OSC or CFA Flat calibration?

[UlteriorModem]

Greetings,

I was attempting to prepare a new master flat for my OSC camera (ASI1600mc)

When I went to calibrate them applying a master dark and master bias I get the following error for each frame.

"No correlation between the master dark and the target frame (Channel 0)"

I tried changing to 'force CFA' but that results in incompatible image geometry and nothing is done.

I 'think' I read somewhere that this is not really a problem and to sort of just ignore it and that the dark will not actually be applied to the calibration. So what is the point of even including a dark?

Thaks!

[dld]

Greetings!

This may not be a problem. What PI tells you is, in your case, by using the master dark over a flat (with exposure time far shorter than the dark frames) will do more harm than good in terms of noise, and will not use it.

If you like to be sure, a flat dark (a dark matching the exposure settings of the flats) should be used. I think this is the best way to go with cameras like yours, where users report "amp glow". Do you have "amp glow"?

Let's see what the ASI1600 users of the forum have to say about this!

[Niall Saunders]

Hi (? ? ?),

In reality, you haven't given enough information to get to the bottom of the issue. You are describing an issue that is associated with OSC Flats, FlatDarks and Biases - just to get started, it would be useful to know the exposure times you used for each of these image groups. It would also be useful to know if TEC was running at the time and, if so, at what % power level - it is never a good idea to try and run TEC at anything near 'full power level', or even at anything near 'zero power level'. Try and find a TEC temperature level that allows you to be able to achieve it on any day of the year, that way it is possible to archive a single collection of calibration frames, at a common temperature that can also be reached during the acquisition of any, and all, Lights.

Which then leaves us with the issue of the image data downloaded and stored by your imaging software - you are using an OSC imager, so all image data files should still contain the full CFA matrix (i.e. they should NOT yet have been deBayered). Further, all pre-processing prior to Image Registration should be on the grey-scale image data, including all Calibration stages. You should only deBayer just before Image Registration and Alignment, and should only perform the Image Integration (stacking) step after this alignment.

The error message that you posted seemed to suggest, to me, that you were dealing with images that may have already been deBayered prior to Calibration (suggested by the reference to 'Channel 0') - but you need to confirm/clarify  that (which you should be able to get from the notes that you wrote into your astronomy paper notepad - you are doing that aren't you  )

Anyway - food for thought. Always happy to try and help, after all we are all learning, all of the time.

[UlteriorModem]

I need to make one correction, when I got the previously mentioned incompatible format error I had tossed in the wrong master dark, it was from my 383L+ oops. 

So I tried the calibration again with "Force CFA" checked and still get the same no correlation error, although then it outputs nothing.

So here are the details although this is sort of a general thing. There are a number of threads on this general Warning message though none dealt specifically with calibrating flats.

The 'flats' were .89 seconds and were not debayered. The cooler was running at -20C cruising along at 20%. The darks were 60sec, and calibrate was checked. Likewise with the Bias also -20C but a fraction of a second. All images are the same temperature and not debayered. The gain and offset is the same for all images. All the same temperature, the same .fits format, size, and binning (1x1) unstreched. With exception of the Master dark and bias which are in xisf format.

I am not 'new' to imaging or PixInsight, just that recently I changed to OSC from shooting narrow band monochrome. So I am pretty famalier with formats, file types, and image sizes.

I should also mention that the resulting flats stack just fine and result in a useable image.

Also when I run a batch pre-process I get this error (with the same dark).

** Warning: The file format reports no CFA pattern for the master dark frame, but it is being forced as per process instance parameters.

And the results of the BPP are fine.

[Niall Saunders]

Try things again - without 'forcing CFA'. That's how I interpretted the error message from your last reply. It seems that PI is being asked to calibrate an image with a CFA matrix using a deBayered MasterDark without a CFA matrix.

FlatDarks are so easy to acquire that I see no advantage whatsoever in trying to 'scale' Darks taken with 60-second exposures. Just get an extra set of FlatDark image data, TE-Cooled to -20°C, at the same exposure time as you used for your flats. And, '0.89 second' exposure times? I would likw to see someone justify why 0.89sec was better than "1 sec"  After all, you could have chosen a TE-Cooling temperature of -23.4°C, and an exposure time of 68.1 seconds - but you didn't 

An exercise that would be worth your effort is to have a look at -20°C integrated (but un-calibrated) image data for your OSC in terms of Biases and and 1- or 2-second FlatDarks. What are the ADU ranges of each? What are the statistics for each?

In other words, do you really need those two groups of images at all? If you are prepared to build up a (very) small set of temperature-correlated LightDarks, to match the typical exposure times that you are likely to use to acquire your Lights, then (IMHO) there is no need for Biases, and quite possibly no need for FlatDarks. So, on a typical imaging session, all you would need to acquire would be Flats and Lights. (And, personally, my imaging train remains set up and constant for months, sometimes years, at a time - so I don't really bother about acquiring Flats, although I do, just because they are so easy to acquire).