Using Histograms – ETTR

December 27, 2010 | 2 Comments |

Review Histograms After Exposure

One big advantage of shooting digitally is the ability to view a histogram in the LCD screen on the back of your camera body. A histogram is a graph of the relative distribution of the data in your image from shadows on the left to highlights on the right. You can use a histogram to evaluate not only the tonal distribution but also the quality of your exposures. By viewing the histogram immediately after exposure, you can determine if you need to make additional exposures at alternate settings to get better exposures. Simply program your camera to display a histogram immediately after exposure. You’ll find this immediate feedback will result in much higher success rates.

Don’t Clip

When evaluating a histogram, the primary concern is clipping or loss of data due to underexposure or overexposure. When a histogram ‘hits the wall’ to the left, the image is underexposed. If the histogram ‘hits the wall’ to the right the image is overexposed. This indicates that you should change exposure settings to get a more balanced exposure.

A Camera’s LCD Displays the Histogram of a JPEG

The histogram displayed on the camera represents the information of an image in a converted JPEG state, even if you are shooting in Raw. Because Raw is so flexible, you won’t know what its histogram will look like until it has been processed. Settings on your digital camera will only be applied to JPEG files it creates but they will also influence the histogram preview. Many digital cameras will allow you to set JPEG contrast to a low setting, which reduces the likelihood of clipping and provides a better Raw preview.

Raw Files Have More Data in the Highlights

If you shoot Raw files, you’ll not only have more data  in your files (16-bit instead of 8-bit), you’ll also have the option of processing files yourself with a RAW converter rather than having the camera process it for you. The Raw files you shoot will have more data in the highlights than is indicated by the histogram on a camera’s LCD. Typically there are many more bits of data in the highlights than the shadows. Assuming your photo has a 6-stop dynamic range (6 f-stops between the darkest and lightest parts of the photo) the progressive increase in bits of data looks like this from dark to light: 64, 128, 256, 512, 1024, 2048 (see above diagram). Because of this, data in shadows tends to be noisier and of lower quality than data in highlights. So, when shooting in RAW format, favor ‘overexposure’ rather than underexposure; optimal Raw exposure looks overexposed but without clipping. Expose to the right (ETTR). Then, later during processing, darken it. You’ll get better shadow and highlight detail this way. You’ll be amazed at how much highlight detail you can recover.

How to Quickly Achieve Optimal Exposure

In short …
1    Review histograms after exposure
2    Set your camera’s JPEG preview to low contrast
3    Expose to the right (ETTR) but don’t clip
4    Darken your Raw files during processing to create a pleasing appearance

1 clipped low

2 poor

3 good

4 better

5 clipped high

6 too contrasty … bracket exposure

Find more online digital photography resources here.

Learn more in my digital photography workshops.

  • Alessandro Rosa

    If you have a dark subject on a dark background and use an incident light meter reading to obtain the correct exposure which then produces a histogram like number 2, would you still recommend compensating and setting an exposure that gives you a histogram like number 4 and adjust using the RAW converter to bring the rendered image back to the correct exposure or does correct exposure based on the light falling on the scene override the results of the histogram? Do you run the risk of introducing a color shift if you do expose to the right for a dark subject/scene?

  • Yes. Even for dark subjects, make lighter exposures as they contain more data / gradation and less noise. Any color shifts that might occur from this are usually minor and much easier to correct than noise, posterization or clipping.


Get the RSS Feed  

Subscribe by Email