›

Dust Lanes

Bruce Donzanti
edited February 2022 in PixInsight
On the image attached, as an example, is there a certain/best/preferred approach to pull out the dust lanes further?  I am under Bortle 7 skies if that makes any difference on the approach utilized.

Comments

  • BlockHeadBlockHead
    Hmmm... I think the approach is to put the question into the language of image processing. "Pull Out" is sufficiently ambiguous and you will not find a "pull out" process in PI.

    So your choices are make brighter, make darker, increase contrast (both darker and brighter) and edge enhancements (sharper).

    Brighter means lower the white point. 
    Darker means raise the black level. 
    Contrast means do both of the above.
    MMT, Unsharp mask, LocalHistogramEqualization means increase contrast at edges.

    (there are also things like non-linear stretches, HDRMT..etc that decrease contrast)

    So... given the above... what do you want to do?, and then the array of processes and techniques are easier to consider.

    -the Blockhead
  • Bruce Donzanti
    Adam- why do you always answer my questions with a question?.....LOL, just kidding!!

    OK- this is interesting as I utilized every one of those techniques in this image.  For example, I used HDRMT to lessen the bright core and LHE and unsharp mask to different points in the workflow to increase edges.  I suspect that perhaps I either used too much or too little of some of them or perhaps in the wrong order.

    So, based on your question and what I'd like to do for this particular image is to increase the contrast and contrast edges (I assume both can be done).  I often have the problem, as stated above, of overdoing or under-doing a process or tool.  

    Bruce  


  • BlockHeadBlockHead
    Ok... now we are getting somewhere.
    I can't speak to the degree of contrast/sharpness/saturation that you find pleasing. 
    For me, this image looks pretty close to complete and self-consistent. I perhaps would have tried (if you did not do it) the color corrected HDRMT method if there is color information near the core. It does look like this might have gotten saturated in your original data based on the flatness I see there.

    With the image in its current form- I do not think a global application of contrast enhancements will help. Your image is already at a critical black level. You need a brighter image (with a brighter background) to do more. If you apply more contrast anything...some faint things will start to disappear into the darkness. Things in the nebula may be OK... but not the dust- say between M42 nad NGC 1977- which is what I think you are talking about when you say you want to pull out the dust.

    You might want to point at images that show something like you are expecting. I am going to predict they have a brighter background... but will wait to see.

    -the Blockhead


  • Bruce Donzanti
    Thanks so much for the feedback and suggestions. 
     
    First- I was not even aware of the color corrected HDRMT method- I just used the default settings on the tool

    Second- Going back and looking at the RGB and luminance linear images, I think the core is overblown So, perhaps I should have used shorter exposure times.

    Third- you are correct in that I am talking about the dust lanes between M42 and NGC1977.  Attached is an example of an image I found on the internet which shows dust lanes as I would have expected to see. 
    tumblr_pobgzwnu7W1u7ypw6_1280.jpg
    1080 x 775 - 196K
  • BlockHeadBlockHead
    Well..that's not fair! That is a different object! 

    I think you fundamentally need to start with a brighter background. However, you can only get away with this if you have enough data so that even the background is not too noisy. Then you apply the contrast enhancements you want..and finally choose the black level (darken things). This is usually the best lever-power-sequence for the effect.

    If you look at my image (with a smaller telescope... ) you will see how bright I made the sky (and left it that way):


    Click on the large version of the Horizontal link... 

    specifically this one:

    -the Blockhead
  • Bruce Donzanti
    Sorry- LOL!

    ok- got it- thanks again for the advice.  I definitely need to collect more hours of data as I know my background would be too noisy to do this.


Sign In or Register to comment.