21 replies to this topic

[Bluephoton]

I have accumulated approximately 20 hours of M101 data over 5 nights of imaging. I captured Flats and Biases for each session and have a dark library for the sensor temp used. I have not altered the image train during this series of imaging sessions.

 

I use Siril so I am thinking I will just dump all the lights, flats and, biases in one directory and use the library darks to process everything. Will this work?

 

I have looked at a lot of tutorials that say to create a master bias, flat etc for each session and then combine but the info I have seen doesn't actually explain how to do this.

[Hesiod]

I use the same master bias but for all sessions, but master dark and master flat for each night since I shot with a dslr.
I assemble the master dark and flat for each session, pre-cook the session subs and only after that dump all the calibrated files together for the subsequent steps

[dx_ron]

A single master bias is fine. Making a new one a couple of times a year is plenty. If you used your 533, then the same applies - darks only need to be re-made a couple of times a year.

 

The issue is flats. You say you did not alter the image train, so it's possible that one of the sets of flats would do. But if you want to play it safe, each night's lights should be calibrated with the same night's flat.

 

Have you seen this thread? https://www.cloudyni...ace/?p=12736512

[Drothgeb]

The only thing that changes night to night is the dust motes. If your system was clean, and you don’t have any, then you can probably get by with one set of flats. I inspect them from the first night. If they look clean that’s it. If they have dust motes (pretty rare), then I shoot flats every night.

[jml79]

Here is a post with a few scripts to help with multi-night stacking in Siril. Basically, setup each night with it's own folders, run the first script, copy the PP_lights from the process folder to a new folder, run the second script on the second night's data and repeat. In the script you can change to offset number to have 3, 4 or more nights. Then load all of the calibrated images into a sequence, register and stack. I usually use the same Biases and Darks for each night but you don't have to. I often use the same flats if I haven't touched the scope but you have to make sure to check each nights data to see if you need new flats before doing anything.

 

https://www.cloudyni...2#entry12374815

[idclimber]

Instead of repeating your bias files nightly, create a master that contains 100-200 samples. This does not take long and only needs to be repeated about once per year. Perhaps every 6 months if you are so motivated. Once I have the master, I delete the individual bias.

 

I also do the same for my master darks, although I use capture 50 of them. 

 

The logic behind this is the same as more data in our lights to increase SNR. In the darks/bias case we want the fixed pattern noise and do not want the random noise. Random noise if still in the master is essentially added into our lights and flats. 

[Spaceman 56]

 

Have you seen this thread? https://www.cloudyni...ace/?p=12736512

thanks to EPinNC for this fine piece of work.  

 

personally I have always found Siril to be a nightmare for manual processing, but using EPinNCs excellent tutorial 

I was able to do a manual process. 

 

I won't say its easy, or that I will try and do it often, but without EPinNCs tutorial I had no hope.

 

Spaceman

[italic]

Check out sirilic for batch processing multiple nights. It's effectively WBPP but very basic. It will output the final integration and I believe each night's stack as well. I got really tired of stacking manually when I used siril, but this tool really helped.

[danny1976]

You can do it in Sirilic which is explained here.

 

I feel more comfortable with taking the pp_lights from the OSC_Preprocessing script from each night, then register all the pp_lights and stack.

[dx_ron]

It didn't take me too terribly long to get used to manual pre-processing. In fact, I quickly realized that it is way easier to have a folder named M3_06-08-23 and switch Siril's home directory there than it is to continually have to fuss with copying the files from M3_06-08-23 to a folder called Lights.

[Dante 34:139]

How do you guys deal with field rotation across multiple nights? I tried yesterday stacking lights taken in separate sessions and it went all out of whack. 

[EPinNC]

How do you guys deal with field rotation across multiple nights? I tried yesterday stacking lights taken in separate sessions and it went all out of whack. 

By "field rotation", do you mean your camera was in a slightly different (rotated) position between sessions?  I would think your stacking software should handle that during the registration step (i.e., move each image left/right/up/down/rotate).

 

Or do you mean field rotation during each sub-exposure?  (like what you would get with an alt-az mount)  I don't know how you would fix that.

 

What exactly went wrong?

[Dante 34:139]

I mean that if you leave everything untouched in your imaging train and simply 'slew and center' on your target across multiple nights, then you'll have the target rotating in your frame each time. Nothing that the stacking software can't deal with but you either need to crop at the end or you need to rotate your sensor a bit every night. I mean, you could ignore it if you shoot for 2 or 3 consecutive nights, but if many days pass by then the cropping would be too large.

 

Anyway, I googled it and the answer is: manual rotator + NINA 'slew center AND rotate' instruction. Luckily enough my lil refractor has a built in rotator so I should be able to deal with it. Not sure how to do this in Kstars/Ekos but I'm sure there's a way.

Edited by Dante 34:139, 09 June 2023 - 07:54 AM.

[idclimber]

I mean that if you leave everything untouched in your imaging train and simply 'slew and center' on your target across multiple nights, then you'll have the target rotating in your frame each time. Nothing that the stacking software can't deal with but you either need to crop at the end or you need to rotate your sensor a bit every night. I mean, you could ignore it if you shoot for 2 or 3 consecutive nights, but if many days pass by then the cropping would be too large.

 

Anyway, I googled it and the answer is: manual rotator + NINA 'slew center AND rotate' instruction. Luckily enough my lil refractor has a built in rotator so I should be able to deal with it. Not sure how to do this in Kstars/Ekos but I'm sure there's a way.

That should not happen with an equatorial mount. The rotation angle that you have at the beginning of the night should be the same as it is at the end. Subsequent nights should also be the same as long as you match that rotation. 

 

On my ZWO if I setup the camera so it is perpendicular to the base of the mount, then I will have an angle of 90/270 degrees. Those two angles are the same and if you start with 90 it will change to 270 as you cross the meridian. Play with the mount with the camera pointed around the meridian and it will begin to make more sense. 

 

I typically plan the rotation of an image prior to capturing my first sub. Lets say I decided to image a target at 115 deg. I do a rough guess how much I need to turn clockwise from 90 and then check by plate solving an image and verifying it is within one degree of my target. There is a routine in Voyager that assists in this task, I assume NINA has something similar? 

[Dante 34:139]

Yep NINA has the Framing Assistant that does exactly that. It can also solve an image taken on a previous session (different day) and slew and center on it.

 

That being said I probably messed up with the scope 'manual rotator' since I was stacking pics done few days ago and I don't remember exactly how I framed them. I googled it and saw it was a 'common issue' and I forgot stars don't move  so it's probably a common issue in the sense that people rotate their framing for each different target and then have to recenter the fov correctly before shooting again an old target. Kinda thing.

 

Anyway the 'slew, center and rotate' instruction in NINA should be all it's needed to have consistent framing across different sessions.

[smiller]

How do you guys deal with field rotation across multiple nights? I tried yesterday stacking lights taken in separate sessions and it went all out of whack. 

As others have said, you shouldn’t have any rotation with an EQ mount.

 

But regardless, ignoring the cause, if you do have a case where you are stacking a bunch of subs that are rotated, as you noted the stacking programs will rotate them.   And APP, SIRIL and PI can also rotate them into a larger master frame so that you can decide exactly how to crop it later, rather than have the stacking program crop it to you reference frame.   APP and PI do this by default, but for SIRIL you need to use the relatively new option called “-framing=max” along with 2 pass registration.   I modified my SIRIL OSC processing scripts to have this option.

 

The nice thing about doing it afterwards is often you can crop to a more optimal, and often larger, final frame than what is represented by a single reference frame.

Edited by smiller, 09 June 2023 - 08:44 AM.

[Alex McConahay]

Did you leave the camera attached from one night to the next?

 

If you had to reinstall it, it may have rotated. Or if you inadvertently rotated it, you may have rotated it. 

 

Alex

[Dante 34:139]

The camera is always attached. But I'm doing all sort of adjustments in these firsts days of AP so I for sure messed up with the rotator. But another 'source of unwanted field rotation' is the polar alignment procedure. I try to be as accurate as possible manually before going in the 3-point polar alignment plugin in NINA, so that means that every night I have to unclutch the RA axis to rotate the polar reticule in order to have the 0 perfectly perpendicular to the center, then put polaris in the circle where it need to be. And then rotate back the RA axis (aka the counterweight shaft) to its park position. And finally reset the home position in the mount's driver and in NINA as well. Then I'm ready to go.

 

I was wondering if you did the same of used a more 'laxed' first polar alignment, meaning that you don't rotate the polar reticule every night and just go to fine tune it via a 'software polar alignment' like platesolving or NINA's plugin.

[dx_ron]

My guess is that very few of us use (or even have) a polar scope any more. But - I'm not seeing how what you described will cause field rotation. As long as you are reasonably polar aligned by the time you start taking subs, the rotation should be minimal. I *think* (but have never worried enough to be sure) that you can get "extra" field rotation if you are guiding on a star that is rather far from the center of the target (much farther away than the stars in an OAG, in particular).

 

Maybe you could show us what came out of your stacking process, that would help us understand your issue.

[danny1976]

If you plate solve an image you get the RA/Dec coordinates AND rotation. So take an image from the previous session, plate solve it and adjust rotation if needed. Usually a good quality telescope has a camera rotator built-in. 

[SuperSirLink]

I have been loving the ASIair for this.  It allows for really quick, easy, and reproducible framing when you need to tear down each night but want to image over multiple nights.

[dx_ron]

All of the major session managers allow you to quickly and easily reproduce framing from night to night.