Using lower iso ?

[The Admiral]

1 minute ago, Galen Gilmore said:

I always thought that dark frames are required for DSLR's, and that only bias frames are required CCD?

I think in an ideal world, yes, but the problem is that the dark noise is very sensitive to temperature, so darks need to be performed at the sensor temperature used for the lights. Problem is, DSLRs are not temperature controlled and the sensor temperature will change as frames are taken. This is not the case with cooled CMOS chips, and darks can be taken whenever convenient, building up a library for different running temperatures, if necessary. The recommendation for DSLRs appears to be to use a master bias instead of a master dark, and to dither and use a bad pixel map. To be honest, I just use the master bias in place of the master dark. This won't necessarily remove things like amp glow though, or so I believe.

Ian

[alacant]

1 hour ago, MARS1960 said:

i have never bothered with darks, just lights, flats and bias

+1. On my Canon, dark frames -as well as being a PITA- serve only to add more noise. HTH.

[Alien 13]

I never use darks, unless I am taking single widefield shots with foreground in them and then I use the inbuilt long exp noise reduction (dark).

Alan

[SAW]

15 hours ago, wxsatuser said:

DSS recommended settings should be fine to start with.

Dark noise is reduced by the square root of the total number of exposures averaged.
I would go for a minimum of 25 or 36 would be even better, only 1/5 the noise will be
added if you use 25 dark frames or 1/6 of the noise if you use 36 dark frames and so on.

BTW
Use ISO 800 or 1600 because this is where most modern cameras have their lowest read noise.
If the target is really bright you can drop the ISO but faint targets use higher ISO.

I'll get some more darks tonight and if it's clear more lights. So does DSS always use the square root of the amount of darks ? Is this the same for lights, flats and bias ?

How many bias should I be using ?

I'll try stacking again with more darks and with no darks.

[Uranium235]

Lower ISOs come into play when you are using very fast optics (ie: f2), and you want to retain star colour. A 5min exposure @ ISO800 (at f2) would more than likely have very little star colour. But if you halve that (ISO400), you are doubling your well depth - and therefore increasing your chances of retaining star colour.

With an uncooled DSLR, you probably want to avoid quite long exposures (over 10min) since the buildup of thermal noise will become an issue - unless its quite cold outside!

[wxsatuser]

27 minutes ago, SAW said:

I'll get some more darks tonight and if it's clear more lights. So does DSS always use the square root of the amount of darks ? Is this the same for lights, flats and bias ?

How many bias should I be using ?

I'll try stacking again with more darks and with no darks.

It's nothing to do with DSS as such, it's the result of the stacking procedure on the SNR.

If you use darks you don't actually need bias as the bias is in the darks and will be subtracted anyway.

[Stub Mandrel]

33 minutes ago, Uranium235 said:

doubling your well depth

Strictly speaking well depth is fixed, but by doubling ISO you can make use of twice as much depth.

[SAW]

So this is stacked with 20 lights, 36 darks and 0 bias. Tried a some basic processing in Pixinsight and then tweaked in LR as you can see the data is there but I'm no good at processing it LOL.

It's all good practice :-)