Jump to content


Photo

Question on derotation

  • Please log in to reply
31 replies to this topic

#1 Mert

Mert

    Skylab

  • *****
  • Posts: 4219
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 12 January 2013 - 06:52 AM

Hi all,

I think this theme has been discussed extensively, with
a lot of exampes and tests by Darryl and others.
Now I'd like to do the following:

I have 3 images, processed and nicely the same size etc.
Each image was taken 180 seconds long and evenly spaced
60 seconds in between.

Q: Can these images be derotated such that I can stack
these 3 images together?

I have seen AVI derotation etc but can't make this
happen :confused:
If someone would be so kind as to show me the right
direction to take I would be more then grateful!!!!!
Thanks a lot for any insight!

#2 Kokatha man

Kokatha man

    Fly Me to the Moon

  • *****
  • Posts: 6800
  • Joined: 13 Sep 2009
  • Loc: "cooker-ta man" downunda...

Posted 12 January 2013 - 07:59 AM

Hi all,

I think this theme has been discussed extensively, with
a lot of exampes and tests by Darryl and others.
Now I'd like to do the following:

I have 3 images, processed and nicely the same size etc.
Each image was taken 180 seconds long and evenly spaced
60 seconds in between.

Q: Can these images be derotated such that I can stack
these 3 images together?

I have seen AVI derotation etc but can't make this
happen :confused:
If someone would be so kind as to show me the right
direction to take I would be more then grateful!!!!!
Thanks a lot for any insight!


.....I think I sent you my tute Mert, and if you have the end-processed images from these 180 second long avi's then that is what the tute concentrates on..!

I'm not sure why you bothered to leave a 60 second spacing between each avi/capture if I'm reading you correctly because running one avi straight after the last would be optimal, but no bother really in your example:

Presuming you are talking colour images you simply place each processed image in WinJupos and enter the correct imaging site coordinates and then go into the Alignment Frame step and place the AF, remembering to check "colour" if it is a colour image we're talking about.....

Saving this creates a .ims file which is then...blah-blah-blah as per the tute instructions! :)

A couple of additions I've added to that tute recently for the sake of clarity are that you can process colour images just as well as mono channel images (I experimented recently with said) and also if your raw data can handle it, you should try using 1.5X or 3X drizzle in AS!2.....larger images are much easier to place the Alignment Frame around precisely to ensure good outcomes..! :)

Finally, it is also perfectly reasonable to use WinJupos to align the individual mono channels together from a "normal" mono capture (ie, the 3 minutes total capture run types) to get your RGB instead of (say) using AstraImage or Photoshop...

#3 wenjha

wenjha

    Vendor

  • -----
  • Vendors
  • Posts: 731
  • Joined: 24 Sep 2008
  • Loc: SuZhou China

Posted 12 January 2013 - 09:29 AM

I think derotation is more important for mono camera to align R G B file.
for color camera I don't see such benifit
and derotate 3 images and stack seems won't improve much

#4 Mert

Mert

    Skylab

  • *****
  • Posts: 4219
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 12 January 2013 - 11:06 AM

Hi Darryl,

Thanks a lot for your feedback, I think you did amil
the Tute, but my laptop crashed and I have a new SSD
in there.
As usual there are no backups made :bawling:
So if it isn't too much to ask for, would you be so
kind as to mail me your tute again??
I'll pm you my mailbox!
Thanks a lot, you really help me out with this!!

#5 Mert

Mert

    Skylab

  • *****
  • Posts: 4219
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 12 January 2013 - 11:42 AM

Hi Sam,

I was just trying to experiment and see if I can raise the
S/N ratio a bit, like having a 9 minute avi result.

#6 sfugardi

sfugardi

    Gemini

  • -----
  • Posts: 3060
  • Joined: 30 Sep 2006
  • Loc: CT

Posted 12 January 2013 - 11:52 AM

Mert, the most important part of WinJupos is the globe alignment and time stamp. You need to make sure N is the true Jupiter N and the P is the direction the global rotates to. Believe me, I've had it backwards for the longest time only to find the derotation going in the wrong direction, making the data smear even more. To be totally accurate each avi ref in your case will be 1.5min + the starting time. 180sec avi times sounds long to me. You could try cutting them into halves using VDub, then derotate and combine all 6. Just be careful of the new ref times if you try this. Darryl's instructions are the way to go, no question. Good luck and keep the Jupiter images coming

Regards,
Steve

#7 Mert

Mert

    Skylab

  • *****
  • Posts: 4219
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 12 January 2013 - 12:06 PM

Thanks a lot Steve for your feedback, I'll certainy take
this into account for!
Right now I'm clouded out so just experimenting a bit!
There is nothing better then good data you nearly don't
have to process at all.
I thought 3 minutes was ok for the aperture I'm using
and the EFL ( aroud 4200 mm )
However when I check the moons, for example Io, it is
really blurred by movement in 3 minutes!
Not that I could possibly get some albedo feature but
just it wuld be nice to have round moons.

#8 sfugardi

sfugardi

    Gemini

  • -----
  • Posts: 3060
  • Joined: 30 Sep 2006
  • Loc: CT

Posted 12 January 2013 - 12:17 PM

Mert, what f-ratio are you using? For sure, you will have to process the moons separately using a 1 small box alignment, then combine with the final globe. There is no way to get both aligned at the same time. Double the work, double the time. Just make sure you pick the best avis before burning hours and hours...

Steve

#9 sfugardi

sfugardi

    Gemini

  • -----
  • Posts: 3060
  • Joined: 30 Sep 2006
  • Loc: CT

Posted 12 January 2013 - 12:55 PM

Mert, at 4200mm fl, you are smearing approx 2.7px in 180sec. To be at 2px you would need to be 134sec. I am not sure what the max amount of smear AS2 can handle and still get a good image, but for sure don't pick alignment points anywhere near the edges

Steve

#10 Mert

Mert

    Skylab

  • *****
  • Posts: 4219
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 12 January 2013 - 01:42 PM

Hi Steve,

Good point!
I did a little trial and cut the original avi of 3 minutes
in 3 parts, calculated the midpoints etc and measured
each resulting image in Winjupos.
Then combined the 3 in the RGB tab, seems to be ok more
or less, but not convincing.
Each of the 3 images of 1 minute are rotated in between
them, it is clearly visible, switching between the 3, so
yes, I am smearing detail out :rainbow:
Then with the derotated combined image there are red
points in the black background, funnny?!
I'll try to do as you said and put the alignment points
for AS!2 not at the edges and see what happens.

Here you have the resulting image, left normal, right
WinJupos derotated and also in the next post a small
animation of the 3 parts of 1 3 minute avi.

Attached Files



#11 Mert

Mert

    Skylab

  • *****
  • Posts: 4219
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 12 January 2013 - 01:45 PM

And here the animation of the 3 sub-frames of 1 avi

Attached Files



#12 Kokatha man

Kokatha man

    Fly Me to the Moon

  • *****
  • Posts: 6800
  • Joined: 13 Sep 2009
  • Loc: "cooker-ta man" downunda...

Posted 12 January 2013 - 09:21 PM

I think derotation is more important for mono camera to align R G B file.
for color camera I don't see such benifit
and derotate 3 images and stack seems won't improve much


Just sent the tute Mert - let's know if it doesn't arrive as I'm having a bit of trouble with Kaspersky anti-virus atm...

Sam, I recently did some trials using rgb images (admittedly from mono composites but that shouldn't make any difference whatsoever! :)) and WinJupos worked fine with them being derotated and combined to enhance the end-image.....I actually did it to see if it was a bit easier for folks than trying to (for example) set 9 mono channel images in WinJupos instead of 3 colour images.....results were no less impressive! :waytogo:

#13 Mert

Mert

    Skylab

  • *****
  • Posts: 4219
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 13 January 2013 - 02:35 AM

Hi Darryl,

Your tute arrived, thanks a bunch :jump:
I was quit surprised by the amount of rotation between
the 1st minute and the 3d minute within the same avi.
It's easily visible in the animation above.

#14 Freddy WILLEMS

Freddy WILLEMS

    Fly Me to the Moon

  • *****
  • Posts: 7228
  • Joined: 13 Nov 2005
  • Loc: Hawaii, Honolulu

Posted 13 January 2013 - 03:06 AM

Amazing captures Mert, both nice processed!

#15 sfugardi

sfugardi

    Gemini

  • -----
  • Posts: 3060
  • Joined: 30 Sep 2006
  • Loc: CT

Posted 13 January 2013 - 08:54 AM

Mert, excellent results! The key is to have enough frames to work with, so I am not sure how many frames are in your 1 min avis. If you have time, it would be interesting to take the first 120sec as RGB1 and recut from the original to the last 120sec as RGB2. This way you still have enough quality frames to choose from. I am not sure if the 1min overlap would be an issue or not for WinJupos. Of course the avi ref times would be now different. It's nice to be able to work with good data. Thanks for posting

Regards,
Steve

#16 Mert

Mert

    Skylab

  • *****
  • Posts: 4219
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 13 January 2013 - 12:16 PM

Not quit sure what you'd like me to try to do now Steve!
I will upload 2 avi's which seem to have pretty decent
data in there but can't get it better then what AS!2 does.
I asume AS!2 takes into account for the rotation within
the avi?
Once uploaded I'll post the dropbox-links to the 2.
Each is 500MB big ( zipped 280Mb ) but are 3 minutes
worth of capture at 30fps, I hope you may want to give
it a shot!
They were captured with Firecapture, SER format and
Autoguided by FC, passed through PIPP to convert to avi.

#17 Mert

Mert

    Skylab

  • *****
  • Posts: 4219
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 13 January 2013 - 04:34 PM

Ok, right now I've uploaded 1 avi.
The next avi will follow lateron, I will post the link
to it so you can download if you like.

I would be more then happy to see what you can do with it!

Thanks.

Here is the link: Zipped AVI 3 minutes at 30fps 280MB

#18 zAmbonii

zAmbonii

    Vanguard

  • -----
  • Posts: 2420
  • Joined: 19 Jan 2008
  • Loc: Ypsilanti, MI

Posted 14 January 2013 - 12:54 AM

Mert, downloaded your .avi and I just have to say is that you had some great seeing, and got a great capture out of it!

I took your video and did a couple of things:
1) took the straight video and stacked 4000 out of the images. Then processed it in R6 the best I could.
2) Split the video in half, and stacked 2700 frames for the first half, and also stacked 2700 frames for the second half. I processed them with the same wavelet settings as I did in #1 above.
3) Took the whole video stacked 4000 frames and did a 1.5x drizzle on it, did wavelets (different from 1 and 2 above), and then resized it back down to the same scale as 1 and 2 to see if I could pull out more details.

At the bottom is an animated .gif of the the processing (I did a little color balance on it also).

Comments: I just want to say that I am not really sure what your goal is in trying to derotate your .avis. At the image scale you're capturing at, I am not sure that you will be able to see a difference in sharpness. If you're able to capture at 60fps, it might be better to do 60fps captures for 1.5 minutes instead to have enough frames to capture.

I would be completely thrilled with the result you got from your 6" at that image scale!

Attached Files



#19 Mert

Mert

    Skylab

  • *****
  • Posts: 4219
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 14 January 2013 - 08:45 AM

Hi zAm,

Great processing you did, I have to take a good look at
it when I get home tonight!

In the meantime, here is the 2nd link to the avi I did
as an example to reprocess with the animation from a
couple of replies above: 2nd avi

Maybe I forgot to upload the log-files from FC so you have
the exact capture times and statistics, if needed I can
do so when I get home tonight!

Edit: Ah, I forgot to comment, when capturing at 60fps,
the DFK or TIS cameras do produce a rather strong ring
effect on the planet's disk.
It looks like some sort of "echo" of the planets limb, it
has been strongly tested and discussed here on CN!!
I did do some trials at 60fps, it's great for moon pics but
on the planets, it gives a very strong limb-echo or
ring-artifact, I don't know how to name it.
Likely I have to upgrade to the ASI120MC like you did,
but for budget reasons ( I bought a concert guitar ) can't
afford it right now.
I am very grateful for the time you took to reprocess my
intents to image Jupiter, it seems like seeing was very
good that night ( the globe didn't move much, nor high
frequency turbulence which kills small detail )
Next time out I will shoot 2 minute avi's without waiting
between and of course Autoguided by FC :cool:


Thanks again!! :waytogo:

#20 sfugardi

sfugardi

    Gemini

  • -----
  • Posts: 3060
  • Joined: 30 Sep 2006
  • Loc: CT

Posted 14 January 2013 - 01:03 PM

Mert, I tried working with your data but had a terrible time trying to suppress the vertical and horizontal noise. For some reason, the image was in color in AS2 but went back to BW in Reg6. I'd say your processing, and Z's, is very good, and my usual tricks don't seem to work. Attached is my attempt at a 1500 frame stack.

Regards,
Steve

Attached Files



#21 Mert

Mert

    Skylab

  • *****
  • Posts: 4219
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 14 January 2013 - 03:03 PM

Hi Steve,

Maybe you have an old version of AS??
You might as well pass the avi through vdub and see
if that helps?
I just tried the same avi and for me it seems to work ok.
On my laptop if I play the avi, it looks like an
undebayered bw avi.
On the image you posted, did you cut the avi just in 3?
Was this the first, second or third part?
Did you do drizzle? since the image is soo big??

#22 zAmbonii

zAmbonii

    Vanguard

  • -----
  • Posts: 2420
  • Joined: 19 Jan 2008
  • Loc: Ypsilanti, MI

Posted 14 January 2013 - 03:46 PM

Mert, I totally forgot about the ringing artifacts that the IS camera had (knew there was another reason why I kinda avoided them ;)

I really would like to have seen what you could have come up with that night of imaging with maybe a little larger image scale. The .avis that you posted actually looked pretty good processed with a 1.5x drizzle. You may have been able to capture a little more detail than the drizzle at a larger capture size.

I processed the second .avi and below is a comparison between the 1.5x drizzle and not resized.

BTW: I did try a couple of nights ago to use my ASI120MC on Jupiter with my 6" but failed horribly. My mirror needed cleaning and I did that, put things together and collimated. But noticed something really wrong with a star image. I never really checked my secondary alignment through the focuser and it was really out of whack. I'm going to try and get that straightened out this afternoon. May have a chance to image tonight even though seeing is not supposed to be good.

Attached Files



#23 MvZ

MvZ

    Surveyor 1

  • *****
  • Posts: 1579
  • Joined: 03 Apr 2007
  • Loc: The Netherlands

Posted 15 January 2013 - 07:36 AM

Steve, you might try the latest beta version of AS!2 to see if that fixes your problem.
http://www.astrokraa...are/latest.html

Remember this is beta software, not a final release. If you find problem with it, please do let me know.

#24 sfugardi

sfugardi

    Gemini

  • -----
  • Posts: 3060
  • Joined: 30 Sep 2006
  • Loc: CT

Posted 15 January 2013 - 12:40 PM

Emil, thanks for the link. The program got stuck at the final step and never finished. I'll try again tonight, since there may be too many programs running on my computer now

Mert, As a rule, I always run a 1.5x drizzle when stacking in AS2, however, for your data, there are a lot of vertical/horizontal lines, so I restacked, no drizzle. Without color its hard to see much difference. If I can get the color working, I'll repost.

Regards,
Steve

Attached Files



#25 MvZ

MvZ

    Surveyor 1

  • *****
  • Posts: 1579
  • Joined: 03 Apr 2007
  • Loc: The Netherlands

Posted 15 January 2013 - 12:55 PM

Let me know if it gets stuck again and please send me a screenshot as well, it would be nice if I can replicate the problem so I can fix it!






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics