Help with calibration

I have just recently encountered some issues with print output. I currently submit my work to several different labs. I just received some prints that came out darker than expected.

I currently use a Monaco Optix XR colorimeter, and I have the Monaco Optix XR software (as in, not Pro). After several attempts (and hours) I’ve been able to get the luminance close to the suggested setting of 120 cd/m2 with the white point at 6500k and the gamma at 2.2. My software does not allow me to suggest or request these settings as ‘targets’.

The difficulty with getting my monitor to these settings is leading me to question if I a) need a new colorimeter (though from what I’ve read, I don’t believe so), b) more advanced software (namely, that lets me specify additional settings) and/or c) a new monitor.

I’m currently working on a Mac Mini (intel gma 950 card) with a Dell E228WFP Monitor.
Any suggestions are much appreciated. I’ve been looking at possibly getting the Coloreyes Pro software if that’s all I need to move forward with more accurate results.

The Optix is a quality device, and if it is working normally you should be able to get good results from it.

If the difficulty you are having reaching your aimpoint targets is because your software does not allow for entering in these values - then there are certainly other software packages that will give you that. ColorEyes Display Pro is one of those and works very well. If your difficulty is because your screen is too bright and that makes it hard to find ways to bring the luminance down, then you could get a new monitor or the ColorEyes software would allow you to use the video card’s color tables to bring the luminance down further.

Ultimately, if your prints are coming out too dark (and assuming you are using profiles correctly), it is usually the fault of the screen being too bright. Try out the “white paper test:” Open a blank image in Photoshop (with a white background) and hold up a sheet of your printing paper. If the white of the screen does not match the white of the paper, you will not have success getting the actual monitor image to match the print.

Thanks Patrick,

I’m currently using the ColorEyes Pro demo. It has been brought to my attention that I need to tame the light that I work in (my setup is next to a three-panel window). Beyond that, it appears that some of the gray levels are off on my monitor which I fear means it may need replacement.

How would I configure ColorEyes to use my video card’s luminance tables. The card is an intel gma 950 integrated card.

Most monitor profiling software has you turning down the brightness at the monitor, and they depend on you being successful at doing that, because the rest of the calibration procedure determines the color adjustments that are made in the graphics card.

ColorEyes Display will automatically make whatever changes it needs to in the graphics card - both for color and luminance - to bring the display to the target aims. It is important that you optimize your screen as much as possible in the Monitor Pre-calibration step, because everything that happens after that step is affecting the graphics card.

Too much yanking the color around in the graphics card can lead to banding of colors, but if you’re dealing with gray uniformity issues, this shouldn’t be a problem.

Are certain portions of your screen off-color for gray?

I am not sure how to determine if certain portions of my screen are off gray (all the calibration has been done in the same general area, FWIW).

The rub with this monitor is that in the pre-calibration step, adjusting the brightness and contrast (I’ve read that they were wrongly termed so) affects the color. So in an endeavor to get the color right, I’m often adjusting down the contrast and brightness just to get the colors to balance – and I’m not sure how low the brightness needs to be.

In the pre-calibration step, the main idea is to get the luminance as close to where it needs to be as possible. A secondary goal is adjustment of color. The purpose of the whole rest of the monitor calibration procedure is for the software to figure out what the monitor is doing and make a profile that reflects that. So basically, if you get it close at the pre-cal, the software will take you the rest of the way. You should not have to spend a lot of time fine-tuning the contrast and brightness controls to try to adjust the color in the pre-cal.

In fact there is a section in the Guided Tour section of ColorEyes Display Pro that recommends that for LCD’s you make no color adjustments in the monitor control buttons themselves - just let the software do the job of bringing you into 6500 or whathaveyou. You might read more on that.

Also, do look at the white paper test described. That gives you an idea of where to set your brightness.

Thanks for the help. I have been pre-calibrating to the point where things remain steady and as many parameters as possible are zeroed out (usually means that brightness, and two colors are at 0 with one at -1 or +1).

I’ve tried it at D50 and D65 with good validation results (as far as deviance – even the grays seem to be back under control).

I’ve gotten curtains to control the lighting and I have a solux lamp on the way to view prints under for verification. In the meantime, I’m looking to get some test prints run off next week so that I can be a bit more assured about what my clients are getting.

Just an update and a question.

I have Coloreyes Pro, I’ve gotten the proper lighting to view prints under (solux lamp) and I have controlled the ambient light in my computing environment (blackout curtains).

I am finding that my monitor calibrates best to D50. At D65, the gray goes all over the place (particularly the 2nd patch, but blues start to fall apart too) when trying to calibrate to maximum luminance.

If the lab that I use specifies that prints should be prepared at D65, 120 cd/m2 white point luminance, and 2.2 gamma – will I inherently have problems working from a monitor calibrated to D50, 102 cd/m2 white point luminance, with a 2.2 gamma?

The reason why most people decide to switch from D65 as an aimpoint to D50 would be because it matches better what you are aiming for, or because your ambient lighting is rather warm. We normally don’t change the whitepoint aim because a different white point gives more consistent grays.

If a gray step wedge is erratic and gives a noticeable color cast, either the calibration software is not doing a good job or the monitor’s consistency is all over the place. Could it be that different portions of your screen are giving you different color? Concerning the first point, ColorEyes Display has a feature that allows you to add more sample points in the colors it samples. (Click the Gray Balance link under Profile Evaluation and add more iteration points where you see the gray balance shifting color.)

This is assuming your Solux lamps are getting you something close to 5000 K. And I assume your monitor luminance is 102 because that is the maximum you can put out?

Normally, you would want to go with what your lab recommends, but it’s a tough call to make. Our eyes get so used to whatever colored white point we are seeing at the moment that just about anything can look “normal” after a while.

Ok. So judging by the first portion of your reply, I should be calibrating to D65 if the lab suggests D65, consistent grays or not, yes?

I can’t help but think the issue with the grays is one of the monitor being all over the place. I believe I purchased this monitor in March of '06 or '07; it may have run it’s course. I will look into adding additional gray points – I will need to look through the guide with respect to that.

The monitor luminance was set at 102 for several reasons. 102 was what was being put out with the profile settings set to “maximum”, the monitor brightness was turned down in an endeavor to address the issue of dark prints, and I was finding that I was getting more consistency in the gray areas when the brightness was lower (my inference was that the monitor performance was degrading at the highest brightness settings).

As far as the Solux lamp is concerned, I’ve only been using the lamps to view prints under. When I’m working in general I just have blackout curtains drawn and the working ambient is pretty dim. The Solux lamps were intended to provide a control on the light source I was viewing prints under when comparing to the monitor. I don’t think I can get my monitor dim enough to match the perceived brightness of the paper while still having a chance of reaching the luminance levels suggested by the lab.

And I have seen my eyes normalize at the different white points, so that was part of what informed my question. In my reading, it appeared that some people suggested calibrating at D50 (and even the native white point) because the monitor performed better under those conditions.

I will attempt to revert back to D65 and read up on the gray points. Please advise if I need to be considering anything else. My best guess at this point is that either I’m not doing something correct in the software, or my monitor is officially the weakest link.

Calibrating to D65 produced a luminance of 123 and a minimum luminance of 23; the 241 gray patch is the frenetic one and has been consistently regardless of what section of the monitor the sensor has been positioned. Running a validation test against a color chart had colors all over the place too. 85% white in Lightroom started to show faint signs of banding.

You should be able to judge if your monitor is the issue. Make a Photoshop document with a solid surface of 241, 241, 241 (if that is the color that is showing the most variation.) Expand it to cover your entire screen. You should visually see a lot of variation in color given your validation results. If you do, then this is entirely due to your monitor.

You can also put a neutral gradient on your screen and see where the colors don’t look gray. The CED Pro software has such a gradient in the Gray Balance procedure, but you can also download a gradient from other sources. (Try the White_Balance.tif from If you see a lot of variation, follow the procedures in the Gray Balance procedure to add more iteration points.

Concerning “calibrating to the native white point” - I know what most people mean when they are recommending this, but it can be confusing. Every LCD display has a “raw” color that its backlighting tubes put out. If the native white point of your LCD is 6900 K and you calibrate to that, then the white of your monitor will obviously be 6900, not 6500. Instead, what people mean when they recommend you calibrate to the native white point, is:

  • Leave the buttons alone on the front of your display that change the color of white in the monitor, keep it at the factory defaults so that the monitor is producing the most white it can at its native white, then
  • Set your calibration software to your desired white point (like 6500) and allow the software to bring the white point in line with your aim using your computer’s graphics card (which is where the profile information is used.)

This procedure can have some benefit for older or poorer-quality monitors since it does not rely on the liquid crystals in the display to accurately adjust color.

I think I wish I had a second set of eyes to look at it, but I’m pretty certain I’m seeing variations on the monitor using the method you recommended in Photoshop (the delay in the reply was due to a computer repair being in progress on the system that utilizes this monitor). It seems like a faint blue gradient is present on the upper left corner of the monitor that is more noticeable than say, in the dead center.

I will try to add more iteration points using the Gray Balance tool in CEP. Should I be getting “more accurate” readings as a result of this procedure?

I’m not sure how much stock I should be putting into the validation test results, but these were the results after adding several iteration points on the gray scale and profiling once more. The validation results were more favorable (avg dE= 0.89, maximum dE = 2.xx) before adding the gray points. I added points at 211, 224, 237, and 247.

Is it unquestionable that this monitor needs replacing?

If your validation results continue to be erratic like this, then that points to either your monitor being unstable or your measurement device is. The Optix is generally a very stable device. You could test it by calibrating someone else’s display and see if you get the same results there. Otherwise, you’d be looking for a new display.

Will it suffice to try calibrating a laptop display? That’s the only other display I have. It’s an older laptop, but it should work well enough, correct?

Big difference between replacing a colorimeter and a monitor.

Yes. The idea is to see if the near-white of your laptop has the same erratic behavior as your other display. A laptop will do fine.

Here are the results of two consecutive readings on a Powerbook G4 (purchased in 2005).

I think the monitor needs to go.