New Siril user! I’m trying to understand why I’m seeing differences when viewing statistics in Siril compared to ASI FITS viewer. Looking at the output from simply stacking bias files (either with average (mean) or median values), when looking at the stacked file in Siril I get the expected average of c. 152. But I’d initially looked at the same file in ASI viewer where I see an average value in the thousands.
I’ve yet to pull this file into my preferred post processing editor, Affinity. But please could someone explain the differences I’m seeing. And wherever there are any gotchas for taking Siril produced files forwarded to other programs. I’ll be working in 32 bit. Thanks very much in advance!
p.s. I’m aware that an offset of 15 on my ASI 2600MM (gain 100) is optimistic, although it seems ok. I’ll likely switch to 50 for safety. And I’ll likely make use of Siril’s synthetic bias facility once I’ve familiarised myself with the program.
Answering my own question! This was obvious after opening one original (from camera) and the stacked images in the online FITS Viewer. And well worded by Nick James at FITS and FIT files suffix – British Astronomical Association:
The raw files you get from the camera are probably 16-bit integer (BITPIX=16). After calibration they may be floating point (BITPIX=-32) … The FP files will be twice the size of the integer ones.
Clearly ASI FITS Viewer isn’t expecting floating point values; no reason why it should. And I’ve learnt something 
.