While doing my digital darkroom work on photographs from Europe, I became dissatisfied with the results I was getting from a few HDR (high dynamic range) images, finding them too flat and without sufficient contrast and depth. Standard methods to increase the contrast on these images were not achieving the results I wanted, perhaps because I am reluctant to commit to contrast adjustments early in the process.  But I didn’t want to simply tolerate images that didn’t quite have sufficient contrast, lacked texture where appropriate, or flattened out shapes and shading.

Pondering this quandary, a technique that is common in digital astrophotography came to mind. The technique is often called “luminance layering” and was developed independently by Dr. Kunihiko Okano and Robert Dalby in 1996. Color digital astrophotographs are usually created by taking a series of grayscale exposures through separate color filters (commonly red, green, and blue), and then combining those series into a color composite. While this can produce good color information, it tends to produce images with poor contrast and a fair amount to noise, due to the filters limiting the transmission of light to the sensor. An unfiltered image produces better contrast and cleaner images, but would lack the color for color imaging.

Luminance layering adds a fourth set of images to the processing, namely an unfiltered black and white image with good contrast. Often the astroimager will use longer exposures for this unfiltered image in order to increase contrast and detail. This exposure records the detailed brightness levels in all colors, or the luminance levels of the object being imaged. These files are then processed so that the color data is taken from the combined color exposures while the brightness level for a given pixel of the image is taken from the luminance layer. In Photoshop, this is accomplished by putting the luminance image in a layer above the color composite and setting its blending mode to “luminosity.”  When done with care, this results in an image with greater detail, improved contrast, less noise, and good color. In other words, it often produces a more aesthetically pleasing image.

What if I were to adapt this technique to terrestrial photography?

Clearly, many of the details of the process will change. We begin with a color image produced by our digital camera.  And since the color filters are built into the camera’s sensor assembly, we don’t have a true luminance image like in the astrophotography process. But one can convert a color digital image to a black and white image.

But what good does it do when one starts with the same image? If you’re wondering that, it’s a reasonable question. Black and white photographs depend on contrast and differences of value (essentially luminance). In converting a color digital image to a black and white photograph, one pays specific attention to these matters without being distracted by color. Further, increasing contrast and other adjustments can introduce unwanted distortions of color, or bring out color noise. It is also easy to block up shadows or blow out highlights and remove color from these areas. During the conversion process, contrast can be increased without these color distortions and attention can be paid to getting the relationship between the values of various portions of the image into good order. This can then be applied as a luminance layer to the base color image.

Isn’t this about photography?  Well, where are the photographs?  Well, here…

This image of a tower in the old city of Heidelberg, Germany, just down a street from the Universitätsplatz, has been enhanced with the luminance layering technique I’ve outlined here. On the left below is the original image, after processing from the RAW camera file and some initial adjustments for exposure, saturation, and so forth.

The enhanced image has better shading, which highlights the roundness of the tower. It also enhances the texture of the tower, especially the stone work of the lower section, which is perhaps a bit flat. Perhaps the most striking difference is the accentuation and increased structure to the clouds, especially those above the tower. To the right is the black and white image that serves as the luminance layer, defining the brightness of each pixel applied to the color data from the left hand image.  Larger versions of these images are available here.

This technique isn’t appropriate for all images, of course, but it adds back depth and contrast to flat HDR images and helps to enhance textures and shape in many standard images as well. The technique can also bring needed clarity to images that look a bit “foggy” due to insufficient dynamic range or contrast. As with all digital darkroom adjustments, an eye to detail and a judicious hand, combined with trial and error, must be one’s guide.

In part two, I will go over the specifics of my technique in greater detail. Further examples are available here.