Ruj Posted January 9, 2023 Posted January 9, 2023 Hello. I have a number of damaged/burned military records that I would like to clean up for my genealogical research. I am a newbie to Affinity/photo editing. Does anyone have any ideas on how to do this or can you point to a tutorial? It's a challenge when the noise is the same colour as the surrounding text. I've already watched the videos on removing the grey bar down a book scan and the remove dust and scratches from the fish, but they don't really apply in my case. I, and probably many other genealogical researchers, would very much appreciate any help. I have dropped an example of a record below that I am trying to clean up and read the writing. Thank you! Quote
NotMyFault Posted January 9, 2023 Posted January 9, 2023 If you have any option to take better scans / photos from the original papers, and save them as tiff/PNG without jpg compression artefacts, in 16 bit color depth per channel, you may have a chance. The example you uploaded does not provide sufficient detail in the areas affected by noise to recover. we would need higher resolution uncompressed files (no jpeg) 16 bit color depth per channel color scans if possible I tried the usual candidates like FFT denies, frequency separation, Tone Map Persona to boost local contrast: nothing helps. You can easily boost contrast in areas without noise, but not in areas with noise. Quote Mac mini M1 A2348 | MBP M3 Windows 11 - AMD Ryzen 9 5900x - 32 GB RAM - Nvidia GTX 1080 LG34WK950U-W, calibrated to DCI-P3 with LG Calibration Studio / Spider 5 | Dell 27“ 4K iPad Air Gen 5 (2022) A2589 Special interest into procedural texture filter, edit alpha channel, RGB/16 and RGB/32 color formats, stacking, finding root causes for misbehaving files, finding creative solutions for unsolvable tasks, finding bugs in Apps. I use iPad screenshots and videos even in the Desktop section of the forum when I expect no relevant difference.
Ruj Posted January 10, 2023 Author Posted January 10, 2023 Thanks NotMyFault. Unfortunately, these are microfilm scans done by the Ministry of Defence. The documents were all damaged in bombing raids/fire during WWII. They are too fragile to handle and so cannot even be seen in person. I have no way to get the additional scans as you suggest. Quote
NotMyFault Posted January 10, 2023 Posted January 10, 2023 Sorry to hear. So the best you can do is to combine "human intelligence" and a bit image enhancing magic. You can try to boost contrast to separate "noise" and actual details by several methods: frequency separation haze removal highpass filter curves adjustments / levels adjustments FF denoise utilising RGB color channels tone map persona, local contrast Unfortunately nothing helps globally in this case globally, as the structure of noise is mainly identical to the structure of desired details. It is easy to recover areas without noise, but for areas with noise it doesn't help, or even degrades readability. The best you can try is to use these methods locally and try to improve local contrast, and use pencil tool to trace what your human eye detects as most probable curves. Then adjust settings for other areas, and step-by-step recover what is possible. Your eye is far more capable than any classic tool or AI in recognising patterns / structures. Sometimes you can detect structure by adjusting sliders fast. Quote Mac mini M1 A2348 | MBP M3 Windows 11 - AMD Ryzen 9 5900x - 32 GB RAM - Nvidia GTX 1080 LG34WK950U-W, calibrated to DCI-P3 with LG Calibration Studio / Spider 5 | Dell 27“ 4K iPad Air Gen 5 (2022) A2589 Special interest into procedural texture filter, edit alpha channel, RGB/16 and RGB/32 color formats, stacking, finding root causes for misbehaving files, finding creative solutions for unsolvable tasks, finding bugs in Apps. I use iPad screenshots and videos even in the Desktop section of the forum when I expect no relevant difference.
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.