Dark and Bias for Narrowband

[WbRaDy]

Good morning SGL, im currently in the transition from DSLR to CCD, namely the zwo1600MMC Pro and the question i have is do i need to take dark and bias frames for the 3 narrow-band filters separately? i .e 30 darks for H-alpha, 30 for O111 and the same for S11 and not 1 set that applies to all? i just thought i could get these done while the 3 month old cloud is still about.

 

Cheers

[SyedT]

Hi Wayne,

You only need to take one set of dark and bias frames for all filters. Flat frames should be taken separately for each filter.  

[Demonperformer]

No.

Bias should be a set with no light getting into the camera, as short an exposure as possible. You can therefore do a lot of them. First run through will give you a masterbias which is all you need.

Darks should be a set with no light getting into the camera, taken with the same exposure length you are using and at the same temperature that you are using for your lights.

[Demonperformer]

Beat me to it!

[WbRaDy]

Thanks SyedT, so for instance am i right in assuming if i had 30 darks using O111 and 30 using S11 i can add these together as a master file because the filter used wouldn't of mattered with the scope being covered? and because i had already taken them with an itchy "capture now" finger.. 

[Demonperformer]

You would process your OIII & SII subs separately, but use the same darks & bias for each - assuming they were taken at the same temperature and exposure length. The results are then combined in PS or PI etc. 

[carastro]

If you have set point cooling, then you can keep the darks and re-use them when doing subs of the same length.  Saves a lot of time.  

Carole 

[vlaiv]

Forget bias files with ASI1600

Take darks, flats and flat darks for calibration.

Darks must be taken at exact settings as lights, meaning temperature, offset, gain, exposure length.

Flat darks also with respect to your flats

[SyedT]

1 hour ago, wayne11 said:

Thanks SyedT, so for instance am i right in assuming if i had 30 darks using O111 and 30 using S11 i can add these together as a master file because the filter used wouldn't of mattered with the scope being covered? and because i had already taken them with an itchy "capture now" finger.. 

That's correct, as you're not capturing any photons when taking darks. 

[John78]

2 hours ago, vlaiv said:

Forget bias files with ASI1600

Take darks, flats and flat darks for calibration.

Darks must be taken at exact settings as lights, meaning temperature, offset, gain, exposure length.

Flat darks also with respect to your flats

Flat darks a necessity?  I thought using a master bias was the norm?

[davew]

1 hour ago, John78 said:

Flat darks a necessity?  I thought using a master bias was the norm?

" Norm " depends on who is talking and about what   There are reports of zero time Bias frames becoming unstable and useless for calibration with certain CMOS cameras. I think the 1600 is one of them but I have never used one so can't say from personal experience. ( I know the actual minimum time for a 1600 is above zero but just try longer anyway )

Dave.

It was a thread on Cloudy Nights that drew my attention to Bias frames on the 1600. I've looked but can't find it. The problem may not exist now if, say, a firmware update has been made. Worth checking out anyway. There is another thread re USB settings and Bias frames.

Edit No. two - https://www.cloudynights.com/topic/593573-cmos-bias-or-no-bias/  That was one thread that I've finally found !

Check post #16. 

[vlaiv]

3 hours ago, John78 said:

Flat darks a necessity?  I thought using a master bias was the norm?

Yes they are for proper calibration. If you happen to use bias only, and this is independent of some CMOS sensors having issues with Bias, you run a risk of having your flats under correct (thus leaving darker spots and vignetting).

It largely depends on dark current and length of flat exposures, but for proper calibration you want both bias and dark current signal removed from your flats. Subtracting bias only leaves dark current signal - not good. I do know that people that use DSLRs tend to do calibration with bias only, and that is because of inability to control the temperature - so properly removing dark signal is essentially impossible. Dark optimization helps in this regard, but it is not optimum solution.

On the matter of bias files with ASI1600, it tends to do something a bit different than some other CMOS sensors I had a chance to work with. Bias is stable, meaning that if you take two sets of bias subs and produce two master bias files and power down camera between sessions - those two will contain same bias signal, so it is not bias instability per se (some CMOS cameras recalibrate internally with power on and give different bias signals each time).

What happens with ASI1600 and it looks like it is feature of sensor it self, not drivers (firmware might solve this, but I did not do update), is that bias level, which you can control with offset settings, is affected by duration of exposure, so depending on length of your darks and lights there is additional bias signal not present in bias it self (or rather to be precise it looks like there is some negative offset added to longer exposures, I guess designer of sensor wanted some level of compensation for dark current in cases where there is no temperature regulation - like consumer cameras, to preserve dynamic range).

This feature of sensor makes it very difficult (I don't think it is impossible, but special algorithm must be employed for it) to calibrate with darks of different exposure length, or to extract dark signal only, without bias. Simply subtracting master bias from master dark will not give you dark signal only, but rather dark signal shifted by some offset (if my assumption about what is going on is correct).

This is why I say that you can effectively skip bias frames, they are not needed for proper calibration, and the reason for using them is effectively made void by design of sensor (like using darks of different exposure length).

 

 

[John78]

I'm starting to get to grips with a qhy174m so your comments are very interesting, I'll take a full set of everything next time I get to image and see the differences post stacking  ??

[ollypenrice]

Bias as dark-for-flats works perfectly well for CCD but we read that this may not be so for CMOS. I think, as always, that experiment is the way forward. Just put the lens cap on, if it's light-proof, and shoot a set of dark-flats after each your set of flats. Not all lens caps are light proof though so be sure to exclude all light.

Olly

[WbRaDy]

Thanks for the input, its a lot to take in being new to the Zwo1600....now where is that Microsoft Lifecam again?