Is there a setting to reject the worst subs in WBPP?
Use Subframe Selector, or Blink.
I wouldn’t trust WBPP alone to reject anything. If theres just a single bad sub, maybe two, and you have 50-100 good ones, pixel rejection will do it’s work and essentially get rid of those bad ones. Beyond that, you’ll want to do manual intervention.
As far as I know no…
You can only adjust the weighting parameters which won’t entirely reject it but will weigh the bad subs so low it won’t have any affect on the final stack.
This is already pretty good on default. However I just cull them using blinking methods and or subframe selector. Keeping bad or subpar subs to me just feels weird.
Brian Puhl · Nov 1, 2025, 12:26 AM
Use Subframe Selector, or Blink.
I wouldn’t trust WBPP alone to reject anything. If theres just a single bad sub, maybe two, and you have 50-100 good ones, pixel rejection will do it’s work and essentially get rid of those bad ones. Beyond that, you’ll want to do manual intervention.
I agree, I would use Blink to page through your subs and just make sure that you have a low number of poor subs to good ones if you’re going to let pixel rejection do its thing. If you have a lot of iffy subs, then go through them yourself and reject them from your batch.
Brian Puhl · Nov 1, 2025 at 12:26 AM
Use Subframe Selector, or Blink.
I wouldn’t trust WBPP alone to reject anything. If theres just a single bad sub, maybe two, and you have 50-100 good ones, pixel rejection will do it’s work and essentially get rid of those bad ones. Beyond that, you’ll want to do manual intervention.
I agree. Even better than blink, Tenmon app on Macbook allows you to do full screen blink with the ability to delete the bad frames. Not sure if it runs on PC. Its been a while so im not sure if you can currently delete directly from the blink in PI.
I wish Nina let you have median values in the name. I have HFR to give an idea of seeing and star count which does help identify background issues ( as well as seeing). Median values tend to be the best identifier of clouds for me at bortle 4. I track it on the image graph, but would be nice to have it in the name of the file to quickly see it