StarTools

Mike in Rancho wrote: ↑Tue May 17, 2022 4:13 am I have noticed that Wipe will act differently depending on the bin/no-bin. Of course I always like to get the extra SNR if I can. I think sometimes the anomalies can return though if bin is done later, after Wipe.
We'll need Ivo on the Bin-Wipe relationship though. I am curious myself if/how it affects Wipe's hunt for undulation. Or lack of undulation, maybe that's it.

Cleaner data (or the same data binned = same thing) will cause two things, one important, the other totally not;

Important
It will help with filtering out dark anomalies. Even if you have no dead or "cold" pixels in your dataset whatsoever, the nature of Poissonian noise is such they may come to exist anyway;

Photons arrive and have a chance to be converted into an electron or not (this chance is equal to your sensor's Quantum Efficiency). It is a binary proposition. Ergo there is always a very small chance that none (or just very, very few) of the photons are converted into electrons. And so, dark outliers in your dataset are born.

You can either filter out such outliers with a spatial filter (say a Gaussian blur or median filter, which is what the Dark Anomaly Filter essentially does). Or you can pool neighbouring pixels into bigger super pixels. The effects are roughly the same. The outliers become less conspicuous.

Outside of locking onto Dark Anomalies, bigger or smaller resolution should not affect Wipe's Undulation frequency analysis (frequency is calculated relative to the image size).

Not at all important
Wipe's "courtesy" AutoDev will similarly be influenced by the Poissonian noise. It will allocate more dynamic range to these Poissonian fluctuations, in order to visualise this "detail". Filter out this "detail" and Wipe's courtesy AutoDev will allocate more dynamic range to stuff that actually matters.

Hope that helps!

Thanks, Ivo.

The bin vs dark anomaly makes perfect sense (and gives me a better understanding of the dark-anomaly filter). I'm away from my processing computer for a week, but I will experiment with Wipe using unbinned + bumped-up dark-anomaly filter vs binned. I expect the 'courtesy' autodev will look a lot more similar then.

My "Wipe-first, then bin" looks OK at a simple 75% aggressiveness when I leave the filter at 1 pixel. But I guess I am leaving color undulations that I would pick up on by binning first or increasing the anomaly filter - and I'm just not yet critical enough to recognize them as a problem as I proceed with processing.