Well after a few tries I'm starting to get an image, not very good but at least there's a picture. 13x30sec, master dark 30sec, master flat, no bias, processed in Pixinsight. The only processing is combining into 1 frame then dynamic crop and histogram correction.
I have a question regarding bias frames which is not a criticism or a
disagreement with anyone[I know far too little to presume to do either.]
I have a number of books on various types of astrophotography; all
written or updated fairly recently. All written by acknowledged experts
in the field. All state categorically that bias frames are totally unnecessary unless using unmatched darks, especially as darks contain bias information. If this is true, why do some members push us novices to use them? I thought that this topic related to this thread, so was not really hijacking it.
raymo
Ray. You are correct. The short explanation is that if you are using dslr you are better off not using bias frames on the darks and lights, BUT, you will need a bias frame for your flats.
Long answer
A DSLR set of frames will be;
Lights (of course)
Darks (to match the lights)
Flats
Bias (to subtract from the flats)
Preprocessing as follows;
Average stack your darks with no pixel rejection to make a master dark
Average stack your bias with no pixel rejection to make a master bias
Calibrate your flats with your master bias to make master flat
Calibrate your lights with your master dark and master flat
Debayer, align and stack.
Why? Because, DSLR data is modified in the camera and is therefore, not linear - unlike dedicated CCD camera data, with wich you can do magical things such as dark scaling.
I have tried this method using PixInsight, AstroArt and Nebulosity with consistent and superior results to those methods typically reserved for 16bit linear data - CCDs.
I hope that helps.
Disclaimer: there may some points of fine tuning that others may suggest. Dont be led astray by folklore. There are good reasons to deal with DSLR data as above. Otherwise you will almost certainly truncate your data. Bottom line, dont touch your lights or darks with a bias frame.
At high iso, you *will* see noise in bias images from non-zero pixel values. It's well worth removing. That said, you'll want a *lot* of bias frames to remove any random noise component you are seeing. I use 100 bias frames and they produce a noticeable improvement in my images.
