Overcorrection of master flat in PixInsight not in CCDStack or Nebulosity?

[akulapanam]

I haven't used PixInsight in awhile to do stacking but I had a particularly bad framing issue occur that CCDStack couldn't register.  In any case I've run into an issue where imagecalibration would take my master flat (already calibrated with flat darks) and overcorrect my light frames.  The light frames now have white dust moles and a dark center but are bright towards the edges.  Using the same master flat frame in CCDStack and Nebulosity calibrates the light frame correctly.  I did make sure PixInsight had calibrate unchecked on the master flat.

[RickS]

If you need a master flat that works with PI calibration then I think you'll need to build it in PI.  The FITS standard doesn't specify the range of floating point values and I'm pretty sure that CCDStack and Nebulosity don't use the same 0..1 range as PI.  Maybe you can fudge around this with the FITS format preferences and/or a bit of PixelMath but flats are very fickle!  Another quick & dirty option would be to convert the master flat to 16-bit unsigned in CCDStack or Nebulosity before using it in PI.  You'd lose precision but at least the range of values would match.

Cheers,
Rick.

[akulapanam]

Yep that fixed it!  Thanks

[STAstro]

Sorry to Hijack the post but I have a question around the 16Bit / 32Bit Calibration frames

Does it matter that the Master Calibration frames are 32Bit FP and the light frames are not?  Obviously after calibration the light frames become 32bit, but my camera is definitely not 32 bit

[pfile]

it doesn't matter and is as you point out pretty much unavoidable that the raw frame from the camera is i16.

rob