Odiug le mars 06, 2019, 09:03:29 21
When thinking about my issue of 12bit imported images limiting the output to12bit as well: How about doing all the image processing in float? That would allow for enough dynamic range without having to worry about saturation. FITS also allows to store float data. Well, image data would double in size, from 16bit integer to 32bit float.


vinvin le mars 07, 2019, 09:27:25 21
Hello Guido, we have planned to go to floating point data for version 1.0. It's very rare to have problems with a 16 bits dynamic however.

Odiug le mars 10, 2019, 05:27:39 17
I agree. This suggestion came from the observation that my 12 bit camera RAWs were imported as 12 bit FITS and so the processing was also stuck at 12 bits. With float that would not have been a problem.

Clear Skies