›

Advice for improvement

Matthew Humphrey
in PixInsight
I have finished watching all of the Fundamentals videos.  Throughout the learning process, I have been processing and re-processing data of the M101 galaxy - hoping for improvement with each attempt. 

One of the things that I have difficulty with is determining if an image can be improved, or if it is as good as can be given the data.  I am wondering if you have any advice for what could be improved or done differently with this image.  Overall, I am pleased with it, and it is far better than anything I could get using the tools I used before getting into PixInsight.  

In this image, I like the detail in the core and the dust lanes.  However, I feel the outer bands of the galaxy aren't great.  

The data for this image was taken in one night - 150 x 1 Minute subs, corresponding Lights, Dark and Bias taken on the same night (30 of each).

I have shared both the XISF and TIF files on my OneDrive:  https://1drv.ms/f/s!Arcl9TtTrZIxh5YWW0lPQ2O06_08RA

Thank you!
-Matt


Comments

  • BlockHeadBlockHead
    Thank you for sharing your image.
    I looked at the TIF file. My gut says this is probably pretty good considering this is only 2+ hours with very short exposures. I am assuming this is OSC data? 

    One of the things that has been suggested for me to do...is to take data like yours and process it. Then I would show my ways. Since I do not know what you did... it can be instructive. The data itself, its quality, ultimately determines the results. 

    But..it still might be good content for Horizons. Would you be interested in having me process it.
    One question though.. (might be show stopper)...did you happen to take any flats?

    -the Blockhead
  • Matthew Humphrey
    Hi,

    I would be very interested to see what you are able to accomplish with this data, and how you would go about processing it. I have processed it 5 times - starting from before I started with PixInsight (DSS and Photoshop), and several times as I progressed through the FastTrack and Fundamentals videos, trying to apply what I’d learned.  

    I see that I failed to mention that I did indeed take flats in addition to the other calibration frames, and you are correct that this is from an OSC camera (ZWO 2600MC). 

    I will upload the data to the same OneDrive folder.  Since it may take some time to upload all the files, I will reply again when it is complete. 

    Thank you for your time and help,
    -Matt




  • Matthew Humphrey
    Hi - Just to let you know all files have been uploaded:


    Thanks!
    -Matt
  • BlockHeadBlockHead
    Thanks Matt.
    Can you please describe your workflow in detail for this data?
    Thanks,
    -the Blockhead
  • BlockHeadBlockHead
    Hi Matt,

    I looked at some of your files.
    I am a bit confused...so I need to follow up.
    It says you binned the data 2x2... but this is a color camera?
    Please let me know..
    Thanks,
    -the Blockhead
  • Matthew Humphrey
    Hi Adam,

    Here are the details of how I acquired the data and then processed it (it's been a few months since I took the pictures, so this is to the best of my memory).  This was taken in urban skies, Bortle 7.

    Equipment:
    • Celestron C11 (Non-EdgeHD), and a Starizona .7 Reducer/Flattener - Focal Length ~ 2060
    • Off-Axis Guider
    • ZWO AsiAir Plus for capturing data and Auto-Guiding
    • ZWO ASI 2600MC Color Camera
    • Optolong L-Pro Filter used for all images.

    Acquisition:
    Binning 2x2  (I chose this using the calculator on Astronomy.tools, given the FL and pixel size of my equipment)
    • 150 x 1 Minute Lights
    • 30 X 1 Minute Darks
    • 30 X Bias
    • 30 X Flats  **(Taken by putting a white acrylic diffuser in front of a telescope, and shining a light in front)
    The calibration frames were taken immediately after the Lights.

    Processing:
    1. WBPP
      • Rejection Algorithm = Auto for all
      • Weighting = PSF Signal Weights
      • Weighting, Registration, AstroSolution, Local Normalization (non-interactive), Integration w/AutoCrop were all used
      • CFA Images
      • Debayer method = VNG
      • Combine RGB
      • No Drizzle
    2. DBE
    3. SPCC
    4. BlurXTerminator
    5. HistogramTransformation
    6. StarXTerminstor
    7. NoiseXTerminator on "Stars" image
    8. Combine Stars and Starless (PixelMath)
    9. CurvesTransformation (Increase Saturation)
    10. NoiseXTerminator
    11. HDRMT
    12. CreateApply mask from GAME and Luminosity to exclude Galaxy
    13. CurvesTransformation w/Mask - (Reduce Liminosity)
    14. NoiseXTerminator w/Mask

    I hope this helps.  It's possible I've missed some of the details as it's been a while since I did this.  For the most part though, this is how it went.

    Thanks,
    -Matt
  • BlockHeadBlockHead
    OK. 
    I believe the 2x2 on-chip (acquisition) binning was an error. 
    It really does give you suboptimal results for an OSC (color filter array/bayer pattern). I am happy to be corrected on this... but a CFA is just not binned in general. You end up mixing pixels. Even if you are oversampled... it muddies things up. 

    So... I think I am not going to move forward on this on grounds there is a prior issue. 

    Regarding step #11- I am surprised you had to use HDRMT. I hope you only applied it to the nucleus and not the entire image. (Which I think is what you said in #12). Also... concerning no drizzle- assuming you are not binning, you will want to use CFA drizzle at 1x1 to improve the color calibration (see the SPCC documentation for more on this).

    -the Blockhead

    P.S. I guess I would be willing to process this as an example... but I will probably strongly highlight the binning choice. :)
  • Matthew Humphrey
    Interesting on the binning.  I had been told (and read) that because of the long focal length and the size of the pixel I should bin to reduce oversampling.  But I did not understand that is for mono cameras.  I guess I just assumed it is for all cameras.  It makes sense though, that the CFA would impact binning.  I will research this, but I am inclined to take your advice on the topic.

    As for the HDRMT - honestly, I was having trouble remembering if I did this step with or without the mask, but in the History it does not specify a mask so I think I did not use one here. 

    Obviously it is up to you if you think there is value in processing it.  I hadn't expected that when I posted the original message.  Mostly a general "this is what's good, this is what's bad and here's what you could try" scenario.  Which I appreciate can be difficult (or impossible) given how many steps and variables are involved.   

    Regardless, I appreciate all of your time and help.



  • BlockHeadBlockHead
    edited September 2023
    Well.. we have identified a few things here. 
    Let me think about it.
    At the moment, another thing came up. Russ Croman enhanced his SXT software at my urging (with regards to batch processing for comets). So...I need to make a video on this ASAP.
    -the Blockhead
Sign In or Register to comment.