Hi Rich, first I want to thank you so much for this tutorial as finaly I decide to upgrade the program to this last version, and for all the video you are doing as they are very helpful and informative for us that are using this wonderful Siril program. On regard of the workflow for comet processing, I found that the sequence you need to pick up after the preprocessing is the registration one 'r_pp_light' not the 'pp_light'. The main reason is that in the registration, the frames has the stars aligned, hence, the comet has a linear path. If you take the pp_light sequence, when is stacking the comet-only / starless-comet, the final result can be bad, with the head of the comet distorted. The reason is that the frames in the pp_light stage are not yet aligned and then the starless comet frame in between can be up or down from the initial/final frame picked for stack. The other thing I notice as other users commented here, is that in the photocolormetric part, seems there is a bug. I have also the 'white' frame issue result after to do that step and I reported in their forum. Odd thing is, if the frame where I apply the photocolormetric is called from Open (left button up) is showed perfect. If the frame is pocked from the window sequence (bottom right) is showed 'white'. The frame before or after are showing normal. Again thanks a lot for all the effort you are doing Best Regard and CS NC
Thanks! Glad the videos have been helpful. Not saying your workflow is incorrect, as there are always different ways to do things, but I mostly followed Siril's tutorial for comet stacking. siril.org/tutorials/comet/
@@DeepSpaceAstro I also had to do an initial star registration first of the sequence, because otherwise my result came out bad. I think it is because I used dithering when capturing the images, so the comet doesn't move in a regular path then. But doing the star registration first saved the day.
Thanks for the tutorial Rich. Indeed a smooth process, the only caveat is that using starnet on noisy images (single frames non stacked images) results in loss of details (e.g. little details in the comet tail). This is why I still prefer to use comet stack on all pictures first, using a magic sigma value, then clean up a bit the remaining star trails with starnet or GIMP (heal selection). Maybe is worth trying and see the difference. All the best and thanks for the high quality content you put on RUclips
Nice video Rich - always good to watch someone go through their processing, can spot the odd extra option that they may use that I don't. Keep up the great work!
Your tutorials are VERY good and professional, really grateful people such as yourself, folks behind Siril, GraXpert and lots of other free software exist out there for those of us ¨on a budget¨. A question for you, though, if you don't mind: how would one go about when dealing with a comet whose green you want to preserve, a DSO (e.g. a galaxy), a very severe gradient (light pollution) and finally, a full Moon to top of all that? I was thinking along the lines of getting 3 images (comet, DSO and stars) that can be further processed separately (Siril? PS?) and then combining them in some way (e.g. stars with DSO in Siril and comet with DSO + stars in PS). Another problem (for me at least) is also removing the green in the background (i used my trusty old Nikon) while keeping the green for the comet but also the colour for the DSO, i.e. for photometric color callibration, do i use the coordinates for the comet of the DSO?... Damn, this type of astrophotography is hard for a newbie, i am much more used to planetary imaging actually :) Thanks again for all the work you do and teach us to do ourselves!
So as you've discovered you don't want to use SCNR in an image with the comet. You may be able to mask the comet in PS and drop the green down for the rest of the image. If you can get good data in one session of the comet and Andromeda, maybe just process the stack twice, once for the comet and once for the galaxy. Break out the stars as you mentioned, and blend everything back together with PS. For light pollution, I use the Optolong L-Pro when imaging galaxies. Thanks and I'm happy to hear the videos have been helpful!
I know I have asked a lot of questions. It's because you gave such a great tutorial! I found that the Photometric Color Calibration destroyed my results. It ran successfully, and the .fit file looked find in and of itself, but the registration and stacking resulted in an image with all the data seemingly wiped. Just a black screen, even with AutoStretch invoked. Skipping the PCC step gave good results, obviously without the benefit of that step. It's a mystery to me why this happened. My camera is a ZWO ASI2600 MC pro and it supplied coordinates, so I just ran the PCC out of the box.
Thank you again for this very concise and to the point tutorial! I ended up doing the background extraction and the stretch after the starnet removal and it works just fine. I'll try if the colour calibration works then as well (yes, after saving my photometric calibrated imagge is a lot brighter than my other ones so I skipped that step); Helped me a lot with Olbers already, now stacking Tsuchinshan ATLAS (bless you! ;) )
@@DeepSpaceAstro of course photometric doesn't work without stars. But I found a good balance doing it by hand. If it just wouldn't take ages to remove the stars 😁
Thanks for this comet processing method, it's pretty cool and effective... To speed up my image processing, I use stacked biases and darks in all of my scripts
Perfect timing Rich !!! I was just going to contact you to request a new comet tutorial using the latest Siril version, so this is perfect ! I got 12P/Pons-Brooks the other night and was struggling with the previous tutorial and the new Siril and so this video will be a BIG help. Looking forward to working on it this weekend - thanks again.
My experiences with C/2023 A3: I was able to register and stack the (starless) comet images without first doing an autostretch. I then worked only on the final stacked starless and stacked starmask. Regarding the starmask, which also was stacked unstretched (as in your workflow), I put that through the regular workflow of Moffatt identification and desaturation along with everything else I would normally do with a deep sky object. I worked separately on each, stretched them separately, and then used the star recomposition for a final re-merging and tweaks at the very end. I never could get away with doing a PCC on a single image before registration. I don't suppose I can post a couple photos of that result here? One thing that always plagued me: if the head of the comet was far apart in some of the subs, I had artifacts. I had to remove quite a few subs to avoid that. And let me hasten to add, this was for the same night of imaging. The comet registration has limited capabilities. It seems to assume that everything falls in a line between the first and last comet heads that you identify. It's a simplifying step to do that. Deep Sky Stack has the user identify the location of every comet in the stacking series. In spite of that, I didn't find that DSS gives a better result than Siril. This is really hard stuff.
Great video. Just wondered at what step did color calibration replicate to all sequence? You did it on frame 29 and then continued with star removal...
@@DeepSpaceAstro thank you, just to understand it better, does it mean that calibration is only aoplied to star color or is the background also affected?
Hi Rich, great video. I'm planning for C/2023 A3 next week. What exposure length did you use for your comet acquisition? I'll only get about an hour from sunset to it dropping below the horizon.
Nice timing! If I recall you were fighting some star trailing or something similar last year with this data. I’m guessing the integration of Starnet this year helped out with that.
@@DeepSpaceAstro I thought this was a better result. I was thinking maybe the availability of Starnet apply to sequence and, of course, you’re a more bad @$$ Astrophotography processor this year!!
I ran into a little problem with still getting trailing from any remnants of the stars left over from the starnet removal. Wondering if anyone else had the same problem and how to fix it. Otherwise thanks for the vid.
I've heard a couple other people mention this last year when I put out my first video. Never did hear back on if they resolved it so I don't have an answer for you, sorry.
I seem to have the same problem. My individual starless images up to that point are pretty clean, no walking noise or star artifacts, but when I stack them, there are still slight startrails again, unless it is "walking noise" ? If I autostretch the starless stack, it is very prominently seen. I also have the issue of the one image with PCC performed on it, that shows very overexposed compared to the rest of the images. I wonder, why is the PCC done only on one of the images? Is it automatically applied to the rest of the images at some point? Just wondering. It would be great to resolve this issue, I know I am asking a lot .. LOL.. but it seems like several people are having these issues. I guess it's possible to minimize these artifacts by keeping the final image pretty dark, but it prevents one from stretching the image more to show more detail. Great video by the way! Thanks for all your tutorials! I learned a lot from them! :)
Hello my dear friend. Thanks for the detailed video. Following your guide, I ran into a problem. After adding together images of a comet from which the stars have been removed, the output is a blurred comet. It's like the program is trying to do star addition. But at the same time, I perform the addition by comet/asteroid, indicating the start and end points of the comet’s location. Please help me figure out what I'm doing wrong? Thanks in advance for your answer.
Yep. Instead of using the OSC_Preprocessing script, use this script gitlab.com/free-astro/siril-scripts/-/blob/main/preprocessing/OSC_Preprocessing_WithoutDBF.ssf?ref_type=heads from Siril. It only requires Lights, and then just follow the reset of the video.
I have followed this (wonderful!) tutorial step by step, but just like with the previous one, every time I save the photometric color calibration in the bkg_cropped_pp_light.fits image and switch to linear... I get a whole blank white image. I played with the "min/max" cursors and proceeded through the processing steps, but again, the comet_stacked.fits file is totally white 😞 (I tryed after weeks and on two different computers... same result)
@@DeepSpaceAstro After several hours on my old PC, everything seemed to work until starnet and histogram transformation. All the images in the starless comet sequence were "normal". Then comet registration, stacking... but again, comet_stacked.fit was totally white 😞
As a newbie, I am on a learning curve and wonder if you can reach down a bit and help me digest this information. I am used to background extraction and other manipulations being performed on stacked images. Can you explain why we’re doing these manipulations on the individual subs?
Following this workflow allows you to also have the files prepared for things like doing an animation, as well as stacking the comet and the stars separately.
Thank you! Can I ask another newbie question? I have some 5 Seconds subs, 15 second subs, 30, 60, and 120s subs of the recent comet. Should this workflow be applied within each of these exposure times separately and then combined together with something like pixel math? Or is there a way to combine different exposure time subs together into one stacked result?
There's no real correct answer for this. For possible issue I see is the rejection algorithm during stacking. It blends light from the same pixel position across multiple images, discarding unusually bright or dark pixels. The remaining averaged values define the final pixel value. If you consistently use the same exposure time, pixel values remain stable. However, varying exposure times result in some pixels being rejected due to their extreme brightness or dimness, altering the average and reducing overall contrast. So you can try and just put all of the images into the lights folder and stack, but it may not be what you're after. The other concern would be with your flats. Assuming those groups of images were taken on different nights where you tore down and setup, you would need a new set of flats for each group. In that case you could use SirilIC to do the stacking/combing of multiple nights/sessions.
Thanks again! So I’m thinking the best solution is either gimp or pixel math in Siril to combine five different stacked images, each of which is the result of the workflow you presented. One for each exposure level. I guess that’s how people do certain galaxies where they don’t want to blow out the core, but they also want granularity in the outer arms.
Deep sky stacker has a comet mode where you can choose in the alignment whether to focus solely on the comet or the comet plus the stars . I guess that’s a sub optimal, but easier workflow. I say easier, but for every submission, you have to click to identify where the head of the comet is. It might be interesting to compare a best effort using deep sky stacker to a best effort using Siril. I’m sure Siril will win out handily.
Hi Rich........First of thx been trying to wrap my head round Siril finally....lol Im old mate brains not good anymore....lol, anyway question why when i stack in siril using GBRG............colors are off eg: orion the reddy pard is green?
I assume you didn't mean to leave this comment on the Comet video? Have you watched my beginners tutorial? ruclips.net/video/9K-V2VIcwfQ/видео.html Is the whole image green, or only part of it? When you say you're stacking with GBRG do you mean you set that in Siril? What are you imaging with?
Nice an simple (sort off, there are a lot of places to go wrong). At 10:56, shouldn't one do a star registration, then do the star removal on r_bkg_cropped_pp_light.seq ? Otherwise you may stack a non aligned wobbly sequence and comet.
There are a number a ways to do just about anything, so you can experiment. In the video I remove the stars to stack the comet and then stack the stars. I wouldn't expect, nor I have I seen a non-aligned result doing it that way, but if you're having issues, then by all means try different approaches.
Hey Rich, great video and extremely useful! However, I am having a couple of issues following your exact process: 1) When I do the Photometric color calibration on the reference image, and I save it, it becomes really dark compared to the rest of the images! BTW, I am only using lights, would that have something to do with it? Even in autostretch mode, it is very dark 2) When is the PCC performed on the rest of the images or in the final image? I can't figure out at which step the PCC done on the single reference image is passed onto the rest of the images or the sequence 3) My stacked comet image has "streaks" on it, similar to walking noise, even though each of my starless comet images is very clean :(. Any idea what could cause that? Lack of dithering? Lack of darks/flats/biases? (BTW, I had a different DSLR image sequence with all calibration frames and it still had this weird walking noise). Any help or hints to solve any of the above issues/questions would be appreciated and thanks in advance ... I am still puzzled at how to process a comet correctly I've learned a lot from your videos!
1. Hard to say why it went dark. You can either skip color calibration of do it manually instead of using PCC. 2. It's used during registration 3. You might try not stretching as much.
Hi Rich, thanks for the reply on 1 and 2, it makes sense I am still scratching my head why the stacked image has prominent "ghost" star trails, even though my starless images look really clean The LFT algorithm seems to perform a little better than WSC, but still shows prominet ghost artifacts Yes I can stretch less but it's hard to bring out detail on the comet I took the starless, aligned images from Siril and stacked them in DSS and still see the same issue I concluded it must be that Starnet is leaving behind artifacts, even though each individual starless image seems very clean I am puzzled how people can stack 1 or 2 hrs of data on this comet (Tsuchinshan-Atlas) without getting these artifacts Do they do in Pixinsight? Is starXterminator doing a better job than Starnet 2 specifically for comets? I've been pretty happy so far with starnet on deep sky objects, but comet processing is still challenging I guess I can always stack just 4 or 8 images, but what's the point then? :) At this point, I get the best image just stacking a few images in DSS on stars (say 10x15 secs) and calling it a day. I was hoping to get more comet detail by stacking 60x15s (15 mins total), but no luck
Excellent video. Only problem I had was StarNet removed the comet from some of my images. I am stacking the images with the stars, so I’m sure that will give me bad stars. How does one keep StarNet from removing the comet?
There are no settings that I'm aware of to adjust what Starnet is doing. Are the images where the comet is being removed possibly bad data? Clouds, out of focus, etc. Sorry that's a tough one to answer.
@@DeepSpaceAstro the data is good. No cloud cover. Just couldn’t figure it out so I processed with the stars. My first comet, so every part of the process is new lol.
Good step by step tutorial. Nevertheless I failed at stretching. I did all exactly as described. But stretching only was performed on my first (reference) picture. All other 41 frames were not or wrongly stretched. My stacked frame shows a red comet core. Does anybody know why? Thanks for a hint. Kind regards, Adrian
This is great. Thanks! It worked very well when I wanted to get an image. Now I want to make all the images (pre-stacking) look nice in order to create a video (via Export Sequence). Sadly, when I try the Histogram Transflormation and hit "Apply to Sequence" it acts as if it's doing it, but all the frames in the "stretch_" sequence except my master remain unstretched. Any ideas what I might be doing wrong? (I understand this migh tbe a hard question to answer without watching me. 😁)
@@DeepSpaceAstro If I open the stretched_ sequence, then open any image in the sequence, it appears linear, unless I open the Reference Image (the one that I stretched) and it will be properly stretched. If I export the sequence as a mov file, the video is dark except for the frame containing the Reference Image, which appears properly stretched. I'll keep playing...
@@DeepSpaceAstro If I stretch the first image in the sequence and apply to sequence, it appears that all the images in the stretch_ sequence are stretched. Maybe my error was stretching the reference image and thinking it would affect the entire sequence. ??
Great tutorial Rich, thanks! Everything worked fine for me until I got to stack the starless files. I selected all (149 pics) but it failed. The process log is asking me to check image #32 but I don’t know what to look for (looks fine to me). I even removed #32 from the group, repeated stacking but it failed again. Any ideas?
@@DeepSpaceAstro Testing for understanding: Normally I apply Photometric to a stacked image, but in this case we are obliged to apply it to the sequence because of how we are registering the comet and stars independently?
Very nice Comet tutorial. I didn't realize you could use Starnet per sub, very cool. However 3 hours later, when removal completed, it looks like Starnet has removed or blurred the core of the comet in 2/3rd of my subs. 1/3rd came out fine. Subs are all about the same (no clouds or other strange things happening). Any ideas why its taking out my comet core?
@@DeepSpaceAstro I think it's just the way starnet is... Sometimes it works great, other times it leaves half of the stars in the image or zaps the cores of your galaxies away (and apparently comets)! I may just have to process this one without the per sub star removal. Thanks again for all the great tutorials!
Not sure I follow. The sequences are created that reference the frames. If you're trying to add after the fact, then you can't and need to start the process over.
Yikes. Something went wrong for me. When I stacked the comet I got a double comet image. One bright and a fainter one below it. (Note this didn't happen when I did the regular OSC stacking as I would normally do for a DSO, but half the core was black for some reason). The stacked comet image was also red. Any idea what might have gone wrong for me?
Sorry I don't. It's hard to say what happened without seeing the data. Maybe the files somehow aren't in order. You can check the file list at the step where you draw a selection around the head of the comet.
I'm new to Siril so this maybe a dumb question. Starting with the first step of folder structure, how or where do I add Dark Flat cal frames? I'm using a 294MC camera and use these rather than Bias frames. Thanks
Thanks for the very prompt reply. I'll try that, although dark flats should only be applied to flat frames, whereas bias images are, i believe, applied to all images, so I'm not sure how this may work. Ive also just discovered that the scripts require the orihinal cal gile sequences in the sub folders. For some inages i only have master cal frames (created in APP).If possible, how do i get the script to work with these? Thanks again@@DeepSpaceAstro
Biases are subtracted from the flats. If you're asking about using master calibration frames, you can do things manually, use SirilIC, or look at using my MARS suite of scripts. ruclips.net/video/87swD3grQds/видео.html
Thanks for the tut, but i have huge problem. After the photometric colorcalibration with the data from astronmetry, when i hit save and go back to linear the image becomes completely white.
Yeah and I have a video on it. I try to keep the videos to just one topic though. I don't disagree about stretching after, but I have a question. If you're saying it's wrong to stretch before stacking, in what use case would that 'stretch sequence' be used for?
Can i still stack in Siril if i dont have all calibration frames? I only have flats and bias. I have NGC 300 and comet C/2020 V2 ZTF captured last year i am keen to process. I had some troubles stacking the comet in deep sky stacker. Not sure if the comet was too faint in the subs...
You should be able to use this scipt if you don't have darsk: gitlab.com/free-astro/siril-scripts/-/blob/main/preprocessing/Mono_Preprocessing_WithoutDark.ssf?ref_type=heads
Where are the fits files which are registered to the comet? I can't seem to find them in the process folder. I need those files to go to PixInsight and take out the stars before stacking. Starnet++ in Siril does strange things like artifacts on my comet.@@DeepSpaceAstro
Thanks for this video! It really helped make sense of what to do with what the Seestar captured. It wasn't a direct translation, but pointed me in the right direction. Also, when registering the stars, it hadn't even dawned on me that with the stars frozen still, the comet would appear to move backwards due to its tail. Derp...haha. Here's what i ended up with; nothing spectacular, but decent enough. ruclips.net/user/shortssdxVJUFwwcs
Glad it was helpful. Actually the direction it's moving in your video is correct. The anti-tail points towards the sun and is the real "tail" of the comet. What we consider the tail is actually particles as it warms up from the sun. At least that's my understanding of it. Great video!
Haha. I always checkin to Rich's Siril tutorials but have to say, honestly, I agree with you. Not only does he speak fast (not helped by his keen use of very closely spaced edits you can see when he's onscreen. I always think "just leave the pause in because I think we'd all be happy with the breather) but (sorry Rich) he also has a very monotonous cadence to his voice. Almost robotic if he isn't, in fact, one of the green aliens in his intro. Great info but holy sh*t! lol 🙂
Unbelievably convenient that you put this video out the day before I get my 12/p data. Like I said before: Best tutorial creator out there. Cheers
Thanks so much!
You continue to knock it out of the park Rich. Thank you.
I appreciate that! Thanks!
Hi Rich, first I want to thank you so much for this tutorial as finaly I decide to upgrade the program to this last version, and for all the video you are doing as they are very helpful and informative for us that are using this wonderful Siril program.
On regard of the workflow for comet processing, I found that the sequence you need to pick up after the preprocessing is the registration one 'r_pp_light' not the 'pp_light'. The main reason is that in the registration, the frames has the stars aligned, hence, the comet has a linear path.
If you take the pp_light sequence, when is stacking the comet-only / starless-comet, the final result can be bad, with the head of the comet distorted. The reason is that the frames in the pp_light stage are not yet aligned and then the starless comet frame in between can be up or down from the initial/final frame picked for stack.
The other thing I notice as other users commented here, is that in the photocolormetric part, seems there is a bug. I have also the 'white' frame issue result after to do that step and I reported in their forum. Odd thing is, if the frame where I apply the photocolormetric is called from Open (left button up) is showed perfect. If the frame is pocked from the window sequence (bottom right) is showed 'white'. The frame before or after are showing normal.
Again thanks a lot for all the effort you are doing
Best Regard and CS
NC
Thanks! Glad the videos have been helpful. Not saying your workflow is incorrect, as there are always different ways to do things, but I mostly followed Siril's tutorial for comet stacking. siril.org/tutorials/comet/
@@DeepSpaceAstro I also had to do an initial star registration first of the sequence, because otherwise my result came out bad. I think it is because I used dithering when capturing the images, so the comet doesn't move in a regular path then. But doing the star registration first saved the day.
Thanks for the tutorial Rich. Indeed a smooth process, the only caveat is that using starnet on noisy images (single frames non stacked images) results in loss of details (e.g. little details in the comet tail). This is why I still prefer to use comet stack on all pictures first, using a magic sigma value, then clean up a bit the remaining star trails with starnet or GIMP (heal selection). Maybe is worth trying and see the difference. All the best and thanks for the high quality content you put on RUclips
Thanks for sharing your process, and glad to hear you enjoy the videos!
Nice video Rich - always good to watch someone go through their processing, can spot the odd extra option that they may use that I don't. Keep up the great work!
Thanks!
Your tutorials are VERY good and professional, really grateful people such as yourself, folks behind Siril, GraXpert and lots of other free software exist out there for those of us ¨on a budget¨. A question for you, though, if you don't mind: how would one go about when dealing with a comet whose green you want to preserve, a DSO (e.g. a galaxy), a very severe gradient (light pollution) and finally, a full Moon to top of all that? I was thinking along the lines of getting 3 images (comet, DSO and stars) that can be further processed separately (Siril? PS?) and then combining them in some way (e.g. stars with DSO in Siril and comet with DSO + stars in PS). Another problem (for me at least) is also removing the green in the background (i used my trusty old Nikon) while keeping the green for the comet but also the colour for the DSO, i.e. for photometric color callibration, do i use the coordinates for the comet of the DSO?...
Damn, this type of astrophotography is hard for a newbie, i am much more used to planetary imaging actually :) Thanks again for all the work you do and teach us to do ourselves!
So as you've discovered you don't want to use SCNR in an image with the comet. You may be able to mask the comet in PS and drop the green down for the rest of the image. If you can get good data in one session of the comet and Andromeda, maybe just process the stack twice, once for the comet and once for the galaxy. Break out the stars as you mentioned, and blend everything back together with PS. For light pollution, I use the Optolong L-Pro when imaging galaxies. Thanks and I'm happy to hear the videos have been helpful!
I know I have asked a lot of questions. It's because you gave such a great tutorial! I found that the Photometric Color Calibration destroyed my results. It ran successfully, and the .fit file looked find in and of itself, but the registration and stacking resulted in an image with all the data seemingly wiped. Just a black screen, even with AutoStretch invoked. Skipping the PCC step gave good results, obviously without the benefit of that step. It's a mystery to me why this happened. My camera is a ZWO ASI2600 MC pro and it supplied coordinates, so I just ran the PCC out of the box.
It's weird that happened, but I'm glad you found a workaround!
Thank you again for this very concise and to the point tutorial! I ended up doing the background extraction and the stretch after the starnet removal and it works just fine. I'll try if the colour calibration works then as well (yes, after saving my photometric calibrated imagge is a lot brighter than my other ones so I skipped that step); Helped me a lot with Olbers already, now stacking Tsuchinshan ATLAS (bless you! ;) )
Glad to hear it was helpful! Thanks!
@@DeepSpaceAstro of course photometric doesn't work without stars. But I found a good balance doing it by hand. If it just wouldn't take ages to remove the stars 😁
Thanks for this comet processing method, it's pretty cool and effective...
To speed up my image processing, I use stacked biases and darks in all of my scripts
Great tip! Thanks!
Perfect timing Rich !!! I was just going to contact you to request a new comet tutorial using the latest Siril version, so this is perfect ! I got 12P/Pons-Brooks the other night and was struggling with the previous tutorial and the new Siril and so this video will be a BIG help. Looking forward to working on it this weekend - thanks again.
Hope it helps! Thanks!
@@DeepSpaceAstro So no need to export to Photoshop/GIMP for layers and all that either then ?
@DrJ_MKE Correct. The star recomposition tool does that for you
Thanks mate. Looking forward to new tutorials. 👌
Thank you Ritch 😊
You're welcome!
Great tutorial as always Rich, many thanks. 👍👍
Glad you enjoyed it! Thanks!
Another great video!
Glad you enjoyed it! Thanks!
Tks!!! Another great video (as always)!
Thanks!
Super impressive videos Rich,thanks. I just bought a Seestar, hoping for some clear skies soon :-(
Thank you! Have fun!
My experiences with C/2023 A3:
I was able to register and stack the (starless) comet images without first doing an autostretch. I then worked only on the final stacked starless and stacked starmask. Regarding the starmask, which also was stacked unstretched (as in your workflow), I put that through the regular workflow of Moffatt identification and desaturation along with everything else I would normally do with a deep sky object. I worked separately on each, stretched them separately, and then used the star recomposition for a final re-merging and tweaks at the very end.
I never could get away with doing a PCC on a single image before registration. I don't suppose I can post a couple photos of that result here?
One thing that always plagued me: if the head of the comet was far apart in some of the subs, I had artifacts. I had to remove quite a few subs to avoid that. And let me hasten to add, this was for the same night of imaging. The comet registration has limited capabilities. It seems to assume that everything falls in a line between the first and last comet heads that you identify. It's a simplifying step to do that. Deep Sky Stack has the user identify the location of every comet in the stacking series. In spite of that, I didn't find that DSS gives a better result than Siril. This is really hard stuff.
Yeah it can be tricky some times. Sorry to hear you're having issues
I made a wonderful image of C2023 A3 - Tsuchinshan Atlas thanks to you. Appreciate your work, thank you
Awesome to hear! Thanks!
Great video. Just wondered at what step did color calibration replicate to all sequence? You did it on frame 29 and then continued with star removal...
Thanks! During the star mask registration.
@@DeepSpaceAstro thank you, just to understand it better, does it mean that calibration is only aoplied to star color or is the background also affected?
Thanks Rich!
Any time!
Hi Rich, great video. I'm planning for C/2023 A3 next week. What exposure length did you use for your comet acquisition? I'll only get about an hour from sunset to it dropping below the horizon.
Those were 60" exposures, taken with my Canon M50 Mark II at ISO 400.
Nice timing! If I recall you were fighting some star trailing or something similar last year with this data. I’m guessing the integration of Starnet this year helped out with that.
Yeah, the Python script helped with that too back then.
@@DeepSpaceAstro I thought this was a better result. I was thinking maybe the availability of Starnet apply to sequence and, of course, you’re a more bad @$$ Astrophotography processor this year!!
@AstroAF 🤣
I ran into a little problem with still getting trailing from any remnants of the stars left over from the starnet removal. Wondering if anyone else had the same problem and how to fix it. Otherwise thanks for the vid.
I've heard a couple other people mention this last year when I put out my first video. Never did hear back on if they resolved it so I don't have an answer for you, sorry.
I seem to have the same problem. My individual starless images up to that point are pretty clean, no walking noise or star artifacts, but when I stack them, there are still slight startrails again, unless it is "walking noise" ? If I autostretch the starless stack, it is very prominently seen. I also have the issue of the one image with PCC performed on it, that shows very overexposed compared to the rest of the images. I wonder, why is the PCC done only on one of the images? Is it automatically applied to the rest of the images at some point? Just wondering. It would be great to resolve this issue, I know I am asking a lot .. LOL.. but it seems like several people are having these issues. I guess it's possible to minimize these artifacts by keeping the final image pretty dark, but it prevents one from stretching the image more to show more detail. Great video by the way! Thanks for all your tutorials! I learned a lot from them! :)
Thank you for the updated guide, Do I have to process in the same way if I'm shooting with LRGB filters? I mean in the initial pre-processing stage.
Yeah you just wouldn't use the OSC Preprocesing script if you shot in mono
excellent video as always! Where can we find the scripts you have installed, like the one for stars shape fix, hubble, star trails, etc? Thanks!
I have videos for each of them. You should be able to search the channel for those names.
@@DeepSpaceAstro thanks, will do. Siril is currently removing the stars from 171 frames so, i have plenty of time :)
Hello my dear friend.
Thanks for the detailed video. Following your guide, I ran into a problem.
After adding together images of a comet from which the stars have been removed, the output is a blurred comet. It's like the program is trying to do star addition. But at the same time, I perform the addition by comet/asteroid, indicating the start and end points of the comet’s location.
Please help me figure out what I'm doing wrong?
Thanks in advance for your answer.
Sorry, hard to say. Maybe try not stretching the comet before stacking, and then stretch it during recomposition.
thank you for this great tutorial! can this be done with just lights? i am using Seestar for comet imaging. If yes, can you please show me how?
Yep. Instead of using the OSC_Preprocessing script, use this script gitlab.com/free-astro/siril-scripts/-/blob/main/preprocessing/OSC_Preprocessing_WithoutDBF.ssf?ref_type=heads from Siril. It only requires Lights, and then just follow the reset of the video.
Thanks!
Thanks for the support!
I have followed this (wonderful!) tutorial step by step, but just like with the previous one, every time I save the photometric color calibration in the bkg_cropped_pp_light.fits image and switch to linear... I get a whole blank white image. I played with the "min/max" cursors and proceeded through the processing steps, but again, the comet_stacked.fits file is totally white 😞
(I tryed after weeks and on two different computers... same result)
What does the resultxxx.fit file look like when you open and autostrech it.
@@DeepSpaceAstro all normal until that step :-(
Odd. Try skipping the background extraction. That'll affect you sequence name though, so take note of that.
@@DeepSpaceAstro After several hours on my old PC, everything seemed to work until starnet and histogram transformation. All the images in the starless comet sequence were "normal". Then comet registration, stacking... but again, comet_stacked.fit was totally white 😞
Sorry really don't know why that's happening. You can try to skip the histogram transformation step next, and then just stretch during recomposition.
супер! теперь хочется снять комету!)
As a newbie, I am on a learning curve and wonder if you can reach down a bit and help me digest this information. I am used to background extraction and other manipulations being performed on stacked images. Can you explain why we’re doing these manipulations on the individual subs?
Following this workflow allows you to also have the files prepared for things like doing an animation, as well as stacking the comet and the stars separately.
Thank you! Can I ask another newbie question? I have some 5 Seconds subs, 15 second subs, 30, 60, and 120s subs of the recent comet. Should this workflow be applied within each of these exposure times separately and then combined together with something like pixel math? Or is there a way to combine different exposure time subs together into one stacked result?
There's no real correct answer for this. For possible issue I see is the rejection algorithm during stacking. It blends light from the same pixel position across multiple images, discarding unusually bright or dark pixels. The remaining averaged values define the final pixel value. If you consistently use the same exposure time, pixel values remain stable. However, varying exposure times result in some pixels being rejected due to their extreme brightness or dimness, altering the average and reducing overall contrast.
So you can try and just put all of the images into the lights folder and stack, but it may not be what you're after. The other concern would be with your flats. Assuming those groups of images were taken on different nights where you tore down and setup, you would need a new set of flats for each group. In that case you could use SirilIC to do the stacking/combing of multiple nights/sessions.
Thanks again! So I’m thinking the best solution is either gimp or pixel math in Siril to combine five different stacked images, each of which is the result of the workflow you presented. One for each exposure level. I guess that’s how people do certain galaxies where they don’t want to blow out the core, but they also want granularity in the outer arms.
Deep sky stacker has a comet mode where you can choose in the alignment whether to focus solely on the comet or the comet plus the stars . I guess that’s a sub optimal, but easier workflow. I say easier, but for every submission, you have to click to identify where the head of the comet is. It might be interesting to compare a best effort using deep sky stacker to a best effort using Siril. I’m sure Siril will win out handily.
Hi Rich........First of thx been trying to wrap my head round Siril finally....lol Im old mate brains not good anymore....lol, anyway question why when i stack in siril using GBRG............colors are off eg: orion the reddy pard is green?
I assume you didn't mean to leave this comment on the Comet video? Have you watched my beginners tutorial? ruclips.net/video/9K-V2VIcwfQ/видео.html Is the whole image green, or only part of it? When you say you're stacking with GBRG do you mean you set that in Siril? What are you imaging with?
Nice an simple (sort off, there are a lot of places to go wrong). At 10:56, shouldn't one do a star registration, then do the star removal on r_bkg_cropped_pp_light.seq ? Otherwise you may stack a non aligned wobbly sequence and comet.
There are a number a ways to do just about anything, so you can experiment. In the video I remove the stars to stack the comet and then stack the stars. I wouldn't expect, nor I have I seen a non-aligned result doing it that way, but if you're having issues, then by all means try different approaches.
Hey Rich, great video and extremely useful! However, I am having a couple of issues following your exact process:
1) When I do the Photometric color calibration on the reference image, and I save it, it becomes really dark compared to the rest of the images!
BTW, I am only using lights, would that have something to do with it? Even in autostretch mode, it is very dark
2) When is the PCC performed on the rest of the images or in the final image? I can't figure out at which step the PCC done on the single reference image is passed onto the rest of the images or the sequence
3) My stacked comet image has "streaks" on it, similar to walking noise, even though each of my starless comet images is very clean :(. Any idea what could cause that? Lack of dithering? Lack of darks/flats/biases? (BTW, I had a different DSLR image sequence with all calibration frames and it still had this weird walking noise).
Any help or hints to solve any of the above issues/questions would be appreciated and thanks in advance ... I am still puzzled at how to process a comet correctly
I've learned a lot from your videos!
1. Hard to say why it went dark. You can either skip color calibration of do it manually instead of using PCC.
2. It's used during registration
3. You might try not stretching as much.
Hi Rich, thanks for the reply on 1 and 2, it makes sense
I am still scratching my head why the stacked image has prominent "ghost" star trails, even though my starless images look really clean
The LFT algorithm seems to perform a little better than WSC, but still shows prominet ghost artifacts
Yes I can stretch less but it's hard to bring out detail on the comet
I took the starless, aligned images from Siril and stacked them in DSS and still see the same issue
I concluded it must be that Starnet is leaving behind artifacts, even though each individual starless image seems very clean
I am puzzled how people can stack 1 or 2 hrs of data on this comet (Tsuchinshan-Atlas) without getting these artifacts
Do they do in Pixinsight? Is starXterminator doing a better job than Starnet 2 specifically for comets?
I've been pretty happy so far with starnet on deep sky objects, but comet processing is still challenging
I guess I can always stack just 4 or 8 images, but what's the point then? :)
At this point, I get the best image just stacking a few images in DSS on stars (say 10x15 secs) and calling it a day. I was hoping to get more comet detail by stacking 60x15s (15 mins total), but no luck
Any chance you can teach me how to shoot through cloud please? ;)
Weirdly im trying to image 12p pons and triangulum tonight:(
🤣 If only!
Excellent video. Only problem I had was StarNet removed the comet from some of my images. I am stacking the images with the stars, so I’m sure that will give me bad stars. How does one keep StarNet from removing the comet?
There are no settings that I'm aware of to adjust what Starnet is doing. Are the images where the comet is being removed possibly bad data? Clouds, out of focus, etc. Sorry that's a tough one to answer.
@@DeepSpaceAstro the data is good. No cloud cover. Just couldn’t figure it out so I processed with the stars. My first comet, so every part of the process is new lol.
Good step by step tutorial.
Nevertheless I failed at stretching.
I did all exactly as described.
But stretching only was performed on my first (reference) picture.
All other 41 frames were not or wrongly stretched.
My stacked frame shows a red comet core.
Does anybody know why?
Thanks for a hint.
Kind regards,
Adrian
You can try skipping applying the stretch to the sequence, and stretch it after you stack.
This is great. Thanks! It worked very well when I wanted to get an image. Now I want to make all the images (pre-stacking) look nice in order to create a video (via Export Sequence). Sadly, when I try the Histogram Transflormation and hit "Apply to Sequence" it acts as if it's doing it, but all the frames in the "stretch_" sequence except my master remain unstretched. Any ideas what I might be doing wrong? (I understand this migh tbe a hard question to answer without watching me. 😁)
Hmm yeah that's a tough one. How are you determining that they aren't stretched?
@@DeepSpaceAstro If I open the stretched_ sequence, then open any image in the sequence, it appears linear, unless I open the Reference Image (the one that I stretched) and it will be properly stretched. If I export the sequence as a mov file, the video is dark except for the frame containing the Reference Image, which appears properly stretched. I'll keep playing...
@@DeepSpaceAstro If I stretch the first image in the sequence and apply to sequence, it appears that all the images in the stretch_ sequence are stretched. Maybe my error was stretching the reference image and thinking it would affect the entire sequence. ??
Sorry I can't be much help. Let me know what you find. Meanwhile, I'll keep thinking about it.
Great tutorial Rich, thanks! Everything worked fine for me until I got to stack the starless files. I selected all (149 pics) but it failed. The process log is asking me to check image #32 but I don’t know what to look for (looks fine to me). I even removed #32 from the group, repeated stacking but it failed again. Any ideas?
Can you post the error form the console as well as a good number of the lines before and after?
How did the Photometric Color Calibration get into the whole Sequence?
During normalization when stacking.
@@DeepSpaceAstro Testing for understanding: Normally I apply Photometric to a stacked image, but in this case we are obliged to apply it to the sequence because of how we are registering the comet and stars independently?
Very nice Comet tutorial. I didn't realize you could use Starnet per sub, very cool. However 3 hours later, when removal completed, it looks like Starnet has removed or blurred the core of the comet in 2/3rd of my subs. 1/3rd came out fine. Subs are all about the same (no clouds or other strange things happening). Any ideas why its taking out my comet core?
Sorry I don't. I've heard this from a couple others as well, but have not figured out why.
@@DeepSpaceAstro I think it's just the way starnet is... Sometimes it works great, other times it leaves half of the stars in the image or zaps the cores of your galaxies away (and apparently comets)! I may just have to process this one without the per sub star removal. Thanks again for all the great tutorials!
tell me, why does Siril not create .seq sequences after adding frames? Thank you.
Not sure I follow. The sequences are created that reference the frames. If you're trying to add after the fact, then you can't and need to start the process over.
Yikes. Something went wrong for me. When I stacked the comet I got a double comet image. One bright and a fainter one below it. (Note this didn't happen when I did the regular OSC stacking as I would normally do for a DSO, but half the core was black for some reason). The stacked comet image was also red. Any idea what might have gone wrong for me?
Sorry I don't. It's hard to say what happened without seeing the data. Maybe the files somehow aren't in order. You can check the file list at the step where you draw a selection around the head of the comet.
I'm new to Siril so this maybe a dumb question. Starting with the first step of folder structure, how or where do I add Dark Flat cal frames? I'm using a 294MC camera and use these rather than Bias frames. Thanks
No dumb questions. Put them in the biases folder
Thanks for the very prompt reply.
I'll try that, although dark flats should only be applied to flat frames, whereas bias images are, i believe, applied to all images, so I'm not sure how this may work. Ive also just discovered that the scripts require the orihinal cal gile sequences in the sub folders. For some inages i only have master cal frames (created in APP).If possible, how do i get the script to work with these? Thanks again@@DeepSpaceAstro
Biases are subtracted from the flats. If you're asking about using master calibration frames, you can do things manually, use SirilIC, or look at using my MARS suite of scripts. ruclips.net/video/87swD3grQds/видео.html
Great, thanks for your help@@DeepSpaceAstro
Thanks for the tut, but i have huge problem. After the photometric colorcalibration with the data from astronmetry, when i hit save and go back to linear the image becomes completely white.
Hmmmm. Sliders below the image maybe?
Do you know you can add astrometry net inside Siril? Also, I did not understand why you do stretch images before stacking. It is always better after.
Yeah and I have a video on it. I try to keep the videos to just one topic though. I don't disagree about stretching after, but I have a question. If you're saying it's wrong to stretch before stacking, in what use case would that 'stretch sequence' be used for?
@@DeepSpaceAstro to make time lapse for example :) like in our comet tutorial;)
Can i still stack in Siril if i dont have all calibration frames? I only have flats and bias. I have NGC 300 and comet C/2020 V2 ZTF captured last year i am keen to process. I had some troubles stacking the comet in deep sky stacker. Not sure if the comet was too faint in the subs...
You should be able to use this scipt if you don't have darsk: gitlab.com/free-astro/siril-scripts/-/blob/main/preprocessing/Mono_Preprocessing_WithoutDark.ssf?ref_type=heads
I love astro... comets, python, siril and anything open source.. bro you kinda look like Cillian Murphy LOL! but great video! thanks!
Thanks, I think! 🤣
Looks like not much star residuals are left over.
Just wonder if you can't stack first and then stretch.
You can and that may actually be a better way for some
Where are the fits files which are registered to the comet? I can't seem to find them in the process folder. I need those files to go to PixInsight and take out the stars before stacking. Starnet++ in Siril does strange things like artifacts on my comet.@@DeepSpaceAstro
Why did you choose file 29 for your reference?
For faming of where I wanted the comet in the final
Thanks for this video! It really helped make sense of what to do with what the Seestar captured. It wasn't a direct translation, but pointed me in the right direction. Also, when registering the stars, it hadn't even dawned on me that with the stars frozen still, the comet would appear to move backwards due to its tail. Derp...haha. Here's what i ended up with; nothing spectacular, but decent enough. ruclips.net/user/shortssdxVJUFwwcs
Glad it was helpful. Actually the direction it's moving in your video is correct. The anti-tail points towards the sun and is the real "tail" of the comet. What we consider the tail is actually particles as it warms up from the sun. At least that's my understanding of it. Great video!
Did you speed up this recording because you are speaking very fast!
I didn't. You can slow the video slow a bit if it's too fast though.
Haha. I always checkin to Rich's Siril tutorials but have to say, honestly, I agree with you. Not only does he speak fast (not helped by his keen use of very closely spaced edits you can see when he's onscreen. I always think "just leave the pause in because I think we'd all be happy with the breather) but (sorry Rich) he also has a very monotonous cadence to his voice. Almost robotic if he isn't, in fact, one of the green aliens in his intro. Great info but holy sh*t! lol 🙂
Thanks!
Thanks so much!