Photographers hate him! One weird trick to extract more dynamic range from your RAW files
14 Mar 2017
Although most displays can only represent 8 bits per channel, modern image sensors can capture much more. For example, even my super old Nikon D5100 can produce 14-bit images in RAW mode. The 14-bit sensor data is then “developed” into an 8-bit image, either in the camera’s firmware or in Lightroom. Both algorithms will choose some range of brightness values to display and clip the other parts of the image.
But we often want to see the full range of brightness values. The solution is to use a tone mapping algorithm. It takes advantage of human perception to reduce the overall brightness range, while preserving local contrast. Although tone mapping was originally developed for HDR images, it’s applicable to any situation where we want to squeeze an image into a smaller bit depth.
For example, Apple Photos has a “brilliance” slider that appears to apply local tone mapping to the image. From the support article:
Brilliance: Applies region-specific adjustments to brighten dark areas, pull in highlights, and add contrast to reveal hidden detail and make your photo look richer and more vibrant. The adjustment is color neutral (no saturation is applied), but there may be a perceived change in color because brighter images with more contrast appear more vibrant.
I’m not 100 percent sure about this but I don’t think Lightroom has an analogous feature. So here’s a scam that achieves the same results:
Tricking Lightroom into tone-mapping a single image
First, we’ll create three copies of the photo. Right-click the image in the filmstrip and select Create Virtual Copy.
In Develop, set the exposure of each image so that the entire range of brightness values is represented:
Finally, select all the images and use Photo > Photo Merge > HDR (Ctrl-H) as usual. Here’s my result:
This process is really inefficient and janky because each image in the first step loses some image data, which we force the HDR tool to recover in the last step. Then the HDR tool runs its tone mapping algorithm (the part we actually wanted). But the results look good, so I’m not going to question it!