— Dane Creek

Monitor Brightness and Dark Prints (Part 2)

A few months back I wrote a blog entry on a common complaint from first time printers: “Why are my prints so dark?”

Since then several people have written to tell me they’ve dropped their monitor brightness all the way to zero, but are still having a hard time judging whether their prints will print with the brightness they envision. I’ve also had clients send me files they believe are print ready, but without a little guidance would really lack the “pop” that people expect high-quality prints to have.

So, here’s part two on dealing with bright monitors and dark prints. Instead of relying on your eye to let you know if you’ve got the print right, it turns out there’s a very handy tool built into both Lightroom and Photoshop for this exact purpose: the histogram.

Three Before Histograms

Here are the histograms from three actual client files I received for printing in the last seven days:

From each of these histograms I can immediately tell that the images are too dark to print. I don’t even have to see the actual images. Let’s discuss each histogram in order to see what the signs are.

Histogram 1


Another sign that the image is too dark is where the midtones fall. Looking at the histogram you can see they’re almost all down between the 1/4 and 1/2 mark of the histogram, near where the shadows live. Getting those up more towards the highlight end of the spectrum will really help.

Histogram 2


The second histogram is even more obvious than the first. There’s almost zero data anywhere in the last 1/4 of the histogram. This means that the image has no highlights anywhere: nothing glinting in the sun, nothing catching nice light from a lamp or flash. Even worse, nearly all the data is shoved down in the shadows (the first 1/4 of the histogram), with very little content in the midtones.

Histogram 3


I’ll admit that at first glance this histogram looks pretty good. Unlike #2, there’s hardly any data in the shadows. There’s a pretty big hump in the midtones, but it’s towards the highlight end again, not the shadow side. And then a reasonable hump that looks near the highlights.

The trick here is the word “near”. The big hump is close to, but not quite, in highlight range. So everything that you might think is nice and bright and adding all sorts of neat little “pop” to the photo actually just isn’t quite making it. A small adjustment here to shove things over just a teeny bit makes a huge difference in the overall effect.

Three After Histograms

So now that we have examples of before histograms, what does a “correct” histogram look like? Well, correct will definitely vary depending on the image content, but here’s the histograms for the three above images just before they went to print:

Whee! So much better! So what changed?

Histogram 1

First, notice how much of the data has moved over to the right side of the image. It’s a massive swing from what the photo was originally. It’s almost all highlight or high-midtones now, with only a smidge in the shadow. While this adjustment was a huge improvement to the original image, it does come with the risk of accentuating noise that was lurking in the darker areas of the image. In this case things turned out ok.

Lightroom adjustments: +1.25 exposure, 33 recovery, 7 fill light, +18 lights.
Image subject: Bride (in a white wedding dress) and groom (in a dark suit) shot from above against a light terazzo floor. If you have a bride in a white wedding dress, it better look white when printed!

Histogram 2

This one wasn’t nearly as drastic as the first, but you can see how some data has moved into the highlights part of the histogram, and more importantly the bump of data that was in the midtone area is more towards the high midtone area of the image.

Lightroom adjustments: +0.89 exposure
Image subject: Three teapots with nice specular highlights on a wood block against a very dark background.

Histogram 3

Talk about subtle changes! If you squint you’ll see it though. The major hump that was in the high midtone area is now in the highlights region. The difference on-screen and in-print is remarkable. The image went from a slightly muddy high-key image to a really crisp high-key image that jumps off the page.

Lightroom adjustments: +0.31 exposure
Image subject: High-key portrait shot on film and scanned for digital printing.

In all three of the above after histograms you may notice that there’s a tiny gap at the very right edge. That’s for printing purposes to ensure that every part of the page has ink on it. For more details on why, see Brooks Jensen’s excellent blog entry on the subject.

Summary

This is the second blog entry on print brightness that’s wound up being much longer than I expected, and I apologize 🙂 Hopefully it’s been useful. There’s a lot to digest, but I’d summarize with two key points:

  1. The histogram is a valuable tool to see if your image is potentially too dark to print.
  2. There’s no rule to come up with a “right” histogram. It’ll vary based on the photographer’s vision and the image content.

Finally, since there was nowhere else to put it, just looking at the histograms I’d say that only one was properly exposed in-camera to begin with. The other two were anywhere from 0.3 to 0.6 stops underexposed.

Questions? Comments? Write ’em below!