Back to forums

Processing Pleiades , newbie here!

Processing Deep Sky M45 Messier Objects
Dimitris Kavallieratosandrea tasselli
•46 replies•1.1k views
Dimitris Kavallieratos avatar
Dimitris Kavallieratos avatar
Hello everyone,

This is my first post here, I am very new to this hobby and recently I bought a SW star adventurer 2i pro pack and combined with a nikon D5300 dslr and 200mm f3.5 lens (L-Bracket with counterweight) , I took a shot at M45 from my bortle 8 home.

300+ 20 secs exposures @ 1600 ISO (stacked with DSS), w/ 30 darks and biased frames and I got this result (tiff from DSS).

I was not sure about the exposure time (PA & balance) and because i did not want to risk it, I went with short exposures. Also heavy LP is a factor there and I undestand that long exposures without filter is a bad tactic.

My processing was carried out in GIMP but I cant get decent results....

Is it because of my data (check a sample RAW) or my processing skills are bad? Can you help me with basic processing at GIMP?

Thank you in advance 
Sean van Drogen avatar
Sean van Drogen avatar
#01 Tip for improvement take flats.
Here a version processed in Pixinsight (very quick and dirty)
Crop, DBE, EZ Denoise, PCC + Background Neutr, Histo Transformation, Curve adjust
Cant guide you on GIMP usage but there is something in this data
andrea tasselli avatar
andrea tasselli avatar
Hi,

I've checked you raw and I can see that tracking isn't great so downsampling is the way to go here (by a factor of 4). It also shows faint wisps of nebulosity so there is definitely room of improvement there. The lens you use has some signficant amount of chromatic aberration so that will require some adjustement during post-processing but it would be better to use some minus-violet filter there (on the lens).

As for processing tips: ditch Gimp. Just not made for this. Same as DSS. Start learning Siril. It is free and it is a great program for AP. There are plenty of tutorials to guide you in processing your astrophotos.

And least I forget it, always capture and apply flats to your APs.
Helpful
Die Launische Diva avatar
Die Launische Diva avatar
Welcome Dimitris!

I did a quick light pollution removal using PixInsight's DynamicBackgroundExtraction tool in a cropped version of your image and performed a default stretch just to save as jpg and present it here and encourage you to continue with your new hobby! No denoising, no saturation boost, just an initial crop, gradient removal, and a default stretch:


 
Well, I have two observations to make. First, you definitely need to learn how to take flats and use them. Flats will correct for vignetting, and more importantly, for the dark blotches due to dust/dirt in your sensor. Dust is inevitable (I am sure that your roommate/SO will probably disagree with this statement ) and if it does not annoy you *a lot* your daytime photography, there is no need to clean your sensor. Proper flat frames will take care of it and will correct for vignetting. You can practice taking and using flat frames during daytime: Collect your flats in a dark room using an even light source and  apply one or a master flat in a single photo of a flat surface like your wall. Does your single/master flat corrects for vignetting and dust shadows? If yes, you are on the right track. With proper flats, light pollution modelling and removal will be far easier. Trust me !

The second thing is that your tracking mount didn't manage to keep all your frames centered, which resulted in the bands towards the bottom and left of your integrated image. Mind you, I have already cropped your image prior modelling for light pollution in order to avoid areas with less coverage from your stack. The solution for this is better polar alignment, better balancing and maybe pausing and carefully reentering/reframing the camera. Also take a look at the Stacking Parameters in DSS and especially what is the purpose of the "Intersection mode".

I believe that GIMP alone won't suffice for light pollution removal. Another free tool dedicated to astrophotography is Siril which is under constant development and probably will meet all your stacking and preprocessing needs. Of course there are commercial alternatives but I am sure mastering Siril will offer you great results and a great skills set! Master these and then decide if you need any filters.

Ah! And please keep it simple. One step at a time. Please avoid complicated workflows. Many astrophotography guides out there offer complicated workflows with operations which may ruin your image if you don't understand if are necessary for your dataset, and if you are not sure how to properly apply them.

Clear Skies!
Helpful Engaging Supportive
Dimitris Kavallieratos avatar
Dimitris Kavallieratos avatar
Hello again!

Thanks everyone for the replies, seems that I truly have no idea what Iam doing...

Iam going to try Siril (seems simpler than DSS) and restack! Pixinsight did a great job (is is paid as I can see,but worth it), Iam truly awed!

Tracking was not good, my first attemp at trying SA I hope it is going to get better...
andrea tasselli:
downsampling is the way to go here (by a factor of 4)

Thank you! And what you mean is...Cut down exposure time/number of frames??
Die Launische Diva:
I believe that GIMP alone won't suffice for light pollution removal. Another free tool dedicated to astrophotography is Siril which is under constant development and probably will meet all your stacking and preprocessing needs. Of course there are commercial alternatives but I am sure mastering Siril will offer you great results and a great skills set! Master these and then decide if you need any filters.

I will definitely try this! Thanks for the feedback and encouragement
andrea tasselli avatar
andrea tasselli avatar
Thank you! And what you mean is...Cut down exposure time/number of frames??


It means reduce the size of the final image by a factor of 4. This also helps a lot with noise.
Well Written
Dimitris Kavallieratos avatar
Dimitris Kavallieratos avatar
andrea tasselli:
Thank you! And what you mean is...Cut down exposure time/number of frames??


It means reduce the size of the final image by a factor of 4. This also helps a lot with noise.


Got it ! Will retry with Siril stacking/post-processing and come back with results...

Edit : I understand this is not the right subforum for this (@moderator if needed I will transfer the post), in order to achieve better balance on my SA I need a lens collar i believe, but my lens is a vintage old lens(telear-n 3.5/200mm) and i cannot find something universal that will do the job...Is there anything like that around that I can buy?

Or it is prefferable to mount my camera on a ballhead and then on to the SA (total weight around 1.3 kilos), or maybe should I try to achieve better balance with the L bracket and counterweight as is without the lens collar? 

Edit n.2 ,bestI could do with Siril. I guess not much can be done with data like these.


Thanks again
Dimitris Kavallieratos avatar
Dimitris Kavallieratos avatar
Die Launische Diva:
I did a quick light pollution removal using PixInsight's DynamicBackgroundExtraction tool in a cropped version of your image and performed a default stretch just to save as jpg and present it here and encourage you to continue with your new hobby! No denoising, no saturation boost, just an initial crop, gradient removal, and a default stretch:

Can you walk me through for these basic steps or direct me to a good tutorial? Iam playing around with the trial version of PI.

Thanks!

Edit: I got around that (with flats & PI).
Die Launische Diva avatar
Die Launische Diva avatar
Can you walk me through for these basic steps or direct me to a good tutorial? Iam playing around with the trial version of PI.

There is an excellent DBE tutorial by @Jon Rista:
https://jonrista.com/the-astrophotographers-guide/pixinsights/dynamicbackgroundextraction/
Andy Wray avatar
Andy Wray avatar
FWIW:  I think you have some nice data there.  The only obvious issues were the vignetting (which flats would get rid of) and the slight tracking issues.  Quick processed image below:


and a cropped image of pleiades:


I note that your stars are weirdly crescent shaped, so was wondering a couple of things:  What stacking method did you use on your lights (Average, Sigma Kappa clipping etc.)?  and also did you go through your 300+ images and discard all the rogue ones?
Helpful Insightful Respectful Engaging Supportive
Dimitris Kavallieratos avatar
Dimitris Kavallieratos avatar
Andy Wray:
FWIW:  I think you have some nice data there.  The only obvious issues were the vignetting (which flats would get rid of) and the slight tracking issues.  Quick processed image below:


and a cropped image of pleiades:


I note that your stars are weirdly crescent shaped, so was wondering a couple of things:  What stacking method did you use on your lights (Average, Sigma Kappa clipping etc.)?  and also did you go through your 300+ images and discard all the rogue ones?


Unfortunately I didn’t discard the bad ones, I just stacked them with DSS(standard I think the default for lights)initially and Siril as proposed with almost the same results .

Is worth it to restack with fewer selected lights and with flats and reprocess? 

What software and workflow did you use ? Great result for a quick processing !
Jonathan Aranow avatar
Jonathan Aranow avatar
My overvalued 2 cents:
Pleiades is actually a fairly challenging target to process well. Dark and reflection nebula with very bright stars is challenging to balance. Andromeda, horsehead, Orion are easier places to start. 

DSS is fine.

agree with flats- they are a must with dslr and should ideally be taken immediately after your night’s shooting so that your image plane (focus, etc) does not change.

you have to review your individual images to discard the grossly poor quality ones- some of your blotches could be from clouds in your subframes.

the dss scoring system is helpful but not completely trustworthy but can be useful to identify other subframes that should be discarded. 

another helpful trick not mentioned above is to use  the gradientxterminator (Rc-Astro.com) plug-in in photoshop. Whatch the video tutorials and it will help you salvage irregularly exposed images like this one. 

Getting back to my original comment about the challenge with the bright stars… starnet++ will help you adjust for the nebulously without overexposing the stars. Of course then you have to get comfortable with layers and masking so that may be pushing it for someone new to the sport!

Good luck!
Helpful
andrea tasselli avatar
andrea tasselli avatar
I you are using PI, assuming you went through the basic processing bits (ImageCalibration and Debayer) use the following process to create a scaling factor for each of the frames so you know which ones to keep (the good ones) and what to discard: SubframeSelector

It looks like the pic below:

In the Expression box use the following expression:

(15*(1-(FWHM-FWHMMin)/(FWHMMax-FWHMMin)) + 15*(1-(Eccentricity-EccentricityMin)/(EccentricityMax-EccentricityMin)) + 20*(SNRWeight-SNRWeightMin)/(SNRWeightMax-SNRWeightMin))+50

Then when you stack everything up use the SSWEIGHT keyword in the Weigth keyword box, as shown below:

Helpful
Andy Wray avatar
Andy Wray avatar
@Dimitris   You asked

"Unfortunately I didn’t discard the bad ones, I just stacked them with DSS(standard I think the default for lights)initially and Siril as proposed with almost the same results .

Is worth it to restack with fewer selected lights and with flats and reprocess? 

What software and workflow did you use ? Great result for a quick processing !"

I would say:  use DSS to sort your lights by score, FWHM, # stars etc and uncheck all those that look poor (star trails, unfocussed or just weird looking compared to the best ones).  You have so many frames that losing even 50 or 100 of them won't make much difference.  In that way you will end up with a sharper image with rounder stars.

Ref flats:  yes they would be worth it ... thankfully with your set-up you should be able to take flats after the event as your light train won't have changed (just make sure you manually focus on infinity which will be close enough).

Ref workflow:  I actually used StarTools to get rid of the vignetting, just because I knew how to do that quickly, used PixInsight and Photoshop for the levels/curves stuff.  To be honest though:  if you'd had the flats in there and had selected the frames, then GIMP would have done all the picture processing just fine.

If you do manage to take the flats, reject the bad images and re-stack using Kappa Sigma clipping on the lights it would be great to see the output and see what the community could make of your data.
Helpful Engaging Supportive
Dimitris Kavallieratos avatar
Dimitris Kavallieratos avatar
Andy Wray:
If you do manage to take the flats, reject the bad images and re-stack using Kappa Sigma clipping on the lights it would be great to see the output and see what the community could make of your data.


Will retry and come back with results, thank you for the reply!
Well Written Respectful
Andy Wray avatar
Andy Wray avatar
Sorry Dimitris:  I should have just clarified the flats stuff:  they should be taken at the same aperture, ISO and focus settings that you used for your lights ... I assume that means F3.5 and roughly at infinity in your case.  Easiest way would be to point the camera flat on at a laptop grey/white screen and let the camera decide on the shutter speed to end up with what should look like a vignetted grey image with maybe some dust bunnies (doughnut shadows).

If you really stretch your master flat in GIMP (curves and levels) you will probably end up with something like mine:
Helpful
Dimitris Kavallieratos avatar
Dimitris Kavallieratos avatar
Andy Wray:
Sorry Dimitris:  I should have just clarified the flats stuff:  they should be taken at the same aperture, ISO and focus settings that you used for your lights ... I assume that means F3.5 and roughly at infinity in your case.  Easiest way would be to point the camera flat on at a laptop grey/white screen and let the camera decide on the shutter speed to end up with what should look like a vignetted grey image with maybe some dust bunnies (doughnut shadows).

If you really stretch your master flat in GIMP (curves and levels) you will probably end up with something like mine:


Thank you !I believe I captured the right flats and restacked nearly half the frames and I believe end up with a good result.

Can you please proccess it? I just did an initial auto-stretch in Siril and the result was very good. Background extraction didn't work (PI or Siril) but maybe I used wrong parameters...
andrea tasselli avatar
andrea tasselli avatar
There seems to be less data than before, so here it goes:
Dimitris Kavallieratos avatar
Dimitris Kavallieratos avatar
andrea tasselli:
There seems to be less data than before, so here it goes:


Yeah there are 1 hour of data instead of around 1:50 , took only the best ```  60% of my data w/flats.

This bad light pollution current , can not be filtered I guess, right?

Die Launische Diva avatar
Die Launische Diva avatar
This bad light pollution current , can not be filtered I guess, right?

This is not light pollution. It is the result of calibrating with mismatched darks, a common problem with uncooled cameras. It wasn't that apparent at your first attempt since the larger number of lights and the randomization due to tracking issues, suppressed this problem. Again, I don't believe you need at the moment filters or a cooled camera, just some study and experimentation on your part. Also, Astrobin is a tremendous resource on your research for what is achievable with equipment and sky quality similar to yours. Use that too !
Helpful Insightful
Dimitris Kavallieratos avatar
Dimitris Kavallieratos avatar
Die Launische Diva:
This bad light pollution current , can not be filtered I guess, right?

This is not light pollution. It is the result of calibrating with mismatched darks, a common problem with uncooled cameras. It wasn't that apparent at your first attempt since the larger number of lights and the randomization due to tracking issues, suppressed this problem. Again, I don't believe you need at the moment filters or a cooled camera, just some study and experimentation on your part. Also, Astrobin is a tremendous resource on your research for what is achievable with equipment and sky quality similar to yours. Use that too !

Nice! So for now on what is the right strategy for darks ,take them after the lights or wait out for the temp to settle in?

Also about these calibration artifacts, can it be processed out?

I agree about astrobin, this is a great community, I did not expect this kind of response, thank you very much!

Edit: By mismatch you mean temperature right?I did not ask the obvious!
andrea tasselli avatar
andrea tasselli avatar
Yeah there are 1 hour of data instead of around 1:50 , took only the best ``` 60% of my data w/flats.

My recommendation is don't. Use a scaling factor as suggested instead. You're throwing out the good data together with the bad. And you're going to need all of it anyway.
Andy Wray avatar
Andy Wray avatar
andrea tasselli:
My recommendation is don't. Use a scaling factor as suggested instead. You're throwing out the good data together with the bad. And you're going to need all of it anyway.


Sorry Dimitis, I suggested throwing out the bad frames to try and understand what was causing your weird crescent-shaped small stars.  That didn't help at all, so try using the majority of your frames and only exclude the ones that have obvious major issues (doubling up of stars, major star trails etc.).  You'll need to look at each frame in DSS individually and de-select the really bad ones.
Well Written Helpful
Dimitris Kavallieratos avatar
Dimitris Kavallieratos avatar
Hello again!

Redid M45 this time with more than 4 hours of integration with 60 sec exposures at 200 ISO0(B8 no filters) with darks/bias/flats and I believe better focus & tracking than last time(definitely not worth posting).

Preprocessing in Siril (with background extraction script) and post process in PI.

Did maybe a 4-5 DBE until the gradients where(somewhat) removed but I can still see that current (like thermal noise?) and my stars are weirdly shaped not round...Is there and issue with my calibration frames or focus maybe? What else can be done processing wise?

Check the export from Siril here.
andrea tasselli avatar
andrea tasselli avatar
I've checked and it seems to me that the stars are focused OKeish. What is left is the best the lens can offer, I'm afraid, so a bit of coma, some astigamtism and longitudinal chromatic aberration. If you're going to use PI for post-processing then it would be wiser to leave ALL of it to it. Do not use the background extraction tool from Siril, just use the DBE in PI. It would be interesting to see what the output would be like without that intervention, just the raw stack-up of the  light frames from Siril with nothing done to it. I suspect you clipped the dark point samewhat there with Siril using the background extraction tool. DBE is waaaay more sofisticated and powerful than that.
Helpful Concise