›

Subframe Selector or NSG

starfield
in PixInsight
Quick question.   After working with the NSG tutorials, it's become a regular part of my workflow.   I've found it makes quite a difference, especially for sessions that I shoot from my light polluted backyard.      It makes me wonder if there's a real need to master sub-frame selector now?    Or has it essentially been replaced by NSG?

Comments

  • BlockHeadBlockHead
    SFS measures images attributes that can be described as affecting the quality of data including things like FWHM, eccentricity...etc. Some of the measures are also related to signal strength and noise (star count). 
    NSG on the other hand measures one attribute (really well)- star flux (signal strength). So in this sense NSG is more focused on a single measure but...

    1. Signal strength via stellar photometry is in my opinion the best method to weight images. Measures such as FWHM and other stellar morphological attributes are somewhat degenerate in their interpretation. People, including Juan, give way to much weight to these measures.  NSG does the normalization and weighting particularly well by fitting gradients and measuring the same sources in each image.

    2. Since SFS is not measuring the same sources in each image- it is not doing relative photometric measurements. This is also true to the new PSFSignal weighting now employed in SFS, Imageintegration. Juan argues that a convolution of these measures in some (complex) way can be used to determine image quality. His new algorithm does now, in many conditions, approximate the results of the photometric measurements. SFS, as a tool, is a manifestation of this philosophy. It gives the user the ability to weight (determine quality) of images based on many measures.

    At the end of the day- I prefer stellar photometry as a primary measure of image quality. You can follow the measurement straight through to the output/result- and it is easy to tell if something does not make sense. With Juan's algorithmic proxies of these measures- the certainty of the result is definitely less obvious and like a differential equation- there can sometimes be surprising behaviors. I don't like that. 

    -the Blockhead
  • starfield
    Thanks, Adam.
Sign In or Register to comment.