Image Calib Refuses To Apply Dark

[Starcruiser]

Been trying to get to the bottom of this for two days.  Gave up and did Preprocessing in CCDStack.  Acts as if the dark is ignored in PI.  Here are two screen shots of my settings.

[Starcruiser]

Second screen shot:

[RickS]

For a CMOS camera you probably want to turn off dark optimization.

Cheers,
Rick.

[sharkmelley]

Second screen shot:

What makes you think the calibration refuses to apply the dark?
If I understand those screenshots correctly, scaling factors of around 0.94-0.95 are being applied to the dark.  That is almost 1.0 i.e. very similar to a straightforward subtraction of the whole dark.
The bit to look for in the Process Console or the log file is where it mentions the dark scaling factor:
e.g.
Dark scaling factors:
t0 = 0.939

Mark

[Starcruiser]

Second screen shot:

What makes you think the calibration refuses to apply the dark?
If I understand those screenshots correctly, scaling factors of around 0.94-0.95 are being applied to the dark.  That is almost 1.0 i.e. very similar to a straightforward subtraction of the whole dark.
The bit to look for in the Process Console or the log file is where it mentions the dark scaling factor:
e.g.
Dark scaling factors:
t0 = 0.939

Mark

Because after calibration the subs are covered with the identical hot pixels which match up similarly to the dark master.  When calibrated in CCDStack this does not occur.  This behavior did not happen until after the first time I got this message a couple days ago: "** Warning: No correlation between the master dark and target frames" and those subs likewise were full of hot pixels as it is my understanding the master is ignored in such a case.

Now I don't get the warning and everything is as shown in my screen shots and the subs are not worth integrating after calibration.  Yes I see what you mentioned about the scaling factors being applied but the results sure don't show it.  After spending hours over two days doing this over and over I gave up and did the calibration in CCDStack.  I don't remember having this issue with PI in the past until now.

Thanks for your response.

[Starcruiser]

For a CMOS camera you probably want to turn off dark optimization.

Cheers,
Rick.

Thanks but I tried that earlier in the game.  Did not fix any of this.

[pfile]

something must have changed - is it a new dark master you are working with?

no correlation means that IC scaled the dark down incrementally and repeatedly subtracted the scaled dark from the light, and found no minimum in noise. this usually happens when the lights are so short that they exhibit almost no dark signal. what is the duration relationship between the lights and dark in your case?

rob

[Enzo De Bernardini]

Because after calibration the subs are covered with the identical hot pixels

Sometimes yes, but ImageIntegration (rejection algorithms) usually takes care of them. Try WinsorizedSigmaClipping.

HTH.

Best regards,

Enzo.

[Starcruiser]

Because after calibration the subs are covered with the identical hot pixels

Sometimes yes, but ImageIntegration (rejection algorithms) usually takes care of them. Try WinsorizedSigmaClipping.

HTH.

Best regards,

Enzo.

Thanks, but thats what I have been using.

[Starcruiser]

something must have changed - is it a new dark master you are working with?

no correlation means that IC scaled the dark down incrementally and repeatedly subtracted the scaled dark from the light, and found no minimum in noise. this usually happens when the lights are so short that they exhibit almost no dark signal. what is the duration relationship between the lights and dark in your case?

rob

Thanks.  I am using 10 min lights with 10 min darks.  Over two days I have made several new dark 10 min. masters, the problem only gets fixed in CCDStack for the calibration phase and extract the color channels.  After that I bring it all back over to PI so I can get something done on this job.  I am starting to wonder if something has gone sour with my ImageCalibration tool.

[pfile]

well i suppose the only way to rule that out is if you post your calibration masters and a light and see what happens when someone else tries...

rob

[Starcruiser]

Well my last ditch attempt as an experiment was to acquire all new -10C darks last night and make a fresh master from those and see how that goes running it on a group of just three lights and see of the problem is gone with ImageCalibration module.

[sharkmelley]

Well my last ditch attempt as an experiment was to acquire all new -10C darks last night and make a fresh master from those and see how that goes running it on a group of just three lights and see of the problem is gone with ImageCalibration module.

And what was the result?
Were the lights at -10C?

[Starcruiser]

Well my last ditch attempt as an experiment was to acquire all new -10C darks last night and make a fresh master from those and see how that goes running it on a group of just three lights and see of the problem is gone with ImageCalibration module.

And what was the result?
Were the lights at -10C?

Hi, thanks for checking in.  Went ahead and reacquired a new set of -10C darks and bias, made the new masters from them and applied them to several of the earlier problematic -10C lights.  All appears to be in order.  Not sure what caused my earlier difficulties but we seem to be good to go.--Thanks.