Hello,
In attempts to use mureDenoise, my results consistently feature what looks like a high frequency noise consisting of light spots. Some call this orange peel. I've used the tool which estimates the gain and noise parameters to use with the script.
I read one thread mentioning possible noise issues if trying to filter an integration that was built with weighted averaging or anything other than simple averaging. I've tested that out and there was no improvement.
I've tried filtering a single frame and still observe the high frequency noise.
The camera is the ASI1600 pro. I have new bias and dark frames and flats and lights. None of this make any difference.
The interpolation method during registration was Lanczos3 and set to that in the script.
The estimate tool param results used are below.
Gain: 0.095 e-/DN
Gaussian Noise: 24.95 DN, tried variations from 20 to 40, no improvement.
Variance scale used: 1.0, also tried 0.9, 0.8, 0.7, 0.6. This HF noise reduced some, however the image became distorted and too rolled off and not usable.
The screen capture of result of integrated (21 frames) is at 100%, not resampled.
Any suggestions would be greatly appreciated.
thanks,
Steve
In attempts to use mureDenoise, my results consistently feature what looks like a high frequency noise consisting of light spots. Some call this orange peel. I've used the tool which estimates the gain and noise parameters to use with the script.
I read one thread mentioning possible noise issues if trying to filter an integration that was built with weighted averaging or anything other than simple averaging. I've tested that out and there was no improvement.
I've tried filtering a single frame and still observe the high frequency noise.
The camera is the ASI1600 pro. I have new bias and dark frames and flats and lights. None of this make any difference.
The interpolation method during registration was Lanczos3 and set to that in the script.
The estimate tool param results used are below.
Gain: 0.095 e-/DN
Gaussian Noise: 24.95 DN, tried variations from 20 to 40, no improvement.
Variance scale used: 1.0, also tried 0.9, 0.8, 0.7, 0.6. This HF noise reduced some, however the image became distorted and too rolled off and not usable.
The screen capture of result of integrated (21 frames) is at 100%, not resampled.
Any suggestions would be greatly appreciated.
thanks,
Steve
Attachments
Last edited: