›

WBPP registration and Local Normalization fail

John Hughes
in PixInsight
Hello
 I Have been trying to use WBPP with limited success. Usually the problem is star alignment and Local normalization..
For example I have 233 mono photos of M15 taken with a ZWO ASI 1600MM pro mounted on a Paramount MYT and a Celestron 9.25 edge telescope. The photos were calibrated with  master dark and a Super Bias and 20 flats each for LRGB. The photos calibrate fine and went through Cosmetic Correction but at registration 228 photos were rejected. Out of 53 Lum photos four were selected to make into a master. For the Red filter one photo was accepted everything else was rejected. This leaves me to go through the drawn out process of dynamic alignment for 228 photos which takes about 15 or 16 hours otherwise I have no finished image to show for two nights of imaging. After dynamic alignment the aligned photos integrate into masters and and combine into a final image pretty much without issue. Can anyone suggest a quicker solution to this issue?

Comments

  • BlockHeadBlockHead
    It does not sound to me like you are investigating the source of the issue. 
    In what way are the images not registered? 
    Is it not detecting stars? 
    How do the images and values compare?
    Is there a calibration issue (zeros/oversubtraction)?
    What about the reference frame? Sometimes WBPP will pick a wrong reference frame because PSF signal weight can be "tricked"? 

    So many questions... 
    Usually what you have to do is show all of things it cannot be... and this will leave only a few things it might be.

    If you upload your data ... the individual files... (don't give me masters) to a server- I can take a look. 
    You will need to upload 10 frames of each calibration type. And 10 frames of both known "good" frames and frame that fail for some reason. 

    Typically these issues are of a data quality one. Some adjustments in star detection can help... but without looking at the data in detail I cannot say.

    If you make the files available remember to set the permission to "anyone that has the link" type of thing.
    -the Blockhead
     
    P.S. I am not a fan of SuperBias. 

  • John Hughes
    sorry but I am not very computer literate. How do I  upload the files to you? I have dropbox but have not had a lot of luck using it.
  • BlockHeadBlockHead
    Hi John,

    Each service (Google Drive, OneDrive...DropBox) has its own ways. They are similar. 
    I have not had great experiences with DropBox either..but this is definitely one way to go if you do not have Gmail (and therefore Google Drive). There isn't a way for me to accept the files... (I would need to own my own server... maybe someday soon I will. You have to put them somewhere I can get them. 

    There are YouTube videos that will explain of course... :)
    -the Blockhead
  • John Hughes
    Thank You  I will work on hooking up a server later today. I do have a Gmail address and I have Google so maybe that is the way to go. Last night I ran another WBPP and chose my own reference frame based on FWHM. .Apparently it was poor choice as only 2 Lum images were accepted and No Reds. It is frustrating because last week I took photos of M 80 and although there was a high Star Alignment rejection rate I did end up using Dynamic alignment and finally got an image out of it.
    Again the Calibration and Cosmetic Correction work fine but then it seems to go off the rails. Under measurement 242 are measured with 13 failed then under Bad Frames it shows 209 rejected. Under Registration the Lum images show 2 registered and 10 failed. For the RGB frames all of them failed. Under local normalization  I did get 1 Lum accepted and 1 failed. The rest all failed as did the Image Integration of course. I have been using Pixinsight for about a year and I subscribed to your Pixinsight Fundamentals. I have produced what I think, at my level at least, are OK images. But the struggle to get Star Alignment to work is consistently a problem and I end up going to Dynamic Alignment and tediously aligning dozens and now hundreds of images
  • John Hughes
    Hello Adam
    I uploaded the images you said you needed. As far as the registered ones go there were only 4 Lum images and 1 Red image that WBPP accepted. The rest were all rejected. I put them into folders that are labeled. The link is posted below:

    https://drive.google.com/drive/folders/1pfjiRddAnaHkqkMhJBX9gs77R_BQN971?usp=drive_link

    Thank You
  • BlockHeadBlockHead
    edited July 2023
    HI John,

    Don't forget my comment above.. you need to set permission to the folder (or files) to be "Anyone that has the link." Currently that is not true.

    -the Blockhead
  • John Hughes
    Sorry, you did tell me that and I messed it up. I think I have it right now.
  • John Hughes
    I hope this is right

    https://drive.google.com/drive/folders/1pfjiRd    dAnaHkqkMhJBX9gs77R_BQN971?usp=sharing
  • BlockHeadBlockHead


    The blockhead
  • BlockHeadBlockHead
    John,

    I was about to download.. but first a question.
    Why are you using LPS?
    Do you really have an issue that requires this??

    -the Blockhead
  • BlockHeadBlockHead
    Hi John,

    OK...I can see part of the problem. I looked at your images. Unfortunately in the world of astronomical image processing there is a sense of data quality. There is a real concrete (mathematical) description of stars in PixInsight that means if your data do not conform to the definition... there will be problems.

    That is what is happening here. Your stars are aberrated significantly. I can't say exactly... but from my experience I would say your telescope is way out of collimation. In addition your stars are not in focus (which could be due to the collimation). PixInisight will simply not bend in such a way to automatically think they are stars. 

    On a note that might come to your mind- some programs are lenient to such a degree they might accept these as detected sources and still register the images. I do not think that is really the issue here. There are two parts of the hobby. There is acquisition and then processing. I think this is an acquisition issue. I think it would be prudent to address the optical issues first and then I think you will find the subsequent processing is much easier. 

    Now... all that being said, it is possible to force PixInsight to detect these as stars? Maybe. I would need to fiddle... but again, I am not convinced that is really the best approach.

    Let me know what you think. If you strongly think this is a reasonable thing based on this data.... I will give it a go. However, you will need to understand it isn't a general solution. Just a solution that is "just-so" for this data.

    -the Blockhead
  • John Hughes
    Hello Adam
    Yeas I have been trying to work with the collimination. The adjustment screws have some tension on them which made me a little nervous reaching around the telescope and adjusting them. I have purchased some Bob's knobs. Unfortunately the weather here on the east coast is atrocious for the next week so it will be a while until I can get things adjusted.
    Thank You for your input. The technology is fabulous when I can get it to work but is maddening when trying stumble through it on your own.
Sign In or Register to comment.