Improve Statistics and saturated pixels detection

[Luca]

Few trivial changes could be introduced to help spot on the fly saturated pixels in our images and identify sub-optimal values of bias and offset immediately.

Due to the automatic stretching is not always easy to understand if large portions/stellar cores are saturated.

The first change I’d add is the possibility to express the statistics as normalized 0-1 values. I know that in a 16bit camera 65535 ADU corresponds to a full well capacity, but I think this number is less immediate than a simple “1”. A simple tickbox to normalize the values would be as simple as implementing a division, but very usefull.

Another useful addition could be the possibility to highlight in the photo clipped pixels (pixels with values equal to either 0 or 65535) similarly to how this is done in Photoshop (https://www.martyfriedel.com/blog/how-to-show-real-time-highlights-and-shadows-clipping-in-photoshop). Basically any pixel with value 0 is colored in bright blue, while any pixel with value equal to 65535 is colored in bright red. Clearly it should be possible to turn this on or off from a button in the “Image” window top bar, as the HRF or the cross-hair toggles. This will inform immediately and extremely clearly the user of how many pixels are clipped and if he needs to readjust its gain and offset values

[Ruediger]

I support the idea of a visualization of saturated pixels, because to know that there are saturated pixels from statists is one part. Now to find them another. Some color overlay indicating saturated pixels would be very helpful. Also some local AP friends were searching for this function too.

Cheers
Rüdiger

[Björn Arnold]

I really appreciate the idea. I’ve been looking for this for quite some time. A thought on it: adjustable thresholds for b/w clipping could be helpful.

CS!

Björn