Astronomy, Photography and Weather
October 23, 2019, 07:24:14 pm
Welcome, Guest. Please login or register.

Login with username, password and session length
News: IAA lecture programme continues alternate Wednesdays from September - an excellent programme of lectures- Queens University Belfast - Bell Lecture Theatre. Also keep an eye out for the Summer Events
 
  Home Help Search Gallery Staff List Login Register  

Stacking alignment in AIP4Win

Pages: [1]
  Print  
Author Topic: Stacking alignment in AIP4Win  (Read 748 times)
Steaphany
Newbie
*
Posts: 5


View Profile
« on: June 13, 2009, 05:08:47 pm »

I have been working to come up to speed using my Sigma SD14 for astrophotography and learning how to use AIP4Win.

I big hurdle was over come when a friend from a Sigma User Forum told me of a tool which can extract raw image data from the various structures of the SD14's X3F file. With this, I have been able to extract the RAW image data and perform a transformation matrix to scale the brightness levels of a set of 24 images that I shot back on November 11th 2008 of Cassiopeia storing the data in 48 bit per pixel TIFF files. The exposure was 16 seconds at an f1.8, ISO 100, with a 28mm focal length. Since these were shot on a static tripod, I knew that alignment from frame to frame would be critical to achieving a clean image.

I went through stacking these images in AIP4Win, using the manual two point alignment and choosing stars located in the Upper Left and Lower Right regions of the reference frame. As time passed from frame to frame, the star positions drifted across the field of view and, obviously, each frame needed a slight adjustment, pointing out the revised locations of the alignment reference stars.

All went well and the process yielded a nice image except that only the stars in the regions surrounding the alignment stars are focused into nice points. As the distance from either alignment star increases, the individual stars end up drawn into ever more pounced star trails, even in the region along the diagonal between the two alignment stars.

AIP4Win did a wonderful job except for the uneven alignment.

What went wrong and how can I perform the stacking so that the alignment yields nice point like stars ?

If anyone needs, I can post a URL to a zip artchive containing the files.

Thanks in advance
Report Spam   Logged

brianb
Hero Member
*****
Posts: 1228



View Profile
« Reply #1 on: June 13, 2009, 05:36:59 pm »

Quote
All went well and the process yielded a nice image except that only the stars in the regions surrounding the alignment stars are focused into nice points. As the distance from either alignment star increases, the individual stars end up drawn into ever more pounced star trails, even in the region along the diagonal between the two alignment stars.
Field rotation? Try Deep Sky Stacker....
Report Spam   Logged
Big Dipper
Events
Hero Member
*
Posts: 1247


Oxford, UK


View Profile WWW
« Reply #2 on: June 14, 2009, 09:34:28 am »

Yes seconded Deep Sky Stacker - works very well for me.
Report Spam   Logged

Remember:- If all else fails, read the Instruction Manual! Grin
 


Andy
Steaphany
Newbie
*
Posts: 5


View Profile
« Reply #3 on: June 14, 2009, 07:34:57 pm »

I have and been using DeepSkyStacker far longer than I have been playing with AIP4Win. I just wanted to see how well AIP4Win performed and my inquiry was prompted from being some what new to AIP4Win.

My bottom line conclusion -

DeepSkyStacker is great, both Free and it works !!!

AIP4Win may have it's place, but to stack images, it sucks !!!

Here is my proof:

DeepSkyStacker yielded


AIP4Win yielded:


A note of distinction between the DeepSkyStacker and AIP4Win images, both needed processing in Photoshop to allow them to be presented here. Since the Milky Way passes through Cassiopeia, I toned down the Milky Way's glow to better bring out the stars in the DeepSkyStacker image. I did not feel going to such efforts would have improved the AIP4Win image.

And here are the stars I had to manually identify for every frame to AIP4Win:

A lot of good that did  Huh

You can see why I immediately concluded that I must have done something wrong. This is the first time that I used AIP4Win's stacking function. Only in the areas surrounding the alignment stars do the stacked stars achieve alignment. The further away from the alignment stars, the worse the distortion. Compounding this, look at the orientation of the distorted stars - the distortion varies and is inconsistent.

Now, for a final nail in the AIP4Win coffin, I contacted AIP4Win support and received a timely reply:

Quote
From:     Jim Burnell
To:    Steaphany Waelder
Subject:    Re: Stacking alignment in AIP4Win

Hello Steaphany,

What you are experiencing is the wide-field astrophotographers lament.
At low elevations,. atmospheric refraction distorts your image, such that
all areas of the image are not seen at the same scale. This is not a simple
correction that can be implemented by shifting and scaling the image.

Another problem occurs when using widefield photographic lenses.
These lenses have pincushion distortion. When images are stacked that
have been shifted relative to each other, the scale of a particular region
is not constant from image to image, as the sky moves across the field.
Again, this is a complex distortion that is not easily corrected during
processing. This is not much of an issue with a well-aligned tracking mount,
but is a real problem shooting off a static tripod.

Jim

This really put me in a bad mood, and I'm not about to reply back. Instead, I'll disect their issues here for everyone to see.

Quote
At low elevations,. atmospheric refraction distorts your image
Hmmm, I shot these images on November 19, 2008 between the times of 10:02 PM and 10:22 PM Central Standard Time, November 20th 4:02 to 422 Universal Time, when Cassiopeia was on the meridian. The location of my camera was at 98 28' 35"W 33 13' 32" N 386m. Calling this up in SkyMap Pro, Cassiopeia is about 60 above the horizon. ( A low elevation ? Huh ) Looking towards the lower left of the image, the stars of Cepheus show that the lowest altitude in the image is between 50 and 55.  ( Again, A low elevation ? Huh )

Quote
when using widefield photographic lenses.
Hmmm, I clearly state that I shot these images with a Sigma SD14 dSLR, a camera which uses an imager smaller than a standard 35mm film frame. In fact, the Sigma SD14 User's Manual, page 121 titled "Specifications" describe the "Picture Angle" as "Equivalent to 1.7x the focal length of a lens when used on a 35mm SLR camera". I also clearly state the focal length of the lens used as being 28mm, yielding 47.6mm on a 35mm SLR. ( Wide angle ? Huh ) I even calculated the combined SD14 & 28mm lens field of view at Horizontal 40.57, Vertical 27.69, and Diagonal 47.91 ( Again, Wide angle ? Huh ) The stars captured in the image are proof - Cephus can be seen in the lower left and Perseus to the upper right.

Quote
These lenses have pincushion distortion
Like Sigma, known across many camera brands for quality lenses, would allow even an out of spec or defective lens to suffer from a problem as indicated by the AIP4Win stacked image. The image looks like it was shot through a lens made of melting ice. The lens used was a Sigma 28mm f/1.8 EX DG Aspherical Macro which I purchased on October 6, 2008 from B&H Photo for $299.00. Please note that this is a DG lens, not a DC lens. I specifically selected this lens for it's ability to accurately operate over a full 35mm film frame, simply to ensure that I'd have an accurate, distortion free, image on the smaller Foveon imager area. This is not a "kit" lens and my SD14 purchase was the individual camera body with lenses purchased separately. I rather spend the money for quality components and instruments rather than getting stuck buying junk that gets sold as a "kit".


Quote
this is a complex distortion that is not easily corrected during processing.
Hmmm, b-u-t DeepSkyStacker took the very same collection of images, accurately identified from 84 to 146 stars in each of the 24 images, 19 of which were used in the stacking after receiving a quality level of 877.67 to 1024.83 and accurately aligned each, completely over the full field of view all on it's own. And remember, these frames where manually shot, leaving the frame to frame time variable, with my SD14 sitting on a static Velbon tripod. ( I couldn't make it more challenging by tieing one hand of my SD14 behind it's back since it doesn't have hands. I could have waited for the Moon to rise, but it was just too cold  Tongue )

I think I've done enough ranting for the time being.
Report Spam   Logged
JohnMurphy
Full Member
***
Posts: 220



View Profile
« Reply #4 on: June 14, 2009, 07:54:51 pm »

Interesting - thanks for the review.
I was thinking of purchasing AIP4Win but will think again given what you've said. DSS does a decent job for me at the moment, but I have had trouble with alignments using a lens at 18mm, although MS Digital image Pro (not an astronomy program) had no trouble aligning and stacking the images.
Sounds like the response you got from AIP4Win was just trying to blind you with science - looks like they picked the wrong guy to try that on. Grin
Report Spam   Logged

Clear Skies,
John Murphy
Check out my Photos
Steaphany
Newbie
*
Posts: 5


View Profile
« Reply #5 on: June 14, 2009, 09:08:36 pm »

One last point that just caught my eye:

Quote
This is not much of an issue with a well-aligned tracking mount,

Hmmm, so a well polar aligned tracking mount would some how compensate for and keep the stars in the sky from dropping below the low, distortion prone, altitude of, I don't know, 60 ?
Report Spam   Logged
brianb
Hero Member
*****
Posts: 1228



View Profile
« Reply #6 on: June 14, 2009, 09:22:37 pm »

No it wouldn't, but I think the problem is field rotation rather than distortion. If so DSS will fix it. AIP is optimized for narrow field frames taken with an equatorial mount and doesn't expect field rotation. There are some things it does really well but stacking of wide field images taken with a static camera, or a tracking altaz mount, is not one of them - DSS is a great deal better for this, is relatively easy to use and is incredibly cheap!

Quote
the low, distortion prone, altitude of, I don't know, 60 ?
Differential refraction increases quite rapidly towards the horizon but the point is that it affects different altitudes differently so the field of view will have an effect - a narrow field of view will be realtively undistorted quite close to the horizon, whereas with a really wide field of view you may have problems matching frames if some of the objects are in the 50-60 degree altitude range and others are in the 5-10 degree range.
Report Spam   Logged


Pages: [1]
  Print  
 
Jump to:  

Bookmark this site! | Upgrade This Forum
SMF For Free - Create your own Forum

Powered by SMF | SMF © 2016, Simple Machines
Privacy Policy