Color management to most seems like a combination of black magic and voodoo.  We read about it, try to figure it out, but just never get it to work as well as those that seem to have “mastered” it.  The goal of color management is to accurately predict from your monitor what your print is going to look like, but accomplishing that seems to be akin to finding the holy grail.

Like most, my success has been somewhat limited.  I remember a long time ago experimenting in Photoshop trying to understand color management, and how to actually make it work.   I had a good color management software package and messed around making profiles that gave me pleasant results on my monitor, but despite good monitor profiles, the prints never quite looked like the monitor. My method to overcome this was pretty straight forward ... just make a print and then adjust the file accordingly.  Sometimes the print was close enough, sometimes it needed a tweak, but for the most part I gave up on getting it to “match”.

This was especially true with density.  Seemed like every print was dark. I got in the habit of creating an adjustment layer that would lighten the print before output.  I assumed this was the norm, and those not fighting this problem were just lucky, or had a better monitor than I had.

A year (or two) ago I noticed the luminance option in Gretag MacBeth’s Color Match software.  The recommended setting was 120, but I wasn’t really sure what that meant.  I started experimenting, and was quite surprised to see that achieving this required me to dim my 30” Apple Cinema  Display substantially ... to less than 50% brightness.  When I first did it, everything seemed a little dark and dingy, but I went ahead and made the profile.  I started working on some images and at first had a hard time, but gradually things seemed “normal” again, as my eyes adjusted to the dimmer display.  The end result was my files were less dense because of the adjustment to the brightness of the screen, and for the first time the density of my prints looked pretty good when I printed them.

I’m not sure if I “found” this on my own, or if I heard it from those that teach it and just didn’t understand it before. This is pretty commonly taught by many sources.  What most of them fail to mention is how important it is when trying to get the correct density from your printer - of course it is so logical you wonder why they would have to mention. Another key point is that the recommended setting is just that ... the recommended setting.  The actual number you use is based on a few other things, perhaps most important being how bright the light is where you view your finished prints.

So now that I understood the density part, the color thing wasn’t nearly as big of an issue.  Most photographs don’t require a real tight tolerance to look good.  If the screen version is a little blue, the print still looks great.  For the most part I felt my screen was a decent match to the printed output, so I didn’t really worry about it too much.  However, now I think I understand why I was happy with the results and why others may struggle with the color portion.

At the recent Photoshop World, there was a session called “Soft Proofing for the Perfect Print”.  I’ve attended classes like this before, but never really walked away with what I felt was an answer to why my monitor couldn’t more closely match my prints.  The class was taught by Randy Hufford, whose company specializes in high end printing, especially art reproduction.  The tolerance for him is pretty tight ... the print has to look like the original as perfectly as possible, and rather than doing tons of test prints, his approach is to have an extremely tight color managed workflow.

I assumed the class would be mostly about the use of Photoshop and a proofing setup, but instead it was mostly about the equipment and steps it took to get your setup appropriate.  As I sat there I realized what was wrong ...

I’VE BEEN LOOKING AT THIS THING BACKWARDS!

You might be as well.  I thought the secret was to create a good monitor profile and a good printer profile, and everything else would fall in line.  I see this advice all the time (this question of matching a monitor to output is asked all the time on various forums).  I have been guilty of offering this advice.  For the most part it’s pretty accurate.

If you think about the process backwards, the problem and Randy’s ideas make sense.  What part of the printing process is constant and what varies?  If you want to match two devices, you have to start with the device that’s constant, and understand how to adjust the device that varies to match it.  You have to match your monitor to your printer, not your printer to your monitor.  Current Epson printers have a Delta E tolerance of less than 1 (that means you probably can’t see the difference).  If the printer is that accurate, then what we have to do is adjust the rest of the process so what we see on the screen matches what the printer produces.

I’ve actually been doing this to a point without realizing it for sometime.  When I create a monitor profile, I normally create 3 or 4 versions, and vary the white point and the gamma slightly to achieve what I feel is a neutral image with pleasing contrast on the screen.  Most of the time I end up with a profile that is around 6000k and 2.0 gamma.  So my profile creates a slightly warmer and less contrasty image than the recommended setting of 6500k and 2.2 gamma.   And this is exactly the process it takes to get a closer match of your monitor to the print ... you literally have to create a profile that matches the print. 

This is what Randy taught in his class, and the logic seems to make sense.

Now I’m so excited about this I’m writing this article while I’m still at Photoshop World.  But I’m quite confident this is the piece of the puzzle that I’ve been missing all of these years.

So here are the steps I plan on implementing (most of these from Randy’s class ... many thanks to him.)

First I’m going to work on making sure i have a little more control over the ambient light of my screen.  I’m having plantation shutters installed which will help tremendously in the daytime.

Second, I have two places I work on prints and I’m going to tighten up the lighting there (both use color corrected lighting but I need to verify the intensity of those lights.

Third, I’m going to make a high quality print from a known test file, using a good printer profile.  Most likely I’ll be using Bill Atkinsons test print ... I love the images on it and I feel it really pushes the envelope on highly saturated natural colors.

Now I have my starting point!.  The next step is pretty straight forward.  I’m going to load the file up in Photoshop, set up the proofing mode to use the printers profile, and judge how close it matches.   Then I will remake my monitor profile, applying any adjustment in color and contrast.  I’ll keep remaking that profile until my screen is as close to matching my print as I can get it.

The end result (I hope) is when I soft proof an image, I can be pretty sure the printer will produce something very close to what I see.

One idea that really intrigues me is using different monitor profiles to help match the contrast of various papers.  Natural fiber papers with matte ink tend to have lower contrast than papers using glossy ink.  It may be possible to create multiple screen profiles, varying the gamma to soften the contrast to more closely match these paper types.

WHY IS THIS NECESSARY?

I believe most that teach color management think this isn’t the way things should work.  If everything is calibrated and profiled, then it should work.  If it doesn’t then I didn’t calibrate and profile things correctly.

My personal theory is pretty simple.  There is one element of this process that exists outside the color management system.  As most systems are set up there is not a standard way to control or measure this one parameter ... the viewing conditions of the print.  What color is that light?  How bright is it?  How much ambient light is involved?  I’m guessing the better the quality of the viewing station, the closer a standard monitor profile will work.  There may be viewings stations that can even be calibrated .... I don’t know.  If so I’m sure they are out of my budget and far beyond my needs.

So what I’m doing with all of this is calibrating the display to match the viewing station that I have set up.  The risk is if my viewing station isn’t at least somewhat decent, matching a print to a monitor wont’ be much use, because the print won’t look good under any other viewing conditions.  It’s important to have a good viewing station, with control over the color, intensity, and glare of the light source.  If this is done well, then a matching print will also look good under a variety of lighting conditions.  This isn’t new  ... viewing stations for color printing have existed longer than I’ve been printing photographs in 1975.  I’ve been using them for that long.

If you are having problems getting your monitor to match your prints, welcome to the club.  Maybe some of these thoughts well help you.