Skip to main content

Stacking in Groups

WGR
WGR's picture
Offline
Joined: 2010-08-17

Hello All

Helmar's comments about keeping up with the data on Nova Del 2013, got me thinking.  I hate to average the frames, because its a PITA to pull into Excel, and then manipulate, only to be rejected by WebObs for some format thing.  But there is a better way, that I just tried.  Long time AAVSO member Jim Jones has written a VBS script that stacks images in groups of x, where you just input x in a GUI.  Its available on the Maxim page, as an "extra" at no charge.  It does work only with Maxim as far as I know. JJ chime in here.  Its fantastic.


I just took 386 images from last night, and boiled it down to 38 stacks of 10.  I chose 10 at my option.  Then redid the PT on the stacks.  Saved as AAVSO extended format.  Then I retracted my 386 observations and replaced them with the 38 from the stacks.  Wah Lah.  All done.  It took about 1 minute for "StackFitsFrames" to produce the stacks.  So its painless.  Thank you Jim. 

I usually am able to reduce the data from a long night by doing it while the camera is warming to room temp, and I am shutting down the equipment.  This includes submitting via webobs. 


I dont know if this would work, but if AAVSOnet observers want to shorten their download times to VPHOT and the Cloud, try stacking your images in Jim Jones' "StackFitsImage" ap first, then download only the stacks to the cloud.  You can choose the number of images in the stack. You would require Maxim locally to work with JJ's ap, for the stacking in groups.

 

Gary

WGR

Resampling
sric
sric's picture
Offline
Joined: 2010-08-03

Thanks, Gary.

A potential problem with StackFitsImages.vbs is the script utilizes "Auto-Star matching and average combine".  You cannot select "Manual 1 star - shift only" with this script.  The resulting stars will have been resampled, which is not ideal for photometry.

Richard Sabo
SRIC

Resampling - more
sric
sric's picture
Offline
Joined: 2010-08-03

I'd like to correct the previous post.  The problem is the way the images are aligned during the resampling.  Ideally, the images should be shifted in x and y rather than "sheared".  Shifting preserves the original ADU count in each pixel.

Richard Sabo

Thankyou Gary
jji
jji's picture
Offline
Joined: 2010-07-24

Gary. 

Thanks for the kind words.  And yes the script does require Maxim.

If I had a dollar for every image I have stacked using that script I would have a new 17" PlaneWave in my new observatory. :)

Richard, If you don't like the method the script uses just open it up in Notepad and change it. But I don't think you will see any difference in the photometry between the two methods.  But you will see a lot of failures in the 1 star method.

Jim Jones

 

 

Thanks Jim
sric
sric's picture
Offline
Joined: 2010-08-03

Thanks, Jim for the comments.

 

I understand that changing the alignment method to "shift" would probably be insignificant.

However, if I wanted to try the "1 star - shift" method how would the parameters of "combineImages" command in the Save_Stack subroutine change?

 

Currently they are:

 

           Img(ii).CombineImages 1, False, 1, False

 

I'd like to try the different methods and see if there is a detectable difference.

Richard

Go to MaxIm
jji
jji's picture
Offline
Joined: 2010-07-24

Go to MaxIm Help>Contents>Scripting>Document Methods>CombineImages and all will be revealed.

Good luck

Jim Jones

Stacking
WGR
WGR's picture
Offline
Joined: 2010-08-17

Hello Richard and all;

I agree with your initial comment.  What I did was I took my raw images, flat fielded them, and then alligned them.  I save these in a folder called "bdffa" which stands for bias, dark, flat field, aligned.  I always have done this, as it saves a lot of time when I redo the photometry.

This "bdffa" folder is the one I drag over to Jim's StackFitsImages script. 

 

Gary

Resampling - more
Richard Wagner
Richard Wagner's picture
Offline
Joined: 2011-11-08

Using the one-star shift only alignment in Maxim will still result in resampling of pixel data - it doesn't shift only by integral pixels.

Rick Wagner

WCR

Thanks, Rick. I didn't know
sric
sric's picture
Offline
Joined: 2010-08-03

Thanks, Rick.
I didn't know that.


Do you know if MiraPro preserves the original pixel information  using
Shift Transformation and Nearest Neighbor Resampling Method?


Richard Sabo

SRIC

I don't know MiraPro, but if
Richard Wagner
Richard Wagner's picture
Offline
Joined: 2011-11-08

I don't know MiraPro, but if the name of the process includes 'resampling' I would suspect it too does more than a strict pixel by pixel translation.

Rick

Stacking
WGR
WGR's picture
Offline
Joined: 2010-08-17

It looks to me that if one uses the "One Star Alignment Method" and checks "Use Centroid, that you get resampling--as it aligns to sub pixel resolution.  I always thought that this was not true, and the method only shifts in integral pixels.

I looked up on the help file, and I do think that resampling does happen.

However, I wonder if a way around this is to not check "use Centroid" when designating the alignment points.  This means that one has to be a little more careful with the selection.  I plan to post a Support Ticket for this and get the true answer.  

Gary

Mira Pro/Re-sampling.
WGR
WGR's picture
Offline
Joined: 2010-08-17

I don't know Mira Pro either, but I do know Mike Newberry, and he is the one who told me about the resampling issue.  Since he is the author of Mira, I am pretty sure that there is a way to just shift in integral pixels to align in Mira.

 

Gary

Mira-Pro Image Registration
WBY
WBY's picture
Offline
Joined: 2010-07-24

Mira allows you to chose how you register images. See attached. I also thought that Maxim allowed you to select whole pixel moves only. I haven't checked recently, however, and my memory may be faulty. 

To see how you make the selection in Mira, see attached. 

Brad Walter

Maxim, Uncheck Use Centroid
WBY
WBY's picture
Offline
Joined: 2010-07-24

Gary, 

I took a look through the Maxim help manual and searched my somewhat foggy memory and I am reasonably sure you are correct. It must be 1 star manual alignment and "Use centroid" must not be checked. Then Maxim just shifts so that the pixels on which you click are shifted only to sit on top of each other, i.e.  at the same row and column index. I know that is the procedure I used when I was doing photometry with Maxim. 

 

Brad Walter

AAVSO 49 Bay State Rd. Cambridge, MA 02138 aavso@aavso.org 617-354-0484