Adding to existing data?

[CaptainShiznit]

Hey,

So I follow AstroBackyard on YouTube, and he often talks about capturing new data to add to and improve existing data. I was hoping someone could clear up how this is done.

Aren't the new images going to be out of alignment with the old ones? Slightly off framing or camera rotation? Does it matter? Can DeepSkyStacker compensate for this or is there a way to reframe a new shot precisely like the old ones? 

Also, I've made an assumption that you just stack the new data with the old in DSS / similar. Is this correct? Are there any particular settings to play with that differ from stacking a single session?

Probably over thinking this.

Cheers!

[bottletopburly]

In DSS you create a new tab and load lights , darks , flats etc  so each imaging session is on a separate tab , the end result is all data from each session , I use Siril slightly different in that you stack and align the final stacked Fit files from each session  together .

[Adam1234]

As bottletopburly says, add the images from each night as a separate tab, the tabs are at the bottom of the table if you haven't seen them already. 

If you try and frame your new shot as close to your old one as best you can, maybe have one of your old shots handy, but DSS should compensate for a slight misalignment and rotate your images to match. Depending on the amount of misalignment, there will be some stacking artifacts around the edges which you can crop out.

Edited March 19, 2020 by Adam1234

[gilesco]

7 hours ago, CaptainShiznit said:

Hey,

So I follow AstroBackyard on YouTube, and he often talks about capturing new data to add to and improve existing data. I was hoping someone could clear up how this is done.

Aren't the new images going to be out of alignment with the old ones? Slightly off framing or camera rotation? Does it matter? Can DeepSkyStacker compensate for this or is there a way to reframe a new shot precisely like the old ones? 

Also, I've made an assumption that you just stack the new data with the old in DSS / similar. Is this correct? Are there any particular settings to play with that differ from stacking a single session?

Probably over thinking this.

Cheers!

APP (Astro Pixel Processor) does this using a process called registration as part of the workflow, the pictures do not have to be aligned, they just have to share features across frames, you can use APP in this way to create large mosaic images of sections of the sky. As long as you have the targets in view it will improve the detail of your targets. APP can also take data from different cameras and telescopes and do the same thing.

[fireballxl5]

When combining data taken many months apart and so cal files (flats in particular) have been refreshed, this is not so straight forward. Whats the best process for adding data on the same target in these cases?

[gilesco]

55 minutes ago, fireballxl5 said:

When combining data taken many months apart and so cal files (flats in particular) have been refreshed, this is not so straight forward. Whats the best process for adding data on the same target in these cases?

APP has multi-session mode, which means you assign Darks, Darkflats, Bias and Lights to each session.

Sessions could mean the same telescope / equipment at different times, or equally different telescopes and equipment at either the same time or at different times.

[almcl]

2 hours ago, fireballxl5 said:

When combining data taken many months apart and so cal files (flats in particular) have been refreshed, this is not so straight forward. Whats the best process for adding data on the same target in these cases?

I find putting only the offset bias files in tab 1 and then putting first sessions lights, flats and darks (if you use them) in tab two, the lights &c from the 2nd night in tab 3 and so on works OK.  Definitely don't want individual night's calibration frames in tab 1 as they will then be applied to all tabs. 

DSS will work out the light with the highest score and use this as the reference frame to stack all the sessions. 

One caveat: the latest (beta) version of DSS sometimes creates spurious stars (some completely outside the frame) and these have to be manually deselected and some real ones used instead otherwise the stack doesn't work.

Edited March 19, 2020 by almcl

[fireballxl5]

35 minutes ago, almcl said:

I find putting only the offset bias files in tab 1 and then putting first sessions lights, flats and darks (if you use them) in tab two, the lights &c from the 2nd night in tab 3 and so on works OK.  Definitely don't want individual night's calibration frames in tab 1 as they will then be applied to all tabs. 

DSS will work out...

BTW, I mainly use MaxImDL for calibration and stacking. I do subsequent processing in PI, so could calibrate using this if it can cope with multiple session calibration.

[oldannie]

Doug German does an excellent tutorial on  Multi-session stacking in DSS.

Annie

[Marvin Jenkins]

Thanks Annie, as usual the thread veered away from the original post with alarming speed.

[gilesco]

3 hours ago, Marvin Jenkins said:

Thanks Annie, as usual the thread veered away from the original post with alarming speed.

I was not trying to hijack the thread, but as I have no experience with DSS, I thought I would give my experience of how this is done with a different, but similar piece of software, from other posts, it appears the same nomenclature and processes apply, so it is still relevant.

[CaptainShiznit]

1 hour ago, gilesco said:

I was not trying to hijack the thread, but as I have no experience with DSS, I thought I would give my experience of how this is done with a different, but similar piece of software, from other posts, it appears the same nomenclature and processes apply, so it is still relevant.

No bother regarding hijacking. All relevant information for me so cheers 😊