Jupiter 17.11.12 - First go with WinJupos Derotate AVI

[theo]

I know a lot of you use WinJupos so I wanted to have a try myself on some captures taken on the 17th November.

These were only 2 min AVI's at 60fps but I wanted to see if it could coax out a bit more detail in the final processed image.

I went with the derotation of the AVI generating a ims file from a centre section of 1000 frames.

Both derotated and non derotated were processed in AS2 and wavelets in R6 with final processing in CS3.

On my monitor I could see a slight improvement in the derotated image especially on the poles but this may be difficult to see here.

Overall for the extra processing time I think the small improvement on a 2min capture is definately worth the effort. Next time I would like to capture maybe a 5min run and see how that goes. Not too sure how to deal with this in WinJupos though as I think its allows a max 2Gb AVI file. I assume this larger file size would require splitting and then each section derotating and then a rejoin? Are separate ims files required for each segment?

Any advice on this would be much appreciated.

Pete.

[cgarry]

I did quite a few tests de-rotating colour AVI files and came to the conclusion that I got a better result if I did not bother and just stacked in AS!2. That was even true with a 10 minute capture!

Cheers,

Chris

[scoobee]

liking the images, sorry cannot offer any advice

[Alan]

I must say Pete that any difference is very, very small at least to my eyes, but congrats on producing such superb images.

Alan

[theo]

Many thanks. The difference is small indeed and I assumed that this is due to the use of a 2 min capture where rotation blurring is minimum. Interesting to hear your results Chris of tests with longer colour files.

Pete

[Baz Pearce]

Excellent images both. Such little difference between the 2.

Baz.

[cgarry]

Hi Pete,

This is the thread with the 10 minute capture: http://stargazerslounge.com/topic/166912-jupiter-03112012-0057/

I was very surprised that the capture did not have more rotation problems. I do wonder if better quality data might work better with derotation.

Cheers,

Chris

[theo]

I agree Chris. I'm still hoping for that night of great seeing before this season is done. Great image that Chris...10mins!!

Pete

[cgarry]

I'm still hoping for that night of great seeing before this season is done.

You are not alone there!

Chris

[Josh Wilson]

Great images theo! Very sharp considering it was ten minutes long. Interesting activity on the left-central meridian.

[theo]

Thanks Josh, for info these were 2min captures.

Pete

[PH-R]

Great detail & thanks for sharing!

[sologuitarist61]

Very nice - well done!!

[Space Cowboy]

Pete have you tried de-rotating stacked images?

I find this a lot easier. Basically I stack say 500-1000 frames from each 2-3 min avi taken over a 10-20 min period then run them through wavelets before putting them through winjupos. So if you are doing 2 min captures then thats 5-10 stacked images. You wont see a huge difference in detail but there willl be a massive reduction in noise. Obviously if seeing is poor on some captures they are not worth including.

To see any real improvement de-rotating avis you need to be using at least 5-6 mins worth.

[neil phillips]

Hi Pete, I just love the smooth natural texture youve achieved here. Its just pure silk, great sharpening routine. less can so sometimes be more This is the second shot ive seen on here thats had that magic texture. Jimness was the other. BTW they look pretty similar apart from one thing, the blueing around the limbs is far less on the winjupos version. For that reason winjupos wins hands down

Nice one

[Freddie]

Very nice. Good job.

[theo]

Stuart, many thanks for your advice. I will give Stacked images a try and see how it goes. Also a longer Avi run is on the cards next clear night.

Neil, i must thank you for your comments here, really appreciate it :-)

Pete

[theo]

Stuart, many thanks for your advice. I will give Stacked images a try and see how it goes. Also a longer Avi run is on the cards next clear night.

Neil, i must thank you for your comments here, really appreciate it :-)

Pete

[neil phillips]

Pete have you tried de-rotating stacked images?

I find this a lot easier. Basically I stack say 500-1000 frames from each 2-3 min avi taken over a 10-20 min period then run them through wavelets before putting them through winjupos. So if you are doing 2 min captures then thats 5-10 stacked images. You wont see a huge difference in detail but there willl be a massive reduction in noise. Obviously if seeing is poor on some captures they are not worth including.

To see any real improvement de-rotating avis you need to be using at least 5-6 mins worth.

In theory yes that would seem to be correct because rotation is small untill you start getting into the many mins. However i think this needs to be tested over time, the reason i say this i could swear i could see a sharpening effect over a 2 min capture recently. I admit like the placebo effect it may have been my imagination. but i could swear the effect was quite dramatic, if you think how much it would take to sharpen a hairline feature on jupiter, could these hairline features be responding to small rotation movements ? I have no idea, but would have thought its possible. Also the higher the focal length the greater the effect i am talking about will show, 4 mins at f20 could be 8 mins at f40 for example.

[cgarry]

In theory yes that would seem to be correct because rotation is small untill you start getting into the many mins. However i think this needs to be tested over time, the reason i say this i could swear i could see a sharpening effect over a 2 min capture recently. I admit like the placebo effect it may have been my imagination. but i could swear the effect was quite dramatic, if you think how much it would take to sharpen a hairline feature on jupiter, could these hairline features be responding to small rotation movements ? I have no idea, but would have thought its possible. Also the higher the focal length the greater the effect i am talking about will show, 4 mins at f20 could be 8 mins at f40 for example.

I could believe that. When we effectively do de-rotation with AS!2 we are just relying on the alignment points tracking features on the planet and shifting them back to where they are in the reference frame. WinJUPOS de-rotation uses a much more accurate method that does not rely on tracking features in the same way, so fine features will naturally be shifted to where they should be without the need for them to be detected.

The data I used for my tests was quite (very) poor and had no fine features in it, this is probably why I could not get an improvement when I used WinJUPOS to de-rotate the AVI file. So we are back to needing a night of good seeing to test this theory out...

Cheers,

Chris

[theo]

WinJupos is very new to me so I intend to try different routines to see what works best for my set up and processing. As Neil mentions I also thought I could see some improvement in a 2min capture in respect to some of the finer detail (on my monitor) but granted its difficult to pick this out on here. Maybe some of that is down to the jpeg conversion process from the Tiff.

Do most if you post in Jpeg format?

Pete