A try to Sh2-140

[Flavio]

Hello PI Group,
after more than 1 year of inactivity due to personal issues, I am posting again - this time a recent try to Sh2-140.
I found this subject particularly difficult and I really feel that the 7 x 1800s exposures I took with the Megrez90 at f/6.2 + SBIG ST8300 are barely enough to obtain a decent (?) image. H-alpha filter used is the Baader 7nm.

http://www.flickr.com/photos/55461835@N04/7386966636/in/photostream/lightbox/

The image has been entirely processed in PI, with histogram transform, 2 rounds of ACDNR, a moderate MT (erosion) on the stars, curves transformation and unsharp mask at the end.

Thanks for any comments and suggestions!

Flavio

PS: I will also try a color shot... sooner or later!

[mschuster]

Hi Flavio, Thanks for you image. I notice dark pixels around small stars in the darker image areas. Were these dark pixels present in your subs or did they get introduced by the processing? If processing, which process step? My preference would be not to see so many of them. In my processing I notice that StarAlighment Lanczos-3 interpolation can introduce dark pixels around small stars, and so I have been using Bicubic Spline instead to avoid them.
Mike

[Geoff]

Hi Flavio
There are some peculiar spikes around some of the brighter stars.  They look like the sort of thing you'd get from a camera lens iris, but that's not the cause here obviously. Also the dark pixels around the stars mentioned by Mike are a problem and the stars themselves are a bit "blocky"  Somewhere I remember reading that upsampling the picture can remove the blockiness, but I've never tried it.
Geoff

[Flavio]

Hello Mike,
many thanks for your attention and comments.
I used the "auto" interpolation mode during the star alignment process - I guess this resulted in the application of the Lanczos algorithm.
To check what has been going on, I made a preview of a dark area (in the upper right corner) from one of the raws and from the integrated image: I saw nothing strange with the integrated image.
Then I went again through the steps of the process: below you can see the intermediate steps:

A/ Result of the integration --> http://www.flickr.com/photos/55461835@N04/7402308702/in/photostream/lightbox/
B/ With HT and 2 rounds of ACDNR --> http://www.flickr.com/photos/55461835@N04/7402315550/in/photostream/lightbox/
C/ With addition of 2 rounds of (masked) MT -->http://www.flickr.com/photos/55461835@N04/7402321056/in/photostream/lightbox/
D/ With UM --> http://www.flickr.com/photos/55461835@N04/7402337716/in/photostream/lightbox/

I indeed begin seeing some "innatural" look on stars already after the noise reduction; where things become really bad is probably during the UM process, which is boosting all issues (including the introduction of the "blockiness" on stars).

I think I will re-start, experimenting more with the noise reduction... Have you faced already the same type of issues during processing?

Thanks again and kind regards,


Flavio

[Flavio]

Hello Geoff,
yes, I noticed the strange spikes.
Looking at the raw data and at the linear integrated image (http://www.flickr.com/photos/55461835@N04/7402513378/in/photostream) these spikes are already there, so they are not artefacts due to processing.
I took all care that the optics are well aligned in the instrument, and I am using high precision, custom made mechanical couplings. I am not sure if some kind of internal reflections can generate a pattern like this one?

Flavio

[mschuster]

Flavio, I agree with your observations. On noise reduction I do not have much experience. For my 4.2"/px undersampled subs, I find I need to be careful with noise reduction. I use an aggressive lightness mask to process only the darker areas. I use MMT on my linear integration with noise reduction applied only to the first layer, and only a minimal amount. I double check the smallest stars to make sure their brightness is reduced by only a small amount. Also I use Script>NoiseEvaluation on a preview containing a dark area. I want the post noise reduction measurement to be at most 40% less than the before measurement. My rational is that if my subs were twice the exposure length signal to noise would have been at most 40% more roughly, and I decided, rather arbitrarily, that I don't want my noise reduction to do any more than that. But all of this is an experiment of course, and I am learning what I like and what I don't.
Mike

[Geoff]

Hi Flavio
From time to time there are discussions on user groups about artifacts shown by refractors around bright stars.  See here for instance http://tech.groups.yahoo.com/group/Refractors/message/20735 for a heated discussion.
Often these artifacts consist of dark lanes intruding into the star image, but I have seen mention of spikes as well, but can't locate it at the moment.  The consensus opinion seems to be that collimation screws or lens spacers intrude ever so slightly into the optical path and cause these artifacts.  Even high end refractors can show this.  I remember seeing some images several years ago but can't locate them now unfortunately
These artifacts are never visible visually (too faint) but with increasingly better technology in ccd imaging they now show up.
Geoff