Every month the ARTstor Blog announces new available collections from an international community of museums, artists, artists’ estates, photographers, scholars, special collections, and photo archives. Many teams in ARTstor work behind the scenes to make this possible: User Services, Library Relations, Production, Communications, Metadata & Cataloging, Collection Development, Finance, Human Resources/Administration, Legal, Software Development, Database Administration/Systems, User Experience, Quality Assurance, and Implementation. This month we begin a new series in which staff members explain the many steps required to share these images with you.
Lily Galib, Production Associate, Image Quality Control, has written a three-part post on the ins-and-outs of light value and color adjustments. Read part 2 and part 3.
At ARTstor, we have a philosophy of maintaining the integrity of the original artworks we feature in the Digital Library and representing them as accurately as possible. Consequently, our focus in the Production department is on image correction rather than image manipulation. This means that preserving detail is a priority when making light value and color corrections, and we never do retouching on top of artwork. For example, if a slide of a painting has been stored in poor conditions and is dirty and color shifted, we will correct for the color shift in order to match the actual painting as closely as possible, but we won’t push our adjustments into the realm where Photoshop is creating false color or detail. We won’t remove dirt from the top of a painting because that would alter the artwork and create an inaccurate representation.
The following is a description of some techniques we use for making light value and color adjustments on images for the Digital Library. In part one we will give a brief explanation of histograms, as having a basic understanding is essential to understanding how light value adjustments work. In part two we will explain how to adjust light values using the Levels Adjustment Tool and using the Curves Adjustment Tool, and in part three we will discuss color correction. When we make adjustments to images at ARTstor, we generally use a combination of Levels and Curves.
A histogram is basically a graph that represents the tonal values within a digital image, from black on the left to white on the right. It’s a very useful tool because it shows the distribution of tones within an image and allows you to see contrast and loss of detail, or clipping.
The numbers 0-255 represent 256 levels of density, which is the full tonal range of a digital image. 0 represents black and 255 represents white, with 254 levels of gray in between. The height of the peaks in the histogram represents the number of pixels of each particular density within the image. Gaps indicate missing tones; there are no pixels of that specific density in the image.
Generally speaking, a “good” histogram will have a curve peaking in the middle and gently sloping towards 0 and 255, representing a full tonal range without loss of fine detail in the shadows and highlights. There are, of course, many exceptions to this and what constitutes a good histogram really depends on what the photographer wants to achieve with an image. For example, an image intentionally shot in a very dark environment will have most of the pixels on the darker end of the histogram and few on the lighter end.
When you make light value adjustments to an image, you are stretching and compressing different parts of the histogram. Stretching will result in increased contrast and compressing will result in a decrease in contrast. You cannot increase contrast in one tonal region without decreasing it in another. You can only move contrast around.