Jupiter Dec 1st Decent UK seeing at last!
Posted 02 December 2012 - 01:34 PM
Taken at 23.29 GMT f26 @30fps Auto Dob 250 & DFK. AVI run through PIPP quality selected before stacking in AS!2.
Posted 02 December 2012 - 07:54 PM
Posted 03 December 2012 - 01:59 AM
Posted 03 December 2012 - 06:44 AM
Cheers Glenn! I know what you mean about AS!2s ability but I have seen an improvement when the avi has been "polished" by PIPP first. Maybe it just helps AS!2 perform more efficiently? There were alignment cracks when using box 50 but after using PIPP this did not happen. Maybe its just a placebo effect??
Posted 04 December 2012 - 06:20 AM
An AP size of 50 sounds really small to me - but I'm not sure of the original recording size, it is smaller than this, right? Have you tried using a size of 100 and just let AS!2 handle the quality estimating part? It would be interesting to see the differences.
>There were alignment cracks when using box 50 but after using PIPP this did not happen. Maybe its just a placebo effect??
It could be that the very bad frames were thrown out, those that might cause AS!2 to have some problems with the likely too small AP size. But I still believe the small AP size is probably not the best idea in the first place.
Posted 04 December 2012 - 10:44 AM
Posted 05 December 2012 - 07:06 AM
Cheers Emil! Yes I do allow AS!2 to analyse the frames before setting the alignment points on the 1st quality selected frame. When seeing is good I use box 50 which gives me about 50 alignment points (sorry forgot to say these are 1.5x drizzle). If seeing is below average I use box 100 as 50 usually produces cracks on poorer data. So you are saying there is no advantage using more (smaller) alignment points even on good data? I use auto alignment too btw.
I have to say normally I don't use PIPP before AS!2 but with good data I sometimes try it.
Posted 05 December 2012 - 07:35 AM
Not really. Generally speaking those small AP sizes are not that good on Jupiter. But if you have a really small Jupiter and/or very high signal to noise ratios, then 50 might actually be quite big. I was trying to generalize that 50 is always bad. It's not, it depends on many things.
Whay you should do to properly test it out, is just try to place the APs manually like shown in this image: http://www.astrokraa...info/jup_ap.jpg
And just play around with the sizes a little bit. First create a stack of AP size 100, then one with 75, then one with 50. And finally carefully compare them (this is not easy!). Do the same for the blue and green channels, as especially blue is often a bit more noisy and requires larger AP sizes.
Even though images with smaller APs might look sharp, they might be not as sharp as they could have been with larger APs.
Posted 05 December 2012 - 09:23 AM
Thankfully being a user of the colour DFK I don't need to worry about the RGB channels.
Posted 05 December 2012 - 10:00 AM
Imagine that you used small APs right on the edge of Jupiter. AS!2 basically tries to minimize the difference between the current image and a reference image. But with a small AP, the compared images are relatively small and are more or less divided by an almost straight line (because the AP is small, the edge hardly looks curved any more) with the planet on one side, and black space on the other side. If you try to minimize the difference between these two images, the effect that the almost straight line has is really big. Any offset right ALONG this line sounds like a good deal to AS!2, the contrast inside the planet is much lower, so that won't contribute nearly as much as the division between the black space and Jupiter. Imagine the division being slightly warped by the seeing - which happens all the time - and AS!2 will put way too much emphasis on getting the slightly curved line back in what it thinks is the right place.
So in the end you can be sure to end up with a very sharp edge of Jupiter, but close by just on Jupiter it might actually be rather blurry. It perfectly tracked along the line, but did not really pay too much attention on what was going on inside Jupiter. You might not notice this at first, but when one blurry AP and a sharp AP meet, there is a chance of some small seem artifacts occurring. So not only is there a chance the image is more blurry in certain places, it can actually introduce some small artifacts in others.
It is impossible to predict what will happen when AS!2 stacks. A lot depends on the seeing conditions, but generally speaking bigger APs and placed a bit away from the edges are a better idea than many small APs all over the edge.
Posted 05 December 2012 - 02:20 PM