Hi All,
first real post, and a relative newbie here.
I use a QHY8 OSC and CCDStack to process the resulting FIT's files. Capture is with AstroArt 4.
In calibrating, I don not use darks, but instead use a bias frame, and a flat, both normally a run of about 5 shots, taken at the same time as the lights. Well nearly always, sometimes I shoot the flats the next day, without any changing of the equipment, I just get lazy at the end of a night of imaging.
To calibrate, I first make a master bias, and then a master flat within the Calibrate routine.
Then use the Bias as the "Dark, as well as in the Bias role, with the Flat in the Flat section.
While I don't fully understand the histogram, the calibrated image is clipped at the left hand end, certainly differently to the raw. To get the histogram visible I opened the resulting images in Maxim and viewed the histogram there.
Is this normal? It would have gone on unnoticed, and likely has for some time, it came about as I had sent a few very southern images to a northern friend who processed them and he commented.
If it is not normal, can someone please help this floundering newbie.
Regards,
Gary BEAL
New Zealand

Hi Gary,

Stan is out of town until 7/15 and will address your questions when he returns.

Thank you John. Likely a simple answer, but at the moment this, and few others have me baffled. Rainy day is a good time to put my hand up and ask.
Gary

quote:

... To get the histogram visible I opened the resulting images in Maxim and viewed the histogram there.

That's probably the problem. When I last checked, MaxIm could not handle negative numbers, though it would add an arbitrary number when opening a file containing negatives. Maybe you saved an unsigned FITS, which would truncate negatives to zero?

In any event, it is likely that the source of your observation is that the image is so thin (underexposed or severely filtered) that the background is hardly any different from the bias (i.e. virtually no sky photons were captured). So when you subtract the bias from that thin image, some image pixels are going to become negative because of read-noise. This is usually not a problem for CCDStack but may become a problem if you transfer images to other software that cannot handle negatives. Or it is a problem if you write-out the data in a format that cannot handle negatives (e.g. unsigned 16 bit FITS or TIFF). CCDStack usually detects when you attempt to do this and issues a prompt to optionally re-scale the data when writing it out.

BTW, software that is unable to handle negative numbers will add a “pedestal” to the results of any subtraction (usually 100 ADU) in order to avoid too many negative numbers. If desired, you can add in your own pedestal via Pixel Math.

So the bottom line is that thin images (underexposed or severely filtered) will have a background close to zero after calibration, which in itself may look “clipped” but can actually become clipped if you transfer the image to some other software without re-scaling or adding a pedestal.

Stan