Jupiter 03/11/2012 00:57

[cgarry]

Finally a clear night on a non-school night! On the face of it, It was not a bad night for planetary imaging with no wind and plenty of moonlight to work by. The downside was it was virtually raining dew and as ever the seeing was nothing special with the focus point jumping about!

I did not bother using the mono camera as I rated my chances of grabbing a half decent sequence of RGB captures as next to zero. Instead I went with the colour (DFK) camera.

I took quite a few 2 and 3 minute captures, but it became obvious that they were only going to give very poor results after some quick processing.

The image here comes from a 10 minute capture at 60 fps (about 36000 frames). This capture was really only taken as I wanted an example of a very large AVI file for PIPP development, but I had a go at processing it anyway. No WinJUPOS used. Simply cropped, centred and the best 1200 frames selected with PIPP, 100% of those frames stacked with AS!2 and some final processing in PS.

Jupiter

03/11/2012 00:57

Cambridge, UK.

C14, x2 barlow, DFK 21AU618.AS

Cheers,

Chris

[cfpendock]

It seems to me that this is a very nice photo, and if no de-rotation software was used, amazing detail for a 10minute exposure considering the moon. But could you please tell me, what is AS!2? Is it similar to Registax? thanks chris

[Freddie]

Not so bad given the conditions Chris. Good job.

It was as bad as I've seen it last night here in Kent. I gave up early and went in for a cup of tea to warm up !!!!

[cgarry]

It seems to me that this is a very nice photo, and if no de-rotation software was used, amazing detail for a 10minute exposure considering the moon. But could you please tell me, what is AS!2? Is it similar to Registax? thanks chris

Yes, AS!2 is another stacking program like Registax (http://www.astrokraa...utostakkert.php). A lot of planetary imagers seem to have moved over to it from registax.

I am surprised that a 10 minute capture time was not too long, and looking at the quality selection log file it does look like frames were selected from all parts of the AVI file, not just from a small section. It has been noted before though, that AS!2's multipoint alignment does go some way to derotate colour captures by aligning the features. I might try derotating in WinJUPOS and compare results at some point.

Yes Freddie, the conditions were far from idea but a clear sky of any kind is a welcome thing these days!

Cheers,

Chris

[Space Cowboy]

10 min avi now thats really pushing the rotation boat! Fingers crossed for tonight looks clear but not expecting great seeing. BTW PIPP just gets better the more I use it, works great for animations Chris after reading your post on Dror's thread.

[si@nite]

Nice image Chris! really gettin confused about capture lengths on jupe?

[Delboy_Hog]

Very nice! 10 minutes, wow. Blows my thinking (and longest capture so far) out of the water! 36000 frames - nicely done!

I think on the next clear night here I'll do a few runs of different length captures, from 30 seconds up to 10 minutes or 12 minutes and see how they compare.

Lovely image Chris!

[cgarry]

Yes guys, 10 minutes holding up so well was quite a surprise to me. I am in the process of running the same AVI through WinJUPOS to compare the results.

That is a good plan Derek, more tests are the way forward. Though I think it will not take long to fill a hard drive up with AVI files!

Cheers,

Chris

[astroman001]

Lovely image, I'm surprised that you get such sharp detail without de-rotating a 10 minute video.

well done

Peter

[space3]

super image

[cgarry]

For comparison, this version was created in the same way, but the 10 minute video was derotated in WinJUPOS first.

Cheers,

Chris

[neil phillips]

Hi Chris that is very suprising how the features havent blurred much on the non de rotated image. Makes me wonder how much shooting 9 one min avis for de rotation is needed. Compared to shooting say 3, three min RGB sequences, ( which is a lot less messing about ) and derotating the images. There will be some rotation, but am i wrong in assuming that done this way. There will only be about 3 mins of rotation ? because it will only be the seperate RGBs that aquire the 3 mins of rotation. If the RED is derotated to the GREEN, and GREEN to BLUE, Surely we are just left with 3 mins of rotation. Over the intire nine mins. Rather than de rotateing the 1 min REDS GREENS AND BLUES before final image de rotation and image colour combine ?

You just need some great seeing Chris for that monster scope. Good detail despite what is clearly not great seeing for you

[cgarry]

Hi Neil,

Right now I would accept half-decent seeing over the pants seeing we are having to live put up with. The C14 really does seriously suffer with bad seeing. Why does the jet stream insist on just sitting over the UK!

I think I am using this time wisely by getting up to speed with WinJUPOS and I think it will pay off when conditions improve. My current experiments have been de-rotating colour AVIs. Actually I could not get WinJUPOS to process a colour AVI, so instead I have been splitting the AVIs into their R, G and B channels and running them through individually. The upshot of this is that AS!2 does a very good job of effectively de-rotating videos and if it can handle frames from a 10 minute AVI so well then the frames from 3 minute AVI should be easy for it.

I wonder if your idea of one set of 3 mins each R, G and B could work. To do this in AS!2 I would use a frame near the end of the R AVI as its reference frame, a frame near the centre of the G AVI as its reference frame and a frame near the start of the B AVI as its reference frame. Doing this I would expect R to G to have 1.5 minutes rotation and G to B to have a further 1.5 minutes rotation (assuming 0 time between captures!). I would like to see this approach tried and compared to the 3 x 1 minute RGB sequence approach which does feel like an awful amount of faffing about to me, both at capture time and processing time. I feel some more experiments coming on next weekend as I move on to my mono camera tests - weather permitting...

Cheers,

Chris

[neil phillips]

Hi Neil,

Right now I would accept half-decent seeing over the pants seeing we are having to live put up with. The C14 really does seriously suffer with bad seeing. Why does the jet stream insist on just sitting over the UK!

I think I am using this time wisely by getting up to speed with WinJUPOS and I think it will pay off when conditions improve. My current experiments have been de-rotating colour AVIs. Actually I could not get WinJUPOS to process a colour AVI, so instead I have been splitting the AVIs into their R, G and B channels and running them through individually. The upshot of this is that AS!2 does a very good job of effectively de-rotating videos and if it can handle frames from a 10 minute AVI so well then the frames from 3 minute AVI should be easy for it.

I wonder if your idea of one set of 3 mins each R, G and B could work. To do this in AS!2 I would use a frame near the end of the R AVI as its reference frame, a frame near the centre of the G AVI as its reference frame and a frame near the start of the B AVI as its reference frame. Doing this I would expect R to G to have 1.5 minutes rotation and G to B to have a further 1.5 minutes rotation (assuming 0 time between captures!). I would like to see this approach tried and compared to the 3 x 1 minute RGB sequence approach which does feel like an awful amount of faffing about to me, both at capture time and processing time. I feel some more experiments coming on next weekend as I move on to my mono camera tests - weather permitting...

Cheers,

Chris

Very interesting Chris, to be honest i didnt even know AS/2 could affect rotation on images. So thats new information to me.

But what i am also interested in is what the total rotation would be seen using winjupos, feeding in the midpoint time of a 3 min red, then green and blue. and just aligning de rotateing and combining. I have already done this and rotation doesnt seem a sever issue at all.

But i wonder how well this really could work, compared to the 1 min sequences. Yes there will be more rotation blur doing it the quick way. But is that rotation blur really a cause for worry or not, is what i am wondering.

And would the total rotation be just 3 mins, from 3 seperate 3min RGBs ? De rotated doing it the quick way ?

I am not actually sure. But i would have thought so. Yes ive seen how C14s hate poor seeing Chris. especially low altitude targets. But when your time comes. you should go a long way in front of what smaller optics can achieve.

[cgarry]

I see what you are saying Neil but I think the total rotation would be 6 minutes between red and blue, with the green rotation being half way between. With this approach I would expect the left side of the planet to be missing red data at the very edge and the right side to be missing blue data (assuming the planet orientation I have pictured in my head!). I think some tests are required to see how much this would be a problem in reality, unless you have some suitable data to play with already...

Chris

[neil phillips]

I see what you are saying Neil but I think the total rotation would be 6 minutes between red and blue, with the green rotation being half way between. With this approach I would expect the left side of the planet to be missing red data at the very edge and the right side to be missing blue data (assuming the planet orientation I have pictured in my head!). I think some tests are required to see how much this would be a problem in reality, unless you have some suitable data to play with already...

Chris

Right Chris, Edge problems are such that often we do get slight artifacts on the edge, some prefer to smooth them, some do not. Either way, i think the amount of missing data would be minimal.

As such i think its intirely viable to take some the quick way. But also worth doing some the hard way. Especially on the very best nights. This is what i plan to do i think after further discussion with you. And seeing this post. Also might be worth doing some two min captures for total capture time of 6 mins rgb. Plenty of frames can be got with those time scales especially at 60 fps. With quick and easy processing. Minimal data loss. A compromise of sorts perhaps. I dont think i plan spending all my time doing 9 runs. every colour capture. what with cleaning up every AVI before its fed into AS/2 and the winjupos process itsefl, I think its getting a bit time consuming with all these processes just doing a RGB run. So i think its a good plan, to do some 1 min 2 min and 3 min sequences and compare results on a night of calm air. As soon as i can i will do that, but i have had some probelms lately, still havent fitted my new focusser, as i need another focus plate to make it fit.

I may refit the old focusser if weather shows signs of improvement before i get the new focusser fitted in about two weeks.

It also doesnt have fine focus on it, as i didnt get it fitted. i was going to use my old fine focus and retro fit it on the new focusser. But i am not sure i want to fit such a old part on a new focusser.

I believe i can fine focus without the 101 reduction fine focus. Though it will be more tricky. I will possibly fit a fine focus on it at a later stage, if i can not adjust to using one without it.

As soon as i have some timescale comparisons taken on the same night. Ill post up if your interested. Or if you do some i will be interested in seeing them

[Space Bat]

Very nice

[cgarry]

Hi Neil,

I like your approach of having a slightly compromised quick method as well as the generally accepted method. Capturing 3 x RGB sequences and especially the processing part it is exceptionally time consuming and I can see myself reserving this approach for nights when the atmosphere is calm enough to make it worthwhile. To be honest on the bad nights that we have had lately I have just been using the colour camera and not even bothered using the mono camera at all, definitely a further compromise but it so much quicker and actually more likely to get a good result if the seeing improves briefly.

Another idea I have had is to modify PIPP to analyse all the AVIs captured in an evening for quality and just output a quality report of some sort so that I can see at a glance which AVI files are the best quality and start by processing them. Running like this PIPP should be quick as it would not be actually processing the files or writing them out again. So with this approach I would capture lots of 1 minute RGB sequences (say 12) back to back and then pick the best 3 back to back sequences to process by looking at the quality report. This is just an idea I have had when looking at a drive full of AVI files after a capturing session and wondering where to start, though it does rely on not changing the focal length I guess. This is just an idea right now that may well come to nothing.

Yes, please do post any comparisons you make and I will do likewise. Good luck with your focuser too.

Cheers,

Chris