SharpCap Forums

Hi Robin,
I tried to use sigma clipping while livestacking. I set the threshold at 3 and was removing .0005% pixels. A plane flew through view and the clipping dimmed track but left it there. Does this seem right? I have used the satellite trail remover and get about the same results. Unfortunately this is becoming more of a problem as the years pass. Any preference to which method you recommend?
I'm using the satellite predictor and it does help

Thank You

Hi
We need a filter with functionality similar to the FWHM and Brightness filters, which discard the image below a certain threshold.
This would be the best option.

Carlo

Hi,

sigma clipping is not 100% perfect - at least not without setting a very very carefully controlled threshold, which is almost impossible to do when live stacking (much easier in traditional post processing where you can stack, tweak the threshold, stack again, etc).

Imagine the situation that trail of the light of the plane is several pixels wide - while the core of the trail will be bright, the edge pixels of the trail might only be a tiny bit lighter than the usual background - they might be close enough to the normal value to slip through without the sigma clipping cutting them off, even if the brighter pixels near the core of the trail are being excluded.

I think Carlo is right that we will need something that is a filter to exclude frames with problems *and* some sort of hybrid approach where you can confirm or deny rejection of a frame that has been flagged as suspicious...

cheers,

Robin

Robin,
The brightness filter deletes frames which are to dim. What if you added removing frames which get to bright as well. Wouldn’t be much use for small satellites but might help for very bright planes flying through

Hi,

right now, the brightness filter works on the brightness of the stars detected in the image rather than the brightness of the whole image, so we are not measuring anything that would give a good handle on a plane passing through... Measuring the stars is key to the current filter working, as if you get cloud come in then the average brightness of the image will most likely increase (unless you are in a *really* dark location), but the stars will get fainter/vanish.

cheers,

Robin