white and white, there should not ideally be any color temperature difference. The paper looks white, the Photoshop white looks warm white.
    Opening up some of my earlier pre-press files on this calibration, the yellows are completely oversaturated, the reds are slightly saturated, and the reds have hardly any detail in them. If I were to send these files again to print, I'd never send them looking like that. I would have to heavily edit that very same file, with the same embedded profile, and my end print from the production press would look absolutely horrid. Alternatively, sending the files unedited as they are now, looking terrible on my screen, they would print fine.
    I don't mean to be a douche, but something is just not right about this calibration.
  4. Actually I just had a thought. I was thinking back to when I first calibrated the monitor using a XRite EYE"X" PRO (which I still have). With that product, while I did have a lot of trouble and spent many hours calibrating, I think the issue might have been with setting the white point. It took many hours to find the exact white point, but at least that software did give the ability to set a custom white point (as opposed to D65, D60, D55, D50). I thought that perhaps the EYE PRO was faulty (due to such time involvement) and that is why I purchased the Colormunki, hoping it would be easier and faster.
    I suppose I should return the Colormunki and bring out the EYE PRO again. One more question, is there anyway to find the exact white point for my monitor, or is it really trial and error?
  5. If its the same software that came with the EyeOne Display2, Eye-One Match 3 has a checkbox for Advanced versus Easy calibration. I haven't tried Easy mode, so don't know if that's what the problem is. My monitor's settings had to move pretty far to register a balance, RGB 37, 35, 47. The contrast point also definitely is measured in calibration. The resulting profile was pretty much a single straight, overlapping curve for all 3 in the results summary graph after calibration.
  6. Welcome to the club. I gave up on the issue and "calibrated" my monitor by hand in its settings using prints from the web lab I use. That works pretty well. I uses a Spyder before, and the result was a green cast, whatever option I choose in the Spyder software. I lived with that for a while assuming that I adopt to that look, but I did not. Moreover, the prints were still off. My assumption now is that my graphics card does not cooperate with that tool.
  7. White point settings (which depending on if they are based on a Standard Illuminant or a Kelvin value may be many differing colors) should be based, like target luminance on the conditions in which you view a print next to the display. And the only way you’d actually get the display to produce D65 (let alone 6500K) would be to heat it so high, it would be a glowing molten pool on your desktop and set your office on fire. IOW, the values are meaningless, the correct settings are those that produce a match. What are you using next to the display to view the prints? What’s the illuminant? High bright? Are you viewing the images that you wish to match with a soft proof from a good profile in a good ICC aware app like Photoshop with the proper soft proof settings on (simulate paper white and ink black)? Is this in full screen mode so that the paper white simulation is not being distorted by the white of the UI elements that do not under go the simulation? I know, a lot of questions. Point is, in order to get good screen to print matching, you have to get all your ducks in order. That includes to a high degree, proper print viewing conditions.
    RGB controls buy you nothing on anything but a high LUT display. Otherwise its just tweaking the 8-bit curves in the display LUT which introduces banding. Start with a native WP and TRC Gamma and adjust if necessary from there (you might need to alter the WP to produce a match). Again, all the setting values are moot unless they produce a visual match.
  8. Thanks for the responses and suggestoins.
    Andrew, at the moment I am not even looking at prints. I am just trying to calibrate the monitor to display good color and neutral gray. Unfortunately the office I work in does have both artificial light and daylight through clear textured glass. I know the environment does affect the viewing of the display, but I do not have an option to control the light in the office. That being said, I had calibrated this same monitor before. It displayed perfect colors to my eye throughout both day and night. Neutral grays were neutral, colors appeared as they should.
    I took out the EyeOne Pro and have been playing around with it for a while now. Overall, even with the same settings as the Colomunki, the EyeOne Pro gives better, closer color. The color is not there yet however. I noticed with the EyeOne Pro, keeping all the settings the same, the monitor's color change every time I run it. At one point the colors were VERY close, but I ran another calibration at 100k higher. It looked worse so I reset the monitor and went down again, but the colors are way different. I can run the calibration 20 times with the same settings, and each time I will get different colors (ranging from slight to moderate). How is this possible? Are these devices really that terrible? I thought the whole point of them was to create a set profile where red will be red, and stay that exact same red regardless of if I calibrate it once, twice, or 50 times.
    One other thing I did notice with the EyeOne Pro, is that if I am setting a custom white point, if I place the EyeOne on the white tile measuring tool, calibrate, and try to measure temperature on the white tile, I get random measurements. Measurements range from 4k to 29000k. Is this normal? Shouldn't the measurement at least be the same every time?
    I really don't know. Right now I'm starting to think these XRite products are absolutely worthless. I will be returning the Colormunki without a doubt, and will contine trying with the EyeOne Pro.
  9. Andrew, at the moment I am not even looking at prints. I am just trying to calibrate the monitor to display good color and neutral gray.​
    Your eye will adapt to gray (and white). Time to move on.
    If you can’t control the light, you can’t control print viewing and the calibration target values can’t be defined and you can’t decide what’s correct (the whole point is to match to an output)
    The ColorMunki is a pretty decent device for the money but if you had an EyeOne Pro Spectrophotometer, why did you even get the Munki?
  10. Andrew,
    I got the Colormunki because I had so much trouble calbirating with the EyeOne that I figured it was faulty. How can the device not be getting consitent readings with the same settings with a time difference of 5 minutes between the calibration? Ultimately I did get it calibrated with the EyeOne Pro roughly 2 years ago, but it took me about 2 days to get it there.
    I know that my eye can get "used" to the color distortion, but that just does not work for me. If I am making a design to send to press my colors will be way off. If I have a green tint in my greys, and I adjust by making them less green by adding warmth to them to the point that the grey is neutral, and send that file off to print, my print is going to look a lot more red or warm than it should be. If my reds are lacking detail, and I darken them and increase contrast, the print will not only be warm, but it will have serious problems with red colors.
    My monitor was perfectly calibrated before. All I am trying to do is get it back to that point.
  11. I got the Colormunki because I had so much trouble calbirating with the EyeOne that I figured it was faulty...​
    There is a diagnostic software product that check this. I suspect there’s nothing wrong with either device.
  12. I did just run the diagnostic software and everything did pass so the device is working properly I would assume.
    I found in the EyeOne Pro software a tool called the "Monitor Validator". I ran it with the calibration resulting in an average Delta Error of 6.88. I reset the monitor to factory settings and got an average DE of 9.26. After I recalibrated the monitor again and got an average DE of 7.10.
    Looking over the DEs for the colors (calibrated) the highest DE for green is 14.43, and the highest DE for red is 18.02. The black has a DE of 0.72 and white 0.04 so without those colors the average DE would be a lot higher. I made an appointment with an XRite technician for tommorow morning over the phone so I will give it a shot.
  13. The validator doesn’t tell you the calibration is correct, its a trending app. It tells you how much the device drifted or changed since the last calibration and profile. You should be doing this at least once a month. Those values are huge. That means your display is very unstable (or you are doing this every year or three). If you keep changing the target calibration and running the validator, you’ll get large deltaE reports too (ie, calibrate at D50, then D65). Again, the idea is to stick to one target calibration, calibrate every month or so, then view the trending. If high, you need to be doing this more often. But it doesn’t tell you anything you’ve set from the get-go is correct because again, it can’t. Correct is a good screen to print match.
  14. Just to update on this, it was apparently a nvidia profile setting issue where for one reason or other it blasted saturation in colors, specifically in the reds. The average DE went down from 7 to 1.11, the red with a DE of 18 is now down to 2.XX
    The colors look fine, thanks for the help. And just to note, the lady I spoke with at XRite was much less than helpful. She insisted that I was the problem and not the hardware when it was clearly a hardware related issue. Anyway, I am glad that this is all over and my monitor is finally displaying accurate color!

Share This Page