What should my master bias look like?

[Diane Miller]

Or probably more to the point, what are the rejection files supposed to look like?

I'm using a DSLR (Canon 5D Mk II) and I shot 200 bias frames at 1/8000 sec, lens cap and eyepiece covered and in a dark room.  I have PI set to Format Explorer > DSLR_RAW > Edit preferences > Pure RAW.

I’m running ImgeIntegration with parameters I think are most recommended:
Combination:  Average
Normalization:  No normalization
Weights:  Don’t care
Scale estimator: Iterative k-sigma
Rejection algorithm: Linear fit clipping (I have 200 frames)
Sigma low and high both to 7

Should I experiment with the low and high values?  If so, which direction?

Attached is a screenshot, with STF enabled.

[pfile]

if you want to reject more pixels, you move the high/low sliders down. with bias frames the sensor is active for such a short time that cosmic ray hits are unlikely, but they still happen. with that many bias frames you might see one or two streaks in the rejection map. however, with so many input frames, if something is rejected from just one frame, it will have intensity 1/200 in the rejection map and could be hard to see. you can use the pixelmath expression iif($T>0,1,$T) on your rejection map to make any rejected pixels show up as bright white.

rob

[Warhen]

 At least visually, your bias looks credible and good. I don't typically generate the rejection maps for calibration masters. Typically, default slider settings for sigma rejection are fine but if you see cosmic ray strikes you can decrease the tolerance by sliding the slider left as Rob says.

[Diane Miller]

Thanks guys!  No obvious comic rays.