Maxim Quality Measurements.

[Tandum (Robin)]

I capture and stack up my images with maxim and try to use the Quality tests to toss out the rubbish. However looking at some subs I have caught recently, these tests seem to pass on some real rubbish in the mix. The bad one here has had clouds floating across it I think.

These 2 O3 subs both passed the tests and gave these figures for the good and bad one respectively :-

Contrast 0.400
Intensity 314785.2
Roundness 0.164
FWHM 1.908

Contrast 0.495
Intensity 267174.9
Roundness 0.128
FWHM 2.029

As you see the results are pretty damn close so as it stands the quality check won't differentiate between them.

Does anyone have any idea of another test we could ask maxim to add that would pick up crap like this?

[bmitchell82 (Brendan)]

I go the full manual way and compare because even CCDStack2 doesn't always pick up the bad ones. My eyes are far better at distinguishing errors than a PC sometimes!

[mill (Martin)]

Yep i do the same as Brendan and look at each picture separately, even if there are a hundred of them.

[RobF (Rob)]

First thing I always do is a quick animation to pick up any funny business and compare frames.

[Tandum (Robin)]

Looks like the standard deviation for bad ones are about triple all the others regardless of the camera filter so I should be able to script it first to check.

Life's too short to spend it in front of a PC blinking images