Hello everyone. Last year I fell into the LED banding trap, taking thousands of photos of a ballet performance, and not realizing about the problem until I loaded the images in my computer… a complete disaster! I was using for the first time the R6 Mark II, with electronic shutter (for quietness in the theater) on high shutter speeds (to freeze the action). That combination worked just fine in other performances during the same week, but on different stages with regular tungsten lights, that’s why I was completely unaware of the potential problem. Since then, I have been searching for a solution to recover those images. Could AI-enhanced software be able to: a) analyze each image at the pixel level and recognize the pattern of common variations of intensity and color, b) apply the reverse pattern on the image… I am not a software expert, just a photographer, but it sounds possible to me… the only difficulty may be the enormous processing power necessary to analyze and process individually each of thousands of images (because the banding is never identical from image to image). I am including a jpeg sample; I will be happy to provide the .raw files of the performances if someone in this brain trust community would be willing to take up the challenge, thank you. (I have since learned how to avoid or minimize the banding problem, but there may be thousands of photographers worldwide who will also be very grateful and happy to be able to recover their photo-disasters)
(Please note that the attached image has at least two patterns of banding, because background and stage lights were on different frequencies)
This is the Dropbox link to the .raw file: Dropbox
Repost it as a “Play Raw”, along with the raw image. This will give people a chance to see what they can do with it.
1 Like