Jupiter 21st April - Stacking Best Frame % Comparison

[Bushy69 (Justin)]

Hi All,

Here is the same AVI stacked in Registax6. Total Frames was 5471 (about 75 seconds), and I was interested to see what effect stacking at various best % frames would have on final image.

The sequence is:

10% Best Frames (547)
20% Best Frames (1094)
40% Best Frames (2188)
60% Best Frames (3282)
Final (which is 60% Best and slightly heavier processing).

Mak Cass 180 Pro
2 X Barlow
HEQ5Pro Mount
ASI120MC-S
Firecapture
Registax6

I suspect outcome is based on seeing conditions, as on a crappy night, the best 20% might yield a better result than best 60% simply due to the fact there are a lower number of good frames. Conversely, on a good seeing night, where there are a lot of good frames, a higher % should yield a slightly sharper image.

From the sequence I shot, not a great deal of difference IMO (I can really only see it when I run them frame by frame overlayed).

Cheers,

Justin

[Atmos (Colin)]

That’s really interesting. Just looking at them in the sequence I cannot tell a different... viewing on an iPhone so not the most scientific comparison hehe

[Bushy69 (Justin)]

Hi Colin,

Viewing on my PC I can only just tell the slightest of difference... moving frame by frame... 10% is coarser than 60%... but we are talking tiny change here.

I do believe if seeing conditions are good, a higher % of best frames would result in a cleaner image... but I also suspect that there is a point of no significant difference. My gut feeling it is around the 30% to 40% mark for 4k to 5k frames. Purely subjective though.

Cheers,

Justin

[Dennis]

Thanks for the comparison Justin, it is quite interesting to see the various examples side by side.

I'm not sure if it wishful thinking, but I believe that I can just make out some very subtle differences. It appears that the noise reduces from L to R as you stack more frames (improved S/N) and there is slightly more "crisp" detail on the 10% image.

I suspect that if you just showed me these one at a time, in random order, I would probably not be able to pick them.

I guess on a night of good seeing the detail would be sharper and more clear on the best frames in the stack and this should then be reflected in the final IQ.

Good stuff!

Cheers

Dennis

[Bushy69 (Justin)]

Hi Dennis,

You have very good eyes. You are exactly right, from left to right, it seems to be a noise reduction issue, with a bit of smoothing occurring.

I also agree that the 10% had a "crisp" feel to it, and this can be seen when viewing/alternating frame by frame. I think the coarser image can highlight certain features.

I have a very fast PC, so the align and stack did not take too long, but for slower PC's, it could be beneficial to use a lower best frame % to speed up processing, with virtually no noticeable detail loss.

Justin