Adding more data, how to then stack etc.

[smr]

Hi all,

For the first time I am imaging the same object again, to gather more data. And whilst I'm imaging I may as well ask what I know nothing about... I'm taking more images, albeit tonight I am taking subs of 1 minute unguided as opposed to 90 seconds unguided. I just feel like I am throwing away too many frames at 90 seconds, so until I get guiding I'll stick to 60 seconds. 

Last week I gathered data on Pleiades with around 2-3 hours of 90 second subs. When imaging again what is the procedure with the new, added light frames? 

Do I just put them into the same folder as the lights I have already collected and stack them all again ?

What about calibration frames? I took flats, darks and bias frames last time I imaged this object, do I need to take them again for this session, and if so do I throw all of the calibration frames together with the last session's calibration frames?

Hope I'm making sense and appreciate any replies.

[vlaiv]

Best way to do it is use same exposure, but you can stack different exposures.

What stacking software are you using? If DSS - there is stacking method that will deal with different exposure lengths (well any stacking method will, but this one is supposed to be rather good at it). I can't remember what its called - need to look it up.

"Auto Adaptive Weighted Average" - try that one first, but if it does not give you good results, revert to regular sigma clip.

Just add new frames to stack together with old frames (redo complete stack with all frames, but assign proper calibration frames to each group).

You need to take new set of calibration frames - all but bias. Bias you can reuse. You will need new flats if you changed anything in optical train - like taking scope apart, or rotating camera with respect to filters. You can try using same flats, but if it does not work - you need to redo them (and it is just best to do them again if you have mobile setup as it is hard or next to impossible to put everything back in the same way just to take flats if you see that previous ones don't work).

You need darks to be of matching length to light subs. You can use old darks if you enable dark optimization. This will try to scale your darks properly, but depending on camera used might not work as expected (CMOS dedicated astro cams have problem with bias and it does not work with such cameras for example).

[steppenwolf]

12 hours ago, smr said:

Do I just put them into the same folder as the lights I have already collected and stack them all again ?

Assuming that you don't have a fixed observatory, I'd place the two sessions in separate folders.

12 hours ago, smr said:

What about calibration frames? I took flats, darks and bias frames last time I imaged this object, do I need to take them again for this session, and if so do I throw all of the calibration frames together with the last session's calibration frames?

Making the same assumption, the Bias frames will be fine for use with both data sets, I suspect that the Darks will also be fine if you can use scaling but it wouldn't take you too long to collect another 20  darks at 60 seconds. I would take a fresh set of Flats for the new data as in my opinion, the Flats are the most important calibration frame!

Calibrate the two sessions separately then copy the calibrated files from both sessions into a common folder then align and stack them.

[smr]

That's great, thanks. The clouds have rolled in so just doing some darks now, and then I'll take some flats. So DSS will automatically know which darks are for which lights, and which flats are for which lights?

[vlaiv]

Have a look here for explanation

It describes how to add files to main group and other groups (including calibration files).

[smr]

Thanks, that's great.

[david_taurus83]

Just put each session into different groups in DSS. For example, night 1 in main group, night 2 in group 2, night 3 in group 3 etc. Shoot new flats with each session. Bias you can reuse. Use your 90s darks for your first group and a new set of 60s darks for the rest assuming the rest of your lights will be 60s.

[ollypenrice]

With as many subs as this, and unguided (so given natural dither) I'd be inclined to try stacking without darks at all. I'd try a master bias instead of a dark. I don't even use darks with a set point cooled CCD because I get cleaner results without them.

Olly

[catburglar]

I agree with Olly- I don’t usually bother with darks if using short frames and a relatively low noise camera.

There’s the natural dither option, or you can use APT’s internal dithering option which works really well even without guiding.I

I set it to dither by up to 5 pixels every frame- it  only adds a couple of seconds between each exposure and means it does a pseudo random walk around the centre of the image. It is very effective a sorting out hot pixels and such like.