M51--ala Wimv

[Rodd]

I have been reprocessing my 51 data for years...literally.  I just am never satisfied with it.  As soon as I saw Wims recent M51 I knew instantly what I needed to do.  Its always been a matter of balance.  I knew the data was good, but I just could not achieve a result that I could live with.  That is why viewing images from excellent astrophotographers can really improve ones processing.  Thanks Wim--Its not perfect (yet), but its back on the rails.  I can now  inch it toward the inflection point (the point beyond which the image begins to degrade.  Its like climbing a mountain and I had lost my way--now I am on the path.  One of my main problems is my screens are fickle.  I can't trust my processing screen, so its back and forth from my phone to the processing screen to other screens  for tweaks.  This has not yet been tweaked on my phone (I find background, sharpness and color can be tweaked effectively on the iphone).  There is still a saturation issue, the Ha is weak and of course the background is like driving down a bumpy road in a low riding car with no shocks (that's my sky and there are limits to what I can do).  Most recently, Olly opined that the image was clipped, which was obvious to me after he said it as the faint extensions were truncated sharply.  They may not be as bright and extensive as they should be for 24 hours of data--but they are not clipped.  Previously I just had to use deconvolution to make the image as sharp and "good" as I could.  Well--it never worked for this data and I tired of applying first aide to my forehead (and having to repaint the door).  I smartened up and forwent the deconvolution.  

TOA 130 and ASI 1600-HaLRGB - about 24 hours

 

 

 

 

Edited May 7, 2022 by Rodd

[almcl]

1 hour ago, Rodd said:

  That is why viewing images from excellent astrophotographers can really improve ones processing. 

Indeed so, Rodd.  May I say that I also find your images inspiring and they prompt me to see if I can do better.  Although I seldom manage to achieve as good results, I much enjoy looking at what you (and others) achieve.

[Sunshine]

A beautiful image, love the contrast!

[wimvb]

Now you make me blush. 😊

Great image, Rodd. And amazing that you get this crisp detail without any deconvolution.

[Rodd]

5 hours ago, wimvb said:

Now you make me blush. 😊

Great image, Rodd. And amazing that you get this crisp detail without any deconvolution.

Thanks Wim--I find deconvolution to be a very slippery tool--like trying to eat soup with a fork.  Sometimes it works wonders, and sometimes it looks terrible.  I revisited this image and reprocessed it again in efforts to get deconvolution to work--but no matter what settings I used, and no matter how careful I was in choosing my stars with the Dynamic PSF tool--the results were always the same - bright knots and lines of over sharpening/brightness.  I find the star mask the most difficult aspect of the deconvolution process, because invariably there are tiny stars and bright spots that are way below the star mask tool to identify, and they pop right up after decon.  But sometimes it is like a miracle.

I decided the Ha regions needed a bit of a boost, and I cropped the image to provide a bit larger view without having to go to full res.  Stars got a saturation boost as well.  There is no star control in this image--perhaps I should use a iteration of morphological transformation--not sure.  That will be a minor adjustment if I decide to try.

[wimvb]

15 minutes ago, Rodd said:

Sometimes it works wonders, and sometimes it looks terrible. 

For deconvolution to work, you need an image with strong signal and low noise. I suspect that I live in an area with less light pollution than you, because I seldom need to use noise reduction on the L master. I also never use deconvolution on an rgb image, but only on L.  If I only have rgb data, I either extract luminance from that, or I combine the colour masters to create a synthetic L to work on.

[Rodd]

1 minute ago, wimvb said:

For deconvolution to work, you need an image with strong signal and low noise. I suspect that I live in an area with less light pollution than you, because I seldom need to use noise reduction on the L master. I also never use deconvolution on an rgb image, but only on L.  If I only have rgb data, I either extract luminance from that, or I combine the colour masters to create a synthetic L to work on.

I don't decon RGB either--sometimes when I try decon on a SL, I run into trouble because the stars are not equal.  So the dynamic PSF tool yields a PSF image that does not work all that well.  Anyway--I usually find my lum is much better than my SL.  This is probably due to my RGB channels can vary wildly in FWHM due to seeing and high cloud.  I don't think I have ever had a project where the RGB and L were all really good.   I think my M100 image was about the best consistently good conditions.  I think the linear RGB image had a FWHM of 2.1. I have subs as low as 1.8, but if I only used sub s with FWHM less than 2 I would need to image for many days per channel.

[wimvb]

23 minutes ago, Rodd said:

This is probably due to my RGB channels can vary wildly in FWHM due to seeing and high cloud

How do you plan your sequence? Because the weather can be somewhat unpredictable here (clear but high cloud or just plain wool blanket clouds), I always try to collect all three channels per night, and I usually have a repeating sequence of 10 x 4 minutes exposures per channel plus 15 x 3 minutes for L. On good nights with low fwhm, I only collect L. I stretch the RGB image much less than the L image, usually only Arcsinh stretch followed by curves to lift the lower parts. I try to keep the curve linear above the 0.5 point, so as to not overstretch the stars. I also have the advantage of a reflector, with virtually no chromatic abberation. (I still have some difference in fwhm between channels, but I suspect this to be from the atmosphere.)

34 minutes ago, Rodd said:

So the dynamic PSF tool yields a PSF image that does not work all that well.

Happens to me too sometimes. I weed the stars in the psf process with a blunt tool. I only keep stars that have a Moffat profile (excluding Moffat4 and above, as well as Gaussian profiles), Amplitude between 0.1 and 0.9, and a Mean Absolute Difference that does not vary too much.

[Rodd]

4 hours ago, wimvb said:

How do you plan your sequence? Because the weather can be somewhat unpredictable here (clear but high cloud or just plain wool blanket clouds), I always try to collect all three channels per night, and I usually have a repeating sequence of 10 x 4 minutes exposures per channel plus 15 x 3 minutes for L. On good nights with low fwhm, I only collect L. I stretch the RGB image much less than the L image, usually only Arcsinh stretch followed by curves to lift the lower parts. I try to keep the curve linear above the 0.5 point, so as to not overstretch the stars. I also have the advantage of a reflector, with virtually no chromatic abberation. (I still have some difference in fwhm between channels, but I suspect this to be from the atmosphere.)

Happens to me too sometimes. I weed the stars in the psf process with a blunt tool. I only keep stars that have a Moffat profile (excluding Moffat4 and above, as well as Gaussian profiles), Amplitude between 0.1 and 0.9, and a Mean Absolute Difference that does not vary too much.

It’s hard for me to take more than one filter per night as I have to shoot flats before any filter change and before a meridian flip due to my camera and filter wheel being a PITA. I tried it one night and the focus offsets had to be set and I am not sure it would be automatic.  I do see less difference in fwhm between filters with the C11Edge. That scope has both mirrors and lenses. But still 

[wimvb]

3 hours ago, Rodd said:

due to my camera and filter wheel being a PITA

I remember you writing about problems with the ZWO filter wheel some time ago. So, you never got that resolved? That's too bad, life's too short for that kind of problems.

[Rodd]

4 hours ago, wimvb said:

I remember you writing about problems with the ZWO filter wheel some time ago. So, you never got that resolved? That's too bad, life's too short for that kind of problems.

It’s even worse now.  My laptop died and  the zwo ascom driver does not work with windows 11.  I have to use ASI studios software to image. It’s not bad, really. In some ways I like if better than maxim dl. But I can’t dither.  It does not support it.  I can’t win.  I want to upgrade to the 2600 anyway.  But I am considering Altair.