OK Carlos, I am hearing what you say - but only for those folks with the luxury of a hyper-cooled CCD.
For everyone else, I disagree that FlatDarks are 'the same as' Biases.
This is simply because the couple of seconds of exposure time often used to obtain Flats (with a 50% ADU level) and therefore the equivalent exposure time for the associated FlatDarks, leads to a
significant difference (IMHO) between FlatDarks and Biases.
If you intend to use Biases in place of Darks (FlatDarks
or LightDarks) then it behoves you to verify, by experiment, that your CCD does actually return images of these two types which have NO DISCERNABLE DIFFERENCE (hence my recommendation to subtract one type from the other, or to divide one type by the other, and actually PROVE that no such difference exists).
To do anything less is, to my mind, nonsensical as you may well be introducing errors that are, in reality, so simple to have avoided in the first place - simply by acquiring FlatDarks in the first place (at a 'penalty' of about 2 seconds per image, for a 30-image FlatDark dataset - probably around ONE MINUTE difference in time between acquiring FlatDarks and Biases, not a huge overhead 'to do it right'
)
As I suggested to Harry, there is a similar argument for those wishing to use Biases instead of LightDarks (because the claim is that there is no 'need' for Darks in a hyper-cooled imager). In other words, the claim is perfectly valid PROVIDING the case has been proved FOR YOUR IMAGER. If you don't take the time to 'check' then the advice would be that it is still 'safer' (statistically) to adopt the 'correct' method (of acquired Darks to correct acquired Lights).
And, if I knew that I could rely on Biases instead of any kind of Darks, I would really want to make sure that my Biases were capable of producing an 'ultra-robust' MasterBias - by using PLENTY of raw subs, and by making sure that the Biases were 'fresh' (in other words that a MasterBias taken, for example a few weeks ago, was statistically IDENTICAL to one acquired from freash Biases taken along with the acquired Lights, this last imaging session).Again, the user should just take the time to run tests - an easy process to do when nights are rubbish for actual imaging.
