Image of the day

Captured by
Michael Scherman

M17 Swan-Omega Celestron C6 Hyperstar2

My Account

New to Astromart?

Register an account...

Need Help?

To Align or not to Align

Started by Halleys_Trauma, 06/23/2007 08:12PM
Posted 06/23/2007 08:12PM | Edited 06/23/2007 08:13PM Opening Post
I have been starting to experiment with "NOT" aligning images I take in the same evening and believe that the finished product is better

I would love to hear some other thoughts on this. smile

My thinking is if i did not have an issue with guiding why bother and I confirm this with the align function in MaxIm using the overlay radio button. I thought perhaps field rotation might creep in but my last session went till 5A with no issue.

My Optima D43M which is now almost 2 years old ran my entire rig all night. What a great investment that has turned out to be.

smile
Posted 06/23/2007 08:50PM #1
Congratulations on a rig that runs so perfectly!

If your pictures are already aligned, then there is no need to align, I should think.

I do find it hard to believe there is such an mount out there, but since you have one, I guess I'll just have to agree with you.

Alex
Posted 06/25/2007 01:14PM #2
Hi John,

One possible counterpoint against using perfectly guided, unaligned subframes is that you lose the benefits of dithering.

When guiding parameters are set correctly, it is possible for a mount to track so well that image alignment is rendered pretty much unnecessary. There will be one unfortunate side-effect of this:

All CCD's, regardless of their class, will have some misbehaving pixels. These pixels will always run hot or cold, and so pepper your image with unwanted noise. If your pixels run excessively hot or cold, dark reduction will only lead to either always dark or bright spots (I have such a group of excessively hot pixels in my ST-8). Aligning and stacking images with these misbehaved pixels will only lead to enhancing these bad pixels.

The answer to dealing with these bad pixels is to dither. This is where the software actually instructs the mount to shift the image around a few pixels (1-3 pixels usually) in a random sequence. In post-processing, the deliberately-misaligned subs are realigned, then stacked using a statistical reject algorithm (Median or better yet, sigma reject). The end result of all this is to statistically remove the bad pixels and lead to a smoother image.

HTH,
Eric