General Processing (and Registax) Tips

[Macavity]

General Processing (and Registax) Tips. Are there any (and I apologise for not looking too HARD!) general tips re. the merits of various processing options? I can (have) "fiddled around" with stacking, wavelets etc. But I sense I might be better off NOT "reinventing the wheel"?

Just wondering about "Rules of thumb", re. Lunar (or whatever) imaging. When to "stack" (how many frames etc.) OR when to "post process" etc. Sometimes I quite LIKE the result of these various latter, but sometimes I go much to FAR (introducing artifacts etc.), I suspect...

Any (LINK TO) "typical numbers" (frames numbers , wavelet values etc.) to work from?

[rawhead]

Will keep an eye on this thread...

[beamer3.6m]

The first two wavelets are the main ones to use and only very gently.

When stacking I usually go no lower than a 75% quality level.

[Andrew*]

Re: the wavelets, you just have to experiment. Sometimes you only need very little, sometimes you can get away with quite a lot. It also depends on the target. Pics of e.g. Saturn in bad seeing need loads of the 6th wavelet and as you go up, less and less. Pics of the moon in good seeing need a little of number 1 and nothing else. At least - that's been my experience

It just depends.

Andrew

[themos]

My experience: low wavelet numbers for when you have 1000s of frames, high wavelet numbers when only a few 100s.

[daz]

There's no 'typical' unfortunately Chris.

For Planetary, here's what I generally do:

1. Use VirtualDub to split into individual frames.

2. Run these through Ninox

3. Start reviewing the sorted frames at about 75% of the way through the filenames until you come to a point you are happy at for discarding the rest

4. Move the discarded frames

5. Recombine in VirtualDub

6. In Registax, set the quality down low to include all your frames

7. Optimise using a search area of 10

8. Generate a reference frame of about 10% of the total frames

9. Apply a gentle sharpening

10. Optimise again

11. Stack

12. Wavelets

Seeing is so important as is the quality of the data. Also, see Trevors thread on FFT here