Wide Bubble nebula in HA

[Rudeviewer]

This my start of a narrowband Bubble nebula.  I will continue this by adding OIII and SII to this over time.  This was taken over 2 nights. the first night i got 8x1200s images, and the second night was 3x1200s images for a total of 3 hours 40 minutes.   Weather stopped me both nights from getting more.

Taken with Williams Optics GT81, Televue .8x flattener/reducer, atik 383l+ mono, baader 7nm AH 36mm filter, Celestron CGEM DX mount, guided with a st80 scope and Orion SSAG Pro. 

[Rodd]

Amazing image for the exposure time.  I took a bubble with 20 hours that falls way short of this.  What kind of processing did you do?

[Rudeviewer]

In PI I did 2 histogram stretches, a curve transformation, a small amount of local histogram equalization and then deconvolution.  Also I used the dark structure enhance script.  Transfered it to photoshop and used used Noel's acti9ns to reduce star size and reduce noise a bit.  Also did just a bit of contrast enhancement.   

[Rodd]

1 hour ago, Rudeviewer said:

In PI I did 2 histogram stretches, a curve transformation, a small amount of local histogram equalization and then deconvolution.  Also I used the dark structure enhance script.  Transfered it to photoshop and used used Noel's acti9ns to reduce star size and reduce noise a bit.  Also did just a bit of contrast enhancement.   

You don't transfer the STF to the histogram?  How do you transfer stretchs to make them permanent if you don't use STF?

[Rudeviewer]

I use the histogram transformation tool and play with the slider triangle at the bottom.  I usually stretch once and then bring the black point down to where its close.  Then i use the curves transformation to fine tune it.  

[Rudeviewer]

Is it me or does anyone think this image is flipped from its normal orientation?  I think I have ever a flip function in APT to change it but I never realized if it was flipped or not.  I have heard people say they've never seen some of the orientationsky of a few of my images.....

[wimvb]

The STF function in PI is mainly for viewing an image while still in the linear stage. Usually you don't (want to) use it as the main stretch.

To keep stars under control you can try the following:

Deconvolution in the linear stage after dbe

If necessary, a small dose of morphology transform on the stars

Masked stretch

Histogram and curve transform

Perhaps followed by the following:

Exponential stretch, gamma stretch, local histogram equalisation if deemed necessary.

MMT with small positive bias in layers 2 - 4 to sharpen slightly.

[Rudeviewer]

I will have to try deconvolution before i stretch, i always did it afterwords with a starmask in place.  Same with morph...Thanks for the heads up.

[PatrickGilliland]

On 8/15/2016 at 15:06, Rudeviewer said:

In PI I did 2 histogram stretches, a curve transformation, a small amount of local histogram equalization and then deconvolution.  Also I used the dark structure enhance script.  Transfered it to photoshop and used used Noel's acti9ns to reduce star size and reduce noise a bit.  Also did just a bit of contrast enhancement.   

Nice image FoV is great with the Claw showing up nicely.  Reading your process was interested that, if i read correctly you ran Decon after the other processes, i find that running much earlier in the process at linear state is more effective and releasing detail.  Did i understand correctly?

Nice work - Paddy

[Rudeviewer]

Thanks Paddy I appreciate the comment.

Yes you understood it correctly.  I usually do decon after i stretch and modify curves.  I will be trying doing it in the linear state now to see what the differences are.  I had no idea.  Still learning new tricks everyday in Pixinsight.  I will try to go back to this image and do it the other way around.  Perform a DBE then before stretching perform the decon and morph for the star reduction and go to the stretch and other processes.  

[wimvb]

I would suggest something like this:

(in the linear stage)

dbe

deconvolution to tighten the stars

slight morphology transform to dim the stars (do not remove them, just dim slightly), using a contour star mask

masked stretch, which stretches nebulosity more than stars

(now you're in the non linear stage)

histogram transform and/or curve transform

gamma stretch and/or exponential stretch to enhance  detail in nebulosity, maybe HDR multiscale transform (depends on target and data quality)

multiscale median transform with small positive bias values for layers 2 - 4 if you want to sharpen detail

more histogram and/or courve transform if necessary

 

This can only be a very general suggestion. Exact process steps depend on what you aim for and what data you have. For my DSLR images, I usually need noise reduction right after dbe, before doing deconvolution and morphology transform.

If you save the project rather than just the image, you can always undo steps in the future. You can also isolate steps from a workflow with the exact settings that you used on the image. This gives an immense flexibility when you test various processing steps.

I can recommend the tutorials by Kayron Mercieca (lightvortexastronomy.com) and Alejandro Tombolini (pixinsight.com.ar).

 

Good luck

[laudropb]

Very nice first attempt.