Basic Color Management For Photographers - Part I

PART I - What Are We Really Asking?

Because I enjoy making prints, one of the topics that I am frequently asked about is color management. At it’s most basic level, what photographers want to know (and one of the questions that color management seeks to answer) is a query that usually goes something along the lines of:

“How do I get my prints to look the way they do on my computer monitor?”

And it frequently is asked with an expression that looks something like this:

Image by Robin Higgins

Image by Robin Higgins

And for good reason……color management can be very frustrating when things go wrong and the answers to that seemingly simple ‘how do I get my prints to look like my monitor?’ question can get complicated.

There are many places one can go to try to get answers and, hopefully, a better understanding of the process, but I have often found that the presentation of color management information is either oversimplified (just do this because it works) or over-complicated (here is a treatise on color theory). The problem is that when things are oversimplified the photographer is left doing things without really knowing why. Then, when a problem occurs (as it invariably does), they have no idea how to troubleshoot. On the other hand, when the information is presented in an over-complicated way…..

Image by Gerd Altmann

Image by Gerd Altmann

This short four part series, then, is going to be my attempt to teach the basics…..just enough so that you can gain an understanding of what’s going on behind the scenes without having your eyes glaze over…..and leave you with the ability to set up your own basic color managed workflow.

In today’s installment I would like to discuss the ‘Basic Science’ of color management. Wait…..don’t roll your eyes!!!

I see it’s too late:

Image By Robin Higgins

Image By Robin Higgins

It really isn’t difficult, and a little understanding of the terminology will be useful down the line. Sometimes you really just have to get a little more explicit about things:

cartoon 1.jpg

So let’s start by asking what one might think would be a very basic question.

What color is this apple?

 
apple-maroon copy.jpg
 

“Red”, you say?

Well what about this one?

 
apple-orange copy.jpg
 

And this one?

 
Apple-red copy.jpg
 

Or this one?

 
applle-yellow copy.jpg
 

Well, you get what I mean. It’s hard to define what one means when they say red. There is yellowish red, orangish red, bluish red, maroonish red, deep red, light red etc. So, when you tell me an apple is red, I might conjure up a picture in my mind of what you mean, but that picture might be very different from what you are seeing.

What is red?

What is red?

Wouldn’t it be nice if there were some way to communicate exactly what color you mean so that the red apple I am thinking of matches the red apple you are looking at. In fact, in the best of all worlds, I could convey an apple color that is exactly and precisely a specific shade of red so that everyone who heard my description ‘saw’ the same ‘redness’? That might be hard to do in our heads, but what about on our computer screens? Surely we should be able to do that!

Well, that is what color management aims to do! Is it perfect……no, but it isn’t bad and it’s the best we have to offer.

Soooo..….now that we understand why we need color management and what, at its most basic level, it aims to achieve, let’s start to examine some of the things we need to know about in order to reach that goal.

To start with, we need some definitions. Sure, definitions can be boring, but we really need them to gain a basic understanding and to get beyond ‘just do this because it works’. So here we go…..


“Color Model” & “Color Space”

We talked about ‘red’ meaning different things to different people. Well, perhaps if we could put together a reference table of sorts that might help. Then I could point to a shade of red and show you exactly what I mean. That might be as accurate as we could possibly get since we can never really calibrate the ‘wiring’ between our eyes, brain, and consciousness. The equivalent of this would be to somehow quantitate each shade of red (and every other color). This is exactly what color scientists attempted to do by devising color models and color spaces.

A “color model” is an abstract mathematical way of defining a color. In the RGB color model for instance (which is the model that one uses for computer monitors and inkjet printing), every color is represented by three numbers, one number for the red component, one for the green component, and one for the blue component. In the CMYK color model (used in most commercial printing) each color is represented by four numbers, one for the cyan component, one for the magenta component, one for the yellow component, and one for the black component.

Sounds pretty simple right? The problem is that while each color is now ‘defined’ by a set of numbers, it is still pretty abstract in nature because one could ask what those numbers actually represent? The answer is that they don’t represent anything real yet……but if we put them into a defined ‘container’ with boundaries then we can start to assign numbers and divisions to all the axes that are within that defined space. Put another way, 2 bits of red don’t mean anything until we have a contained linear red scale and can see what two bits of red look like, counting from the boundry that is 0 red and extending towards the boundry that is as red as the container holds. That is what a “color space” does. It provides the boundaries into which we fit our color model. Now those three (in the case of RGB) or four (in the case of CMYK) numbers actually mean something and define a color that we can look at.

So lets see if we can represent what I just said graphically. Here we go…..

Chromaticity_Diagram.jpg

In these graphs, the black triangles define the limits or serve as the ‘container’ of the color space. Any colors inside the triangles are contained in the color space and defined by points along the axes. All three of these graphs pertain to different color spaces that utilize the RGB color model, but they each have different sized containers to hold the colors in. As you can see, ProPhoto RGB is the largest color space followed by Adobe RGB and then sRGB. In a very real sense the ProPhoto RGB color space contains, or can represent, more colors than can the sRGB color space.


Now that we understand these definitions, we can point to a color and assign it a set of numbers that define it. Theoretically, I should be able to pull up the color space graphs on my screen and look at the exact color you are referring to and know exactly what color you mean. However…….you know how when you walk into a TV store and all the TVs are tuned to the same channel but they all look a little different??? Yeah, we have a problem! While there may be a standard somewhere that shows what RGB=0, 12, 73 looks like, the problem is that we have to look at that depiction on something and, just like the TVs, the colors on my screen don’t necessarily look like the colors on yours. So I might be able to point to the very exact spot on the graph you are referring to, but it might not look the same on my screen as yours. It therefore appears that we might have solved the color problem from a quantitative standpoint but not from a functional, day to day use, standpoint.

And now you know why monitor calibration and profiling is important. It is a very valiant attempt to solve that problem, and I will discuss that in the next post.

Comments or questions? Just click on comments link below and I will do my best to answer them.