›

Local Normalization parameters (WBPP vs Process)

TxAstro
in PixInsight
Hello,

When using Local Normalization in WBPP, two subs resulted in the same error.



*** Error: Unable to compute a valid scale estimate (channel 1)



I reset WBPP to make sure I'm using the current version. I am also
using default parameters. I reran using "Single Best Frame" as the
reference, but I received the same error.



Using LN process, I increased the Scale Evaluation noise reduction
parameter from 0 to 1, and the files were normalized without problem.



LN in WBPP does not have this noise reduction parameter. Are there
other LN (WBPP) parameters which might enabled the subs to be
normalized.



Thanks,

Xavier

Comments

  • BlockHeadBlockHead
    In this situation the pipeline (WBPP) is usually trying to tell you something is likely wrong with your data. It could be a calibration problem... an oversubtraction/pedestal problem... or perhaps simply poor star quality and resorting to the multiscale method is all that is necessary.

    Doing the brute force thing and then trying to find the similar parameter I guess (without looking at the data) does not seem like a good way to proceed. In the ABS Universe... knowing why this is happening is more important. I think the data needs to be looked at closely. 

    -the Blockhead
  • TxAstro
    For whatever reason, I've thought pedestal applied to NB images.  Does it apply to any image?  These are RGB OSC without any filter.  They are also my first images with a dedicated astro OSC.  .I could simply drop the subs (2) and proceed with integration, but then I wouldn't know what is happening.  I'll revisit the calibration steps for starters.

    Thanks,
    Xavier
  • TxAstro
    Quick update.  I didn't find an pixels with value=0.  There are thin clouds in the frames that I missed in my initial pass to delete frames though.  There are thin clouds in other frames on a different night, but those weren't tagged as invalid.

    Xavier
Sign In or Register to comment.