I think those smudges around the prominence are from the radial blur step. I was able to get rid of the smudging by zooming into the prominences and spot healing on the subtracted layer (the gray one, e.g. at 22:48)
Thanks! Someone else mentioned that too and it looks like it's the culprit. Stared at my data for hours and couldn't figure out where the smudging was coming from. Sometimes you need a fresh set of eyes :) I'll work on that in my next rendition.
Thanks for another great video. I followed along and got a great stack from my photos. My wife and I took the rv to Batesville, AR and had great skies and temps. I used SetnC, connected to my DSLR, with a 400mm lens and 1.4 extender. Tracking with the az-gti went well. and thank you for the solar glasses!
Awesome! Glad the video was helpful. Glad your trip worked out as well. We ended up going to northern Vermont to avoid clouds. And glad to hear the solar glasses made their way to you! if you use discord and want to share your photos, the Naztronomy discord server (invite link in the description) has a section for sharing your images.
So, to summarize the process outlined in this vid: 1) Use multiple exposures (different shutter speeds covering at least 12-14 stops, or more, of dynamic range) 2) Align the multiple exposures to 'a reference layer' (usually use the fasted shutter speed exposure as the reference frame to align to) 3) Perform an HDR merge of the exposure stack - called OG in this video (this compiles the exposures to a single HDR image and increases the signal to noise ratio of the image) 4) The next "steps" essentially creates a high pass frequency separation of the HDR image. i. duplicate the HDR image - OG (place it above the HDR image in the layer stack) ii. perform a radial 'spin' blur to the duplicate layer/image (result is renamed 'blur' in this video iii. create an image in which the radial blur image ('blur') is subtracted from the HDR merge (OG) and the result is the 'subtracted' image (For those watching who use Affinity Photo: duplicate the HDR merge layer/image (OG) and perform a frequency separation on the copied layer and then discard the low frequency layer - retaining the high frequency layer. The high frequency layer is basically the same as the 'subtracted' layer in this vid) 5) Blend the 'subtracted' image into the OG image using 'multiply' as the blend mode. I would recommend trying 'overlay' or 'soft light' as the blend mode and observe the results. The 'multiply' blend mode is a blend mode that is used to darken. Hence the really dark resulting image at 12:00 6) Use your preferred techniques to adjust brightness, contrast, clarity, etc .... of the resulted image to taste. Correct me if I missed something. For the record, I employed a completely different 'HDR stacking' method that could be described as a 'manual adjusted gradient clipping merge' method for my exposures. I also added 'sky color' back into the image because, as we all know, to our naked eye: the sky is not black during a TSE- only the moon is black. 😉
Great summary! For #2, I think the 'reference' layer should be what you find the best layer to align everything else to. Not necessarily the fastest exposure. I have some stacks where the second or third fastest is the best one. No real formula there. I like the idea of adding sky color back in. Makes total sense! If you use discord and want to share your results, check out the description for an invite link to the discord channel.
At 17:33 I think that's not stacking artifacts, but because when you did the radial blur and subtract, the prominences would get smeared into a pink blur, then subtracted from white corona. White - pink = green. It might make sense to edit out the prominences before doing the blur.
@@Naztronomy Good idea here - as I commented above, I did a heal on the difference layer to address the issue, but I like the idea of removing them before the blur step. worth a try, but it might reduce the brightness or definition of them... will be interesting to try different methods.
Awesome video! It was super helpful. I ended up creating 10 stacks with each set of bracketed images (10 sets from 1/2000-2 seconds). Some of the early shots had only the peak of the prominence not the whole thing and it creates some weird artifacts when I combine the stacks. Did you have to deal with that at all?
Thank you! Yeah, I have similar issues trying to stack all of my stacks from beginning to end. You'll have to just make a big black circle in the middle to hide the imperfections at the edges because the moon moves quite a bit throughout the eclipse. So it may be better to align all of your exposures to the corona itself and not the moon/sun or the prominences. Then just overlay one good frame with the prominence on top. Aligning on the prominence is tough, especially if you had totality of more than a minute because the moon moves fast!
@@Naztronomy Thanks for the tip! I was trying to do this with all the stacks combined but I'll try it on the individual stacks and then combine it. Thanks!
Hi Naz, this is a great video adding to the collection of HDR videos out there. It is going to help people. I am going to try your techniques with my eclipse images. Thanks for posting this.
How have you resolved the green effects around the prominences? I'm having a similar problem. Also, the stacking alignment is still a bit of a challenge. I'm getting that bumpy look around the disc of the Moon even though I'm taking great care in aligning the images. Even if I stack only two images, the look is the same. Thoughts?
Near the end of the video, I mentioned that I used the clone stamp tool to fix the edges and get rid of the green effects after I was done with everything. A few others mentioned that you can use the clone stamp effect after you do the "blur" because that's what's causing the green effect. The prominences get smudged and the later steps exacerbate the smudging.
You can do a separate alignment on the moon instead of the prominence and stack that. The corona and prominences will look weird but then you can use a circular marquee tool or something to cut the moon out into a different layer and play with the settings to see what you get.
Ok, I finally figured out GIMP for those of you without PS to do the Pellet method. For each exposure, duplicate the sole layer of the original image into another layer. Select the lower layer - this is the one we will blur. In GIMP, this is "Motion blur" in the "radial" mode. Make sure the center is in the center of your inage, and set your desired angle, then apply the filter. Now select the top, duplicated layer, and change the layer mode to "Grain extract". The resulting image is exactly what you get with the apply image, subtraction, 128 offset step. Flatten and save this as your subtraction image, then repeat for all your exposures. Once done, you have two choices. One is to apply the subtraction image to each original image: open the original exposure, then add the subtraction image in a layer above it, but change the layer mode to "Grain merge". This will apply the enhanced detail to the original image, which you can then save and stack as described here. The other option is that you can layer up all of the subtraction images together, and play with what that detail is applied to. Make a new image with all the subtraction images as layers, but change their blending mode to "Grain merge". Underneath all these layers, add another layer that is one of you exposures or a basic merged HDR, and the enhanced detail will be applied to that.
Thank you! Would there be any benefit to doing the blur/subtract/multiply on each image before stacking? In your video with the 2017 data, you mentioned that you were only stacking first to demo the Pellett method quickly, and that we might want to do the Pellet method on each layer. But here you also stacked first, so now I'm second guessing the ideal order. Pellet then stack or stack then Pellet? Also, is there any advantage to utilizing your other bracketed sets? And if so, I have the same question: Pellet before or after stacking?
The benefit is that you may get finer details to appear when you apply the blur/subtract/multiply method on each image before stacking. You can stack then in two different ways. And you have it figured out (Pellet then Stack or Stack then Pellet). I think you'll end up getting similar results. I've been in touch with several people who have done it both ways and there doesn't seem to be a preference of one over the other. Doing it layer by layer is time consuming and can take hours to do so I haven't done it yet. In this video, I tried to come up with an alternate method of 'double stacking' the Pellet Method which gives me much sharper results. I've gotten good feedback from others that it works for them and it also saves a ton of time. I do want to one day sit down and do the Pellet Method on each image. I'm actually working on creating a photoshop script that'll do it for me automatically. Then I can share that with the world. It's currently not working as I'm expecting so I'm kind of stalled there. That's been my one motivation keeping me from doing it all manually for now. You can use other bracketed sets BUT you will need to align on the corona itself which is much harder to do. The moon may have moved significantly enough that aligning on it with both be difficult and negatively impact the corona. And aligning on prominences may be impossible because they may not be visible in some frames. I have not tried this but again, I hope to if I can automate some of it, otherwise it's too much manual work for me at the moment. And You can do pellet before or after stacking.
@@Naztronomy That all makes perfect sense thank you! I've spent MANY hours on this already, so I am relieved to hear that others have found that the extra time doesn't pay huge dividends. However, I'm still not happy with my results, but I think it's more a limitation of not having good images on the more exposed end. My brain short-circuited during the awesomeness of it all and I forgot to adjust my base exposure. Thankfully I was bracketing 9 frames to +/-4 stops but that only got me to 1/8 on the bright end. My darker images are great but I'm lacking the brighter ones I need to get the really big corona I'm after. It's tending to be short and stubby with a rapid fall off, rather than the long fingers. Family and friends think they're amazing, but we're our own worst critics, and therefore, must get to the next eclipse!
@@pastorjesse49 I know the feeling of your brain short circuiting. Happened to me this eclipse and I didn't take my solar filters off until almost 15 seconds later than I expected. Missed some baily's beads shots :( And also know what it's like to be your own critic. I don't like the end result of the video but others are telling me it's great 😊 If you have discord, you can share your images on the Naztronomy server (invite link is in the description). I'd love to see what you got and I'll give you honest feedback.
First of all, let me thank You for these excellent tutoriales on processing eclipse images, which are surely very benefical for all of us who want to process our images. Unfortunately I had clouds during My entire sequence of photographs of the event. However, I Will try to do the processing that You are showing us on your channel. In the case of this tutorial You performed the procedure for a single set of bracketed exposure images. When You have several sets of already processed AEB images, what method is used to combined all the sets...? Thank You very much in advance and I Will be watching your next videos for new information about how to avoid smearing in the prominences after the blur filter in Photoshop. Best regards
Thank you! And good question. If you have several sets of processed images, you can try to align the different images on the prominence itself and then do another mean stack. The challenge with using different brackets of exposures is that the moon moves a pretty significant distance from the first bracket to the last one. Especially if you're in a place with 1+ minutes of totality. So you should expect your moon to be egg shaped. The best thing to do t here would be to ignore it, align on the prominence, then superimpose one single image of the moon on top of it all. If you blend it correctly, it should look fine. I've had a few people ask me if I can show that process and I've added it to my backlog but it may take me a few weeks to do. The prominence smearing is higher on my priority list. Thanks again for the kind words and let me know if you have more questions!
I shot up in Burlington VT. I stacked two runs of approx 14 images each run and got similar results. I also went back and attempted to group 7 sets of each exposure as averages then stack those together into one image, but that method failed miserably. I'm gonna go back and try this again but im pretty happy with my first run. Cheers and thanks for sharing
Ah I need to find time to attempt to stack all my brackets as well. You have to try to align to the corona itself when You stack those. It's definitely not easy! It's also very time consuming so I'm not sure when I'll have time to do that.
I also wanted to ask you something. I didn’t bracket my shots because my camera was having issues the day before and I didn’t want to risk missing everything trying to play with the settings. But I did manually take a bunch of different exposures. When I went to stack them for HDR, the stacking was not uniform as the sun was in different alignments in each frame. Any way I can still align and do an HDR composite if the totality? Thanks
You'll have to manually align the frames. You can use the manual alignment method I show you in the video try that. You can also set the layers to like 50% opacity and try to manually align it and try to be as perfect as possible. It'll be a lot of work but totally worth it. Best of luck!
It's interesting that the shape of the corona and prominences tell us what year the eclipse photo was taken. I followed your video (the previous one) to edit my totality photo. It helped a lot. Thanks. But instead of stacking multiple photos of different exposures, I only used one, duplicated it to make several layers, and adjusted the exposure of each layer. The end result has a lot of noise, though, but I'm still trying to find a way to remove the noise.
That's an interesting way of doing it. If you want to share the photo with the community and you use discord, there's a section in the Naztronomy discord server (invite link in the description) for you to share your results.
There is no benefit of duplicating layers and using them for stacking to increase signal to noise. You can duplicate your original image a million times, adjust the exposure level of each image, and it will still have the exact MEAN (signal to noise) value after stacking as your original. If you want to benefit from stacking, each image must be unique as stacking causes the signal in your data to be reinforced while reducing noise which is random.
Trying to figure out what I have done wrong, but the file is so large (like 3Gb) that PS has to save it as a .psb and not .psd. Still working through it to see if it makes a difference as I try to play along. Not like I stacked 100 layers or anything. I did use my .NEF files (not JPG) and maybe had a dozen.
Ah wow, how big are your individual .NEF files? I'm stacking RAW files here and they're about 26MB each. My final stack file is about 1GB and as I flatten my frames, the file gets smaller.
I couldn't figure out getting GIMP to do the offset for the subtraction step, so I had to dust off a 19 year old laptop with old CS to do that (though actually that matches the age of my DSLR I captured with...). I did go back to GIMP though to stack up the resulting subtraction frames because I could leave them as live layers and play with how everything blended together.
@@Naztronomy CS6 It has all I need for my regular astrophotography needs...until today.....sighhhhh I'd love to see a Gimp example. Meant to ask you: is there a way to automatically align the Ha shots I got through the partial phases (in and out). I have nearly 800 of them and want animate them. Despite polar aligning the mount the night before, there is movement between those frames.
@@hael8680 I have CS6 as well and I just tested stacking and it works. After you load all the layers, make sure you convert it all to a smart object. Then you should have the option in Layer >> Smart Object >> Stack >> Mean. I can't test the full process in CS6 because of other compatibility issues but the stacking definitely worked. I'll post the gimp steps when I have them ready :) As for aligning your images, look into PIPP. It'll automatically center objects for you. You may have to do it in batches but it's definitely easier than doing it manually.
When I get to the point of converting to a smart object and setting the blending mode to mean, it's drastically reducing the extent of the visible corona. I do have a lot of exposures but cutting out some of the shorter exposures doesn't seem to help at all. Any suggestions?
If you're moon is not wonky looking, you're not aligning properly. You want to align to the background stars as they track similar to the corona. If you align to the moon, then your not aligning the corona.
@@hael8680 For the fastest shutter speed shots, the solar prominences will be properly exposed and you can use them for aligning. Key point being you want to stack your images with the corona aligned, otherwise you'll just blur the data from the corona after stacking. There are a couple of bright stars in the FOV for this eclipse that can be used for alignment even in the fastest shutter speeds. One if a very nice double star, Zeta Piscium which is very close to the Sun. If the background is still too dark, you can add a temporary adjustment layer to the top of your layers and increase the brightness of the background sky to make the stars more easily visible while aligning. THen when done, you can delete the adjustment layer.
@@StillaRebel2489 There are definitely stars in there if you look. There is a bright double, Zeta Piscium close to the Sun, and on the opposite side, there is a very bright star. If you zoom in and look, you'll find them. If the sky background is still too dark, you can add a temporary levels or curves adjustment layer to temporarily increase the sky background to better aid seeing these stars during alignment and then delete the adjustment layer when completed.
@dhmccauley it worked!! It worked quickly and was easy to do. Those stars reverted to a black spot in the faster shutter shot. That black spot showed up nicely when the image mode for that layer was switched to "difference ".
I think those smudges around the prominence are from the radial blur step. I was able to get rid of the smudging by zooming into the prominences and spot healing on the subtracted layer (the gray one, e.g. at 22:48)
Thanks! Someone else mentioned that too and it looks like it's the culprit. Stared at my data for hours and couldn't figure out where the smudging was coming from. Sometimes you need a fresh set of eyes :) I'll work on that in my next rendition.
Thanks for another great video. I followed along and got a great stack from my photos. My wife and I took the rv to Batesville, AR and had great skies and temps. I used SetnC, connected to my DSLR, with a 400mm lens and 1.4 extender. Tracking with the az-gti went well. and thank you for the solar glasses!
Awesome! Glad the video was helpful.
Glad your trip worked out as well. We ended up going to northern Vermont to avoid clouds.
And glad to hear the solar glasses made their way to you!
if you use discord and want to share your photos, the Naztronomy discord server (invite link in the description) has a section for sharing your images.
So, to summarize the process outlined in this vid:
1) Use multiple exposures (different shutter speeds covering at least 12-14 stops, or more, of dynamic range)
2) Align the multiple exposures to 'a reference layer' (usually use the fasted shutter speed exposure as the reference frame to align to)
3) Perform an HDR merge of the exposure stack - called OG in this video (this compiles the exposures to a single HDR image and increases the signal to noise ratio of the image)
4) The next "steps" essentially creates a high pass frequency separation of the HDR image.
i. duplicate the HDR image - OG (place it above the HDR image in the layer stack)
ii. perform a radial 'spin' blur to the duplicate layer/image (result is renamed 'blur' in this video
iii. create an image in which the radial blur image ('blur') is subtracted from the HDR merge (OG) and the result is the 'subtracted' image
(For those watching who use Affinity Photo: duplicate the HDR merge layer/image (OG) and perform a frequency separation on the copied layer and then discard the low frequency layer - retaining the high frequency layer. The high frequency layer is basically the same as the 'subtracted' layer in this vid)
5) Blend the 'subtracted' image into the OG image using 'multiply' as the blend mode. I would recommend trying 'overlay' or 'soft light' as the blend mode and observe the results. The 'multiply' blend mode is a blend mode that is used to darken. Hence the really dark resulting image at 12:00
6) Use your preferred techniques to adjust brightness, contrast, clarity, etc .... of the resulted image to taste.
Correct me if I missed something.
For the record, I employed a completely different 'HDR stacking' method that could be described as a 'manual adjusted gradient clipping merge' method for my exposures. I also added 'sky color' back into the image because, as we all know, to our naked eye: the sky is not black during a TSE- only the moon is black. 😉
Great summary! For #2, I think the 'reference' layer should be what you find the best layer to align everything else to. Not necessarily the fastest exposure. I have some stacks where the second or third fastest is the best one. No real formula there.
I like the idea of adding sky color back in. Makes total sense!
If you use discord and want to share your results, check out the description for an invite link to the discord channel.
At 17:33 I think that's not stacking artifacts, but because when you did the radial blur and subtract, the prominences would get smeared into a pink blur, then subtracted from white corona. White - pink = green. It might make sense to edit out the prominences before doing the blur.
Ah that's a great point and it makes sense. Thank you! I have more data to process so I'll edit those out and put them back in later.
Thanks!
@@Naztronomy Good idea here - as I commented above, I did a heal on the difference layer to address the issue, but I like the idea of removing them before the blur step. worth a try, but it might reduce the brightness or definition of them... will be interesting to try different methods.
Definitely worth a try! I'm also talking to someone tomorrow who worked on the original Pellett Method back in the 90s to discuss tactics.
Awesome video! It was super helpful. I ended up creating 10 stacks with each set of bracketed images (10 sets from 1/2000-2 seconds). Some of the early shots had only the peak of the prominence not the whole thing and it creates some weird artifacts when I combine the stacks. Did you have to deal with that at all?
Thank you!
Yeah, I have similar issues trying to stack all of my stacks from beginning to end. You'll have to just make a big black circle in the middle to hide the imperfections at the edges because the moon moves quite a bit throughout the eclipse. So it may be better to align all of your exposures to the corona itself and not the moon/sun or the prominences. Then just overlay one good frame with the prominence on top. Aligning on the prominence is tough, especially if you had totality of more than a minute because the moon moves fast!
@@Naztronomy Thanks for the tip! I was trying to do this with all the stacks combined but I'll try it on the individual stacks and then combine it. Thanks!
@@MrJaimec94 Anytime and good luck!
Hi Naz, this is a great video adding to the collection of HDR videos out there. It is going to help people. I am going to try your techniques with my eclipse images. Thanks for posting this.
Thanks so much, Gordon! Best of luck with your images and let me know if you have any questions or need anything!
How have you resolved the green effects around the prominences? I'm having a similar problem. Also, the stacking alignment is still a bit of a challenge. I'm getting that bumpy look around the disc of the Moon even though I'm taking great care in aligning the images. Even if I stack only two images, the look is the same. Thoughts?
Near the end of the video, I mentioned that I used the clone stamp tool to fix the edges and get rid of the green effects after I was done with everything. A few others mentioned that you can use the clone stamp effect after you do the "blur" because that's what's causing the green effect. The prominences get smudged and the later steps exacerbate the smudging.
Thank you very much for posting this very helpful video!
Glad this was helpful 😊
Is there any way to cut the moon out and stack it separately and add it back?
You can do a separate alignment on the moon instead of the prominence and stack that. The corona and prominences will look weird but then you can use a circular marquee tool or something to cut the moon out into a different layer and play with the settings to see what you get.
Ok, I finally figured out GIMP for those of you without PS to do the Pellet method. For each exposure, duplicate the sole layer of the original image into another layer. Select the lower layer - this is the one we will blur. In GIMP, this is "Motion blur" in the "radial" mode. Make sure the center is in the center of your inage, and set your desired angle, then apply the filter. Now select the top, duplicated layer, and change the layer mode to "Grain extract". The resulting image is exactly what you get with the apply image, subtraction, 128 offset step. Flatten and save this as your subtraction image, then repeat for all your exposures. Once done, you have two choices. One is to apply the subtraction image to each original image: open the original exposure, then add the subtraction image in a layer above it, but change the layer mode to "Grain merge". This will apply the enhanced detail to the original image, which you can then save and stack as described here. The other option is that you can layer up all of the subtraction images together, and play with what that detail is applied to. Make a new image with all the subtraction images as layers, but change their blending mode to "Grain merge". Underneath all these layers, add another layer that is one of you exposures or a basic merged HDR, and the enhanced detail will be applied to that.
Video! Video! Video? Please 😉👍
I guess Gimp can stack?
Thank you for this! It sounds like it is not possible to stack in Gimp? What do you mean by "sole layer"?
Thank you! Would there be any benefit to doing the blur/subtract/multiply on each image before stacking? In your video with the 2017 data, you mentioned that you were only stacking first to demo the Pellett method quickly, and that we might want to do the Pellet method on each layer. But here you also stacked first, so now I'm second guessing the ideal order. Pellet then stack or stack then Pellet? Also, is there any advantage to utilizing your other bracketed sets? And if so, I have the same question: Pellet before or after stacking?
The benefit is that you may get finer details to appear when you apply the blur/subtract/multiply method on each image before stacking. You can stack then in two different ways. And you have it figured out (Pellet then Stack or Stack then Pellet). I think you'll end up getting similar results. I've been in touch with several people who have done it both ways and there doesn't seem to be a preference of one over the other.
Doing it layer by layer is time consuming and can take hours to do so I haven't done it yet. In this video, I tried to come up with an alternate method of 'double stacking' the Pellet Method which gives me much sharper results. I've gotten good feedback from others that it works for them and it also saves a ton of time.
I do want to one day sit down and do the Pellet Method on each image. I'm actually working on creating a photoshop script that'll do it for me automatically. Then I can share that with the world. It's currently not working as I'm expecting so I'm kind of stalled there. That's been my one motivation keeping me from doing it all manually for now.
You can use other bracketed sets BUT you will need to align on the corona itself which is much harder to do. The moon may have moved significantly enough that aligning on it with both be difficult and negatively impact the corona. And aligning on prominences may be impossible because they may not be visible in some frames. I have not tried this but again, I hope to if I can automate some of it, otherwise it's too much manual work for me at the moment. And You can do pellet before or after stacking.
@@Naztronomy That all makes perfect sense thank you!
I've spent MANY hours on this already, so I am relieved to hear that others have found that the extra time doesn't pay huge dividends. However, I'm still not happy with my results, but I think it's more a limitation of not having good images on the more exposed end.
My brain short-circuited during the awesomeness of it all and I forgot to adjust my base exposure.
Thankfully I was bracketing 9 frames to +/-4 stops but that only got me to 1/8 on the bright end. My darker images are great but I'm lacking the brighter ones I need to get the really big corona I'm after. It's tending to be short and stubby with a rapid fall off, rather than the long fingers.
Family and friends think they're amazing, but we're our own worst critics, and therefore, must get to the next eclipse!
@@pastorjesse49 I know the feeling of your brain short circuiting. Happened to me this eclipse and I didn't take my solar filters off until almost 15 seconds later than I expected. Missed some baily's beads shots :(
And also know what it's like to be your own critic. I don't like the end result of the video but others are telling me it's great 😊 If you have discord, you can share your images on the Naztronomy server (invite link is in the description). I'd love to see what you got and I'll give you honest feedback.
Such a great capture! Thanks for showing us your workflow.
Thanks, Rich ☺
First of all, let me thank You for these excellent tutoriales on processing eclipse images, which are surely very benefical for all of us who want to process our images. Unfortunately I had clouds during My entire sequence of photographs of the event. However, I Will try to do the processing that You are showing us on your channel. In the case of this tutorial You performed the procedure for a single set of bracketed exposure images. When You have several sets of already processed AEB images, what method is used to combined all the sets...? Thank You very much in advance and I Will be watching your next videos for new information about how to avoid smearing in the prominences after the blur filter in Photoshop.
Best regards
Thank you! And good question.
If you have several sets of processed images, you can try to align the different images on the prominence itself and then do another mean stack. The challenge with using different brackets of exposures is that the moon moves a pretty significant distance from the first bracket to the last one. Especially if you're in a place with 1+ minutes of totality. So you should expect your moon to be egg shaped. The best thing to do t here would be to ignore it, align on the prominence, then superimpose one single image of the moon on top of it all. If you blend it correctly, it should look fine.
I've had a few people ask me if I can show that process and I've added it to my backlog but it may take me a few weeks to do. The prominence smearing is higher on my priority list.
Thanks again for the kind words and let me know if you have more questions!
I shot up in Burlington VT. I stacked two runs of approx 14 images each run and got similar results. I also went back and attempted to group 7 sets of each exposure as averages then stack those together into one image, but that method failed miserably. I'm gonna go back and try this again but im pretty happy with my first run. Cheers and thanks for sharing
Ah I need to find time to attempt to stack all my brackets as well. You have to try to align to the corona itself when You stack those. It's definitely not easy! It's also very time consuming so I'm not sure when I'll have time to do that.
Very cool! The detail you brought out is excellent! Thanks for the tutorial!
Thank you ☺
Naz, thank you for this video. I will try to manually align my non-bracketed 7-images and hope to get a good result. You're the best. Thanks man.
Best of luck 😊
I also wanted to ask you something. I didn’t bracket my shots because my camera was having issues the day before and I didn’t want to risk missing everything trying to play with the settings. But I did manually take a bunch of different exposures. When I went to stack them for HDR, the stacking was not uniform as the sun was in different alignments in each frame. Any way I can still align and do an HDR composite if the totality? Thanks
You'll have to manually align the frames. You can use the manual alignment method I show you in the video try that. You can also set the layers to like 50% opacity and try to manually align it and try to be as perfect as possible. It'll be a lot of work but totally worth it.
Best of luck!
Thank you so much.
It's interesting that the shape of the corona and prominences tell us what year the eclipse photo was taken. I followed your video (the previous one) to edit my totality photo. It helped a lot. Thanks. But instead of stacking multiple photos of different exposures, I only used one, duplicated it to make several layers, and adjusted the exposure of each layer. The end result has a lot of noise, though, but I'm still trying to find a way to remove the noise.
That's an interesting way of doing it. If you want to share the photo with the community and you use discord, there's a section in the Naztronomy discord server (invite link in the description) for you to share your results.
There is no benefit of duplicating layers and using them for stacking to increase signal to noise. You can duplicate your original image a million times, adjust the exposure level of each image, and it will still have the exact MEAN (signal to noise) value after stacking as your original. If you want to benefit from stacking, each image must be unique as stacking causes the signal in your data to be reinforced while reducing noise which is random.
Shared! Thanks!
What ISO?
All taken with ISO-200.
Trying to figure out what I have done wrong, but the file is so large (like 3Gb) that PS has to save it as a .psb and not .psd. Still working through it to see if it makes a difference as I try to play along. Not like I stacked 100 layers or anything. I did use my .NEF files (not JPG) and maybe had a dozen.
Ah wow, how big are your individual .NEF files? I'm stacking RAW files here and they're about 26MB each. My final stack file is about 1GB and as I flatten my frames, the file gets smaller.
Great tutorial. Unfortunately, my version of PS doesn't allow stacking....😢
Ahh what version do you have? I'm going to see if I can translate this workflow into Gimp and Krita soon.
I couldn't figure out getting GIMP to do the offset for the subtraction step, so I had to dust off a 19 year old laptop with old CS to do that (though actually that matches the age of my DSLR I captured with...). I did go back to GIMP though to stack up the resulting subtraction frames because I could leave them as live layers and play with how everything blended together.
I should clarify that I did a more typical Pellet method.
@@Naztronomy CS6 It has all I need for my regular astrophotography needs...until today.....sighhhhh
I'd love to see a Gimp example.
Meant to ask you: is there a way to automatically align the Ha shots I got through the partial phases (in and out). I have nearly 800 of them and want animate them. Despite polar aligning the mount the night before, there is movement between those frames.
@@hael8680 I have CS6 as well and I just tested stacking and it works.
After you load all the layers, make sure you convert it all to a smart object. Then you should have the option in Layer >> Smart Object >> Stack >> Mean.
I can't test the full process in CS6 because of other compatibility issues but the stacking definitely worked.
I'll post the gimp steps when I have them ready :)
As for aligning your images, look into PIPP. It'll automatically center objects for you. You may have to do it in batches but it's definitely easier than doing it manually.
Thank you. I can relate about the day job. lol.
When I get to the point of converting to a smart object and setting the blending mode to mean, it's drastically reducing the extent of the visible corona. I do have a lot of exposures but cutting out some of the shorter exposures doesn't seem to help at all. Any suggestions?
I just found a comet in my long shutter speed images and its not 12P.
Awesome! It's probably SOHO 5008. A recent sungrazer found which has since disintegrated. I found it in a few of my shots as well. Super cool!
Hi!! Were you able to do the circular alignment composite? Thanks.
That video is coming up this weekend. It's taking me a while to make it through my data. If only I didn't have a day job 😅
If you're moon is not wonky looking, you're not aligning properly. You want to align to the background stars as they track similar to the corona. If you align to the moon, then your not aligning the corona.
How do you see the stars in the fast shutter speed shots?
Can’t align to stars when there are no stars (except the sun) in the pictures.
@@hael8680 For the fastest shutter speed shots, the solar prominences will be properly exposed and you can use them for aligning. Key point being you want to stack your images with the corona aligned, otherwise you'll just blur the data from the corona after stacking. There are a couple of bright stars in the FOV for this eclipse that can be used for alignment even in the fastest shutter speeds. One if a very nice double star, Zeta Piscium which is very close to the Sun. If the background is still too dark, you can add a temporary adjustment layer to the top of your layers and increase the brightness of the background sky to make the stars more easily visible while aligning. THen when done, you can delete the adjustment layer.
@@StillaRebel2489 There are definitely stars in there if you look. There is a bright double, Zeta Piscium close to the Sun, and on the opposite side, there is a very bright star. If you zoom in and look, you'll find them. If the sky background is still too dark, you can add a temporary levels or curves adjustment layer to temporarily increase the sky background to better aid seeing these stars during alignment and then delete the adjustment layer when completed.
@dhmccauley it worked!! It worked quickly and was easy to do. Those stars reverted to a black spot in the faster shutter shot. That black spot showed up nicely when the image mode for that layer was switched to "difference ".