hmm still trying to figure out the best processing method.. I think it really comes down to seeing and the quality of the image in the first place.
As an example, here are 2 images from the same avi from last night (the same avi that the teaser in general chat came from).
Processed identically (convert to bmp's, split into r/g/b), except for the following differences:
Image on the left
- Harder wavelets on each channel after stacking in registax
- ME deconvolution on each channel (7/1.4) in AstraImage
Image on the right
- Medium wavelets on each channel after stacking in registax
- LR deconvolution on each channel (7/1.4) in AstraImage
Both gamma reduced to 0.7 in AstraImage after recombine.
That's the only processing.. normally I would do a bit extra in Photoshop afterwards, but for the purposes of this example I left as is.
My preference is the right one, the one on the left looks overprocessed. But that's using identical iterations/curve radius. So what I'm getting at, is because the data was already quite sharp, the hard wavelets in registax did most of the processing the image could handle.
The ME deconv just overprocessed it.
I need to go back and try different methods on the "hard wavelets" version, to process it to a point where it doesn't look overprocessed. Then we can see the difference between:
a) hard wavelets in registax and soft processing in AstraImage, versus
b) medium wavelets in registax and harder processing in AstraImage
On lesser quality data, i've done the same steps as above and preferred the hard-registax version. So it's important to process based on the quality of your raw data.
Anyway.. just rambling
