Re: [APML] Veil Complex Reprocessed

["Bradley" <bhope@thorngarden.net>]

Hi, Alan;

----- Original Message ----- 
From: "Alan Voetsch" <alanv12952@yahoo.com>
To: "Discussion of Film Astrophotography" <astro-photo@seds.org>
Sent: Friday, September 16, 2005 11:36 AM
Subject: Re: [APML] Veil Complex Reprocessed


> Due to the slowness of the image downloads, I was not really able to
> compare the two. I wouldn't have been much help anyway. 8-)

I've compressed them so they should download faster.

Original:
http://astronomy.thorngarden.net/images/veil-orig.jpg

Latest:
http://astronomy.thorngarden.net/images/veil-latest.jpg

> I do have a question concerning the procedures you describe below.
> Could you explain the how and why of:
>
> 1. Offset scanning

I just shift the scan box a tiny bit in both directions for the second scan
of the same frame. If there are any irregularities in the scanning process
itself that are attributable to the scanner, this should make for two
slightly different images with the irregularities offset a bit from each
other. Then, a median combine in Registax, after aligning and correcting, of
those two scans. I don't have the means of doing multiple scans and
combining them in the scanner software (unless there's something I'm missing
in PolaColor). Does it really help all that much? I don't know, but I
remembered others talking about doing multiple scans of the same frame to
correct for scanner irregularities...so I thought I'd give it a try, doing
it the hard way. I assume that doing additional scans would be even more
beneficial, but it would take forever and a day to do it my way...so I
settled for doing two scans of each frame.

> 3. What is the difference between median combine and adaptive add, and why
do both?

I'm not an expert at this, so I'm going to paste Mile Unsold's explanation
of median combine from the APML archives.

------------------------------------------------------------------------
1) From each image in the set select the same pixel with respect to its x
and y  coordinate and build a stack of pixels. For example, choose the top
left pixel from each image in the set and form a subset of pixel values.
This is called a pixel stack.

2) Using this subset of pixel values calculate their median value. This is
done by sorting all pixel values in the subset by increasing value then
assign the middle value in the sort order as the median value.

3) Set the corresponding pixel value in the combination image equal to the
median calcuated in step 2. So in the example, set the top left pixel of the
combination image equal to calculated median from step 2.

Now repeat the above process accross each row of the images on a
pixel-by-pixel basis.

So the top left pixel of a median combination may have come from image #1,
the next pixel to the right from image #2,  and the 2nd pixel to the right
may have come from image 1# again.    It all depends on the sorted order of
tthe pixel stack not the image stack.

Mike
---------------------------------------------------------------------------

The end result is to, hopefully, give me two nice smooth scans, without
scanner artifacts, to work with for additional processing...though after
reviewing the above, it would seem that I might consider doing at least
three scans of each frame to median combine.

And from the  Images Plus help section on Adaptive Addition.

------------------------------------------------------------------
Several short exposure images can be added together to create a single long
exposure image using this option. Adaptive addition adds lower intensity
values from a set of images without overflowing the bright areas of the
combination image. If necessary, the bright areas of the combined image are
scaled to minimize overexposed areas.
------------------------------------------------------------------

Since film doesn't react the same way as CCD or CMOS, I doubt I'm getting
all the potential benefits from adaptive addition, but...I think it did help
bring up the nebulosity a little, and I also think I ended up with a little
'crisper' image (though that could have been as a result of my handling the
image differently while processing it in PS..dunno).


Bradley

Thorn Garden Observatory
http://astronomy.thorngarden.net

_______________________________________________
Astro-Photo mailing list
Astro-Photo@seds.org
http://seds.org/mailman/listinfo/astro-photo