• Welcome to Orpington Astronomical Society.
 

News:

New version SMF 2.1.4 installed. You may need to clear cookies and login again...

Main Menu

Jupiter Thur 17 March

Started by MarkS, Mar 17, 2016, 21:26:26

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

MarkS

Jupiter is looking quite good tonight - seeing is reasonable as well  ;)

There seems to be a shadow passing across right now (9:25pm) - Calisto apparently.

Mark

RobertM

The shadow happens to be clouds here :(

I hope you manage to get something.

Robert

MarkS

A very quick process of some of the data. Imaging Source DBK21 camera with 2x Barlow on Celestron C11.

500 frames shot at 15 frames/sec stacked using AutoStakkert.  Deconvolved using my own software.



I also have some data showing the red spot shot at 2pm this morning just before the clouds came over.

Mark

The Thing

Your Jupiter is looking excellent. What percentage of frames were keepers?

Great to see you back on the planet trail, get that technique honed, Mars will be at it's best at the end of May and Saturn will follow in early June.

Seeing was excellent here, will be posting something later! I'd like to have a go with your deconvolution software - is it something you could share?

Carole

Another fine Jupiter Mark.  Well done.

Carole

MarkS

Thanks Duncan, Carole.

I kept 50% of 1000 frames.  I'm looking forward to a good Mars season this year - a close opposition  ;)

Yes - you can have a copy of DCON - in fact I'm sure you took a copy at a DSC when I demonstrated it.  I'll put together a tutorial because it has a few quirks and foibles.  But it allows you to do a few things I've never seen in any other software.

Mark

MarkS

Conditions deteriorated in the early hours of Friday morning but I still managed to get a reasonable image just before the clouds came in.

Again this is 500 frames at 15frames/sec.  DBK21AU04 on Celstron C11 with 2x Barlow.



Mark

P.S. Rick - is it possible to move this thread to Astrophotography?

Fay

It is healthier to be mutton dressed as lamb, than mutton dressed as mutton!

MarkS

I've put my first image through the colour matrix processing stage (referred to in this thread http://forum.orpington-astronomy.org.uk/index.php?topic=10688.0).
I used the colour matrix for the Sony A7S, which probably isn't correct but maybe is not too far off.  I need to image a colour chart with the DBK21Au04 then I can generate a colour matrix specific to that camera.  Anyway, here is the result:



I rotated it as well because seeing Duncan's version made me realise I flipped mine left-right instead of up-down. It now has a consistent orientation to my red spot image.

Mark

P.S. Thanks Rick, for moving this thread to astrophotography.

Carole

#9
Looking good.  I wonder whether you might be able to sharpen the image a little to bring out some of the detail.

Carole

The Thing

Quote from: MarkS on Mar 20, 2016, 00:16:48
I've put my first image through the colour matrix processing stage (referred to in this thread http://forum.orpington-astronomy.org.uk/index.php?topic=10688.0).
I used the colour matrix for the Sony A7S, which probably isn't correct but maybe is not too far off.  I need to image a colour chart with the DBK21Au04 then I can generate a colour matrix specific to that camera.

The colour looks very nice using that method, more subtle. I am balancing the histogram RGB channels when I capture at the moment but that sort of assumes that overall Jupiter is white. Probably not. However it's a starting point.

MarkS

Quote from: Carole
Looking good.  I wonder whether you might be able to sharpen the image a little to bring out some of the detail.

It's already sharpened as much as I dare: it already has too much noise.

Quote from: The Thing
The colour looks very nice using that method, more subtle. I am balancing the histogram RGB channels when I capture at the moment but that sort of assumes that overall Jupiter is white. Probably not. However it's a starting point.

I do more or less the same thing.  The colour matrix has to be applied to white balanced data. The only problem with the colour matrix is that it does increase noise quite significantly. 

Mark

Fay

I like the softer look, I think too much sharpening would make it look too harsh
It is healthier to be mutton dressed as lamb, than mutton dressed as mutton!

Carole

When I do sharpening I don't necessarily sharpen the whole image, I just do a selective sharpening so only bits get sharpened.

Carole

MarkS

This one has 1500 frames stacked:


DBK21AU04 on Celestron C11 with x2 Barlow at 15 frames/sec.

Mark

Fay

It is healthier to be mutton dressed as lamb, than mutton dressed as mutton!

Carole

Looking even better.

Carole

The Thing

Excellent job. Even a bit of detail on the moon.

MarkS

Here's the final attempt - 4000 frames out of 5000 (5.5 minutes):



Compared with the previous version (1500 frames):



And the original (500 frames) which needs 180 degree rotation:




The interesting thing is that Jupiter rotates quite a lot in 5 minutes (8 pixels at the equator at this scale) but AutoStakkert2 has clearly coped incredibly well and lined up all the features.  The AutoStakkert site has a link to an article ( http://www.planetary-astronomy-and-imaging.com/en/maximum-video-time-jupiter/ ) suggesting that Winjupos is not needed a lot of the time (I've never used it).

You may be right Duncan, that I've caught some detail on the moon but I'm a bit cautious in claiming that.

Mark

Carole

Latest version looks really good.

Carole

MarkS

#20
Quote from: MarkS
Yes - you can have a copy of DCON - in fact I'm sure you took a copy at a DSC when I demonstrated it.  I'll put together a tutorial because it has a few quirks and foibles.  But it allows you to do a few things I've never seen in any other software.

Here it is with a brief tutorial included on the functionality required for deconvolving a planetary image (with example images including the 17 March Jupiter) - also works really well for deep sky images as a final stage.
http://www.markshelley.co.uk/webdisk/DCON03.zip

Right click the above link and do a "Save As".  Make a folder on you drive and extract the contents of the zip file there.  Take a look at the readme.txt and the tutorial document.

If you try it, let me know how you get on - especially if you get better results than Registax.

Mark

Fay

my computer advises me not to download your file Mark,as it can be dangerous! not had that warning before 
It is healthier to be mutton dressed as lamb, than mutton dressed as mutton!

RobertM

What an excellent capture of Jupiter, it sure does look like a small dust mote on that moon though ;)

Robert

JohnP

Definitely looks like detail on the moon on my monitor - very nice shot - How does it compare with your previous attempts - I seem to remember you having some great captures before...

John

MarkS

Quote from: JohnP
Definitely looks like detail on the moon on my monitor - very nice shot - How does it compare with your previous attempts - I seem to remember you having some great captures before...

I think it's my best ever.  My previous best was this one, taken when Jupiter had a band missing:


AutoStakkert!2 has a lot to do with it - I'll try running that old data through AS!2

Quote from: Fay
my computer advises me not to download your file Mark,as it can be dangerous! not had that warning before 

Interesting.  Maybe because it's a zip file or maybe because it's a zip file with an executable in it.  In general it's not a good idea to download stuff unless you are sure of exactly what it is and you trust the author.

Mark



The Thing

The improvement is palpable. It's amazing what improvements there have been in the processing that enable Joe Bloggs to get amazing planetary images. It's a bit magical to see a defined sharp image appear from a blurry video :)

Quote from: MarkS on Mar 22, 2016, 19:56:02
Quote from: Fay
my computer advises me not to download your file Mark,as it can be dangerous! not had that warning before 
Interesting.  Maybe because it's a zip file or maybe because it's a zip file with an executable in it.  In general it's not a good idea to download stuff unless you are sure of exactly what it is and you trust the author.

Mark
If it's like Win10/Chrome it's because 'this file is not often downloaded' or some such as twaddle based on the lists they compile of popular downloads. So anything astronomical will fail this check. Ignore it.

MarkS

Interesting that Ganymede in my older picture is so much brighter than Callisto in my recent one, even though they are not that much different in size.  So I looked it up:  Ganymede has an albedo of 43% whilst Callisto is 20%.  Now I know!

Mark

Fay

Really nice Mark,  I have dug my colour SPC webcam out to see what I could get with it, not sure on the quality
It is healthier to be mutton dressed as lamb, than mutton dressed as mutton!

JohnP

Fay - you got rid of your high speed B&W camera?

MarkS

#29
Quote from: The Thing
The improvement is palpable. It's amazing what improvements there have been in the processing that enable Joe Bloggs to get amazing planetary images. It's a bit magical to see a defined sharp image appear from a blurry video :)

Take a look at the incredible improvement Damian Peach achieved using AutoStakkert and WinJupos
http://www.damianpeach.com/barbados10rw.htm

Interesting that he can now use up to 15 minutes of Jupiter data in one image.  I must give that a try.  The great thing about so much data is that the signal to noise level reaches very high levels and allows the deconvolution to be pushed to extreme limits in terms of the maximum amplification that can be applied to data in the wavenumber (Fourier) domain.

I'm sure there's more detail I can extract from my recent image - I just need to process 12000 frames instead of 4000  ;)

Mark

MarkS

Here's my final version - my first time using WinJupos:


It stacks 7000 frames from a total of 9000 over 11 minutes.

Compare to the previous 4000 frame (5.5minutes) version using AS!2:


For the latest version I created 3 frames using AS!2 - each from 3.5 minutes of data.  These were de-rotated and stacked in WinJupos and the result was then deconvolved.

My impression is that there is little difference probably because WinJupos does a lot of resampling of the data during the planetary de-rotation and the rotation to bring the North pole to the top.

Mark

The Thing

Excellent result. The moon shadow and the moon really show the difference, WinJupos appears to have smeared them a bit.

I have heard that Autostakkert does an even better job if the features are horizontal or vertical to start with, which is how I tend to capture Jupiter anyway. Maybe straighten the video before stacking?

I've never used WinJupos for the actual video processing as I've read it's not so good. I de-rotate single stacked and sharpened images, which sometimes makes a big difference and other times not. I've never been able to find out how WinJupos can know how much de-rotation to apply to a single frame. Yes the mid-point time in the file name or entered by you gives a target moment in the planets rotation to de-rotate to but the duration stacked each side of that point is unknown - you are never asked for it and I can't see it in the image tags, so how does it know how much smearing there is? I should try de-rotating unsharpened images as that is actually more logical - de-smear then sharpen.

Carole

Am I right in thinking that Winjupos de-rotates Jupiter somehow in the software?  If so, this would provide more data for processing and therefore more detail will be revealed and less noise instead of smearing it out. 
Also does this mean you can take a lot more data and not worry so much about rotation?

This is my non technical way of thinking of it.

I can certainly see more defined detail in Mark's Winjupos version, to the non WJ version.
Is it free?
Is it easy to use?

Carole


The Thing

Quote from: Carole on Mar 28, 2016, 11:14:31
Am I right in thinking that Winjupos de-rotates Jupiter somehow in the software?  If so, this would provide more data for processing and therefore more detail will be revealed and less noise instead of smearing it out. 
Also does this mean you can take a lot more data and not worry so much about rotation?

This is my non technical way of thinking of it.

I can certainly see more defined detail in Mark's Winjupos version, to the non WJ version.
Is it free?
Is it easy to use?

Carole


Yes, and yes, and Yes if you seek out tutorials on YouTube and elsewhere.

And I have just tested de-rotating unsharpened images and it works a lot better. So much for on-line tutorials!

Carole

I have just downloaded WinJupos and watched a tutorial.  Am I right in saying that it's the already stacked images that you load into Winjupos, not the videos themselves. 

Carole

The Thing

You can do either, but the consensus is AutoStakkert stacks better, so do that first.