On top is how the image appeared on screen; below is how the print appeared.
As you can see, the images are close, but the printed version is darker in the skin tones, with more saturation, and a slight reddish cast.
Color management is designed for the express purpose of ensuring that what comes out of the printer matches what you see on screen. The way it is typically taught starts with a discussion of monitor calibration and moves to the use of printer profiles and soft-proofing. I don’t want to suggest that this isn’t useful information, but even when applied correctly, ink jet prints still don’t match what is displayed on the monitor. There are some technical reasons for this that may be cited: screen contrast is much greater than that of the print and the range of colors an ink jet print is capable of reproducing is smaller than the range a monitor can display. But as a practical matter, images which contain colors outside the range of the printer’s capabilities are the exception. The real problem lies with the printer driver software – this is the software that determines what size and mixture of cyan, magenta, yellow, and black ink droplets will be used to represent the various colors and tones of the image. The software that comes with the printer simply doesn’t do a good job of this. The most efficient solution is to buy third party software to drive the printer, but it is expensive, and I’ve never taught at a school that uses it.
Fortunately, if you are using Photoshop, there is a workaround for this problem. The best part is that it is free and it doesn’t require a profiled and calibrated monitor. (If you’re printing at school, who knows the last time the monitor was calibrated?) I came up with this technique to give students a better way of printing. Typically, when the print came out dark and off color, you’d add some adjustment layers to compensate, but it was really guesswork as to how much lighter to make it or how much to adjust the color. You went through a process of trial and error, until you got something that looked OK, or gave up. My solution was to eliminate the guesswork.
The first step is to compare the print to the image on screen. When you do this, you want to be sure that the print is illuminated by the ambient light of the room. A lot of monitors have hoods to block off the ambient light. You don’t want to hold the print so that it is under the hood. I’ve found that it helps to pull away from the monitor so that when holding the print it matches the general size of the image on screen.
Next, make a series of adjustment layers to match the screen image to the print. Typically, I’ll use a curves layer to darken the screen image, followed by a hue/saturation layer if necessary and then a color balance layer to fine-tune the color match. Your three adjustment layers might look like this in the layer stack:
It wouldn't be a bad idea to label each adjustment layer as a print matching layer, and perhaps include the model of the printer, just to prevent them from mistakenly becoming a permanent part of the file.
At this point, the image on screen should match the print as closely as possible. If I’m demonstrating this technique in class, usually someone asks if I haven’t made a mistake: after all, we’re trying to match the print to the screen and not the screen to the print. But this is what we have accomplished through the initial set of adjustment layers: we’ve found the difference between the two. Now we know exactly how far off from the monitor image the print is. To get the print to match the screen, make the opposite correction with all of the adjustment layers. So, if my color balance layer reads +6, 0, -1, I simply switch the signs to -6, 0, +1.
If the overall saturation was raised to +16 to match the print, I’ll invert that to -16. You can click on these screenshots to see a larger image.
The curves layer can be the most tricky. The easiest way to invert the curve is to use the input and output boxes, entering the adjusted values in the output box using the keyboard. Click on the point or points on the curve to select them, and one by one determine the difference between the input and output for each point. Below are the matching and inverted curves. Notice that the arc of the curve as it sags below the diagonal unchanged line is mirrored in the inverted curve.
And this shows the way to make the curve inversion for the upper right point. Of course, the other two points on the curve in this example must be selected and inverted next.
After each adjustment layer has been inverted, you will have an image on screen which will vary from what you intend the print to look like, but will account for the un-cooperativeness of the printer. This is what my adjusted file looked like on screen. As you can see it is lighter and less saturated than how I want the print to appear. Printing from this adjusted file will give me a print which matches the desired look of the image.
The principle behind the technique is simple: find the difference and compensate; however, it may seem like a lot of work the first couple of times. The trick is in accurately matching the screen image with the print when making the initial comparison. With some practice, you'll get the hang of it and improve your ability to discern the nuances of color and tone to boot. It certainly is better and quicker than guessing. One thing to note: the adjustments you are making apply to the particular monitor you are using and the exact model printer. If you are moving from workstation to workstation, or print to a variety of printers, you'll have to redo the process for each printer/monitor combination. If you are a teacher, feel free to try this with your class. If you share it, all I ask is a link back to this page.