When is colour matching critical?

[erie patsellis]

Exactly what Peter said, I rely on the RGB values of the Q13 patches and while it may not always look exact on screen, capturing with a properly profiled camera and printing to a properly profiled printer is usually (96%) a visual match to the original. From there is becomes a tweaking game, some clients want more saturation, or contrast, but it's usually dead on first pass.

[Frank Petronio]

Not that many people print on real printing presses anymore, but their color gambit is severely limited (and skewed) compared to the inkjet printers most of the hobby togs are used to.

Which should be quite a surprise to some of you once you get that vanity book project and learn that the printer really, truly can't print the sky blue.

[bob carnie]

I agree with Peter here.

The proof is always in the print for me, with that said I am completely happy with the output of our first tests, usually a small tweak and go to final, sometimes more, but I rely on the numbers rather than looking at the screen. My info pallet is always set with RGB on left and LAB on right and I use the LAB numbers when thinking about adjusting an image.
I try to make a habit of not staring at the screen but always looking at my second monitor which is set up to nuetral grey background and give my eyes a blast of grey when adjusting files.
The A and B channels are much like the dichroic wheels of an enlarger and if you have done a lot of colour enlarger printing , you will appreciate working in LAB.
As well I like the L channel for readings as it is in a 0-100 or ten zone way of reading density.
For each paper I get to know the significant highlight and significant shadow readings for the L channel and set these for printing. In between is the fun part.

Those who say they get a perfect match between print and screen should go on the road and sell their services Colour Management Geek Squad . com as there are millions of others out there not able to do this. There must be a good buck in that.
My 2 cents is that if you do not have to go back to your file and make major moves and simple adjustments at output are all that required , you have a good workflow.

Color management will only get pretty close to a screen/print match. Computer displays and prints are two very different technologies with different characteristics. They can be made similar but not the same. If you're really concerned about how the print looks, you won't be able to avoid making test prints. If possible, make these judgments using the same light source as the print will be displayed in.

Professional post production people will tell you that "the numbers don't lie." What they mean is that they trust the numbers more than how a monitor looks. A good post production person can color correct on a black and white monitor. Heck, even I can do it. It's easier on a good and color-managed display, but not by all that much more.

Recently, my neighbor asked me to photograph his house and color it in Photoshop such that the house colors matched some color swatches that he gave me. This was a non-trivial task. I used my spectro photometer to read the colors of the swatches. I got readings like a lab A reading of 6.86. Well, photoshop only allows whole numbers with LAB channels. So the best I could do was to get close. This type of thing will pop up, for example, if you try to exactly match the colors of a painting with a computer produced print.

So, if someone says that they're monitor to print match is perfect, then don't believe them. How close you should get is a value judgment that only you can make. Like anything else, at the highest level you can pay a tremendous amount of money for a very small gain.

[IanMazursky]

Its especially important when the image is destined for press or even inkjet. Having your proofs match your screen and press output is the holy grail for pre press.
The standards set forth in SWOP, Gracol and SNAP will help with color and density (for press and press proofing) but they are not perfect. The standards are still very vague but are a good starting point.
If you calibrate your monitor to your printer/proofer to a standard like SWOP or GRACOL, it will keep things consistent and reduce your overall expenditure on ink, paper and most important time.
Although they are a smaller gamut then a profile not based on those standards.

The numbers can and do often lie! They only tell you what the current profile or the soft proof profile sees (assuming you have one from the printer).
There are so many variables when going to press like how the pressman slept last night or the temperature of the press room....

Inkjet on the other hand is much easier (sort of) but it still has a number of variables that can throw things off.
Room temp, clogged head, printer drift, media humidity.....all affect how a printer lays down ink. They can vary a lot during a day.
Linearization before profiling is a really good thing to do. You can re linearize a profile without re printing and reading the whole 1500 patch target.
It will also help to keep the profile stable even if the machine has a slight drift.

With a device like the ColorMunki or an i1 Spectro, you can get close and the numbers would be more truthful.
Monitor gamut and printer gamut are two different beasts, then you have reflective and transmissive color. Gloss, Satin, Semi Gloss, Matte...medias...AHHHH!!!
They all have a hand in throwing color accuracy out the window. With profiling you can get much closer to WYSIWYG.

One other thing to mention is profiling yourself into a corner. Some people swear that a 2000-3000++ patch target is the only way to go.
With one of those, you can create more problems for yourself. They leave no room for anything like a misfiring nozzle, slight drifts or out of gamut colors.
Instead of a nice curve, they give a jumpy wavy curve that can cause reversed or strange colors or ones that will take forever to dry.
They are printer/ink/media dependent and sometimes trial and error is the only way to go. Fun, fun, but a necessary evil.

Sometimes less is more and the last 5% is a killer!
In the end, a good scanner, monitor, printer profile will save you time and money.
Over a number of years that can add up to a lot more $$ in your pocket then in the waste bin.