Thank you for excellent work on the script. I started to use it recently to convert Vuescan RGBI files. And the issue is that the script identifies dust and scratches on contrasty edges in the images, i.e. candle flames. After running dust removal macro the image does not look good.
Maybe there is a way to play with the script variables (global or in the dust removal section), but without any explanation of their meaning it's a trial and error exercise.
I attach samples presenting the issue - original Vuescan raw file, converted file and the file after dust removal. All have been downsampled to fit the github size limit.
If you could provide the explanation of variables used in the script, I can try to find better values. Now denoising is a pain, as I need to manually remove all the false positives.
Hello Signyts,
Thank you for excellent work on the script. I started to use it recently to convert Vuescan RGBI files. And the issue is that the script identifies dust and scratches on contrasty edges in the images, i.e. candle flames. After running dust removal macro the image does not look good. Maybe there is a way to play with the script variables (global or in the dust removal section), but without any explanation of their meaning it's a trial and error exercise. I attach samples presenting the issue - original Vuescan raw file, converted file and the file after dust removal. All have been downsampled to fit the github size limit. If you could provide the explanation of variables used in the script, I can try to find better values. Now denoising is a pain, as I need to manually remove all the false positives.
Thank you
samples.zip