I have been binge watching these videos all day. I am just getting into the hobby and have found these videos to be incredibly useful. I am surprised that this series doesn't have more views! The pacing of these videos is just perfect. Not so fast that beginners like me can't follow nor too slow where a newbie might get bored of it. Thank you for these videos. This series deserves more recognition in the community. Subbed as I would love to see more of these videos! On another note, could you please make some tutorial style videos on using stacking softwares like Deep Sky Stacker (and what the several settings within the software mean). Maybe a video on astrophoto editing (preferably in free softwares like GIMP)? Hell maybe even a extended series just on editing. I am finding the learning curve on the editing process a bit too steep and could definitely use some guidance.
Thank you for your very encouraging comment Amol! This means a lot to me! I follow a prefixed curriculum - all your points will be covered but it will take me some time. I personally are a big fan of all open source and free to use programs, so GIMP will play a role, too! Deep Sky Stacker and all the other necessary software-parts will be covered, too. Stay tuned! Clear skies! -Chris
From sampling theory, I can say this: you need to have enough noise to use averaging to grab signals that are weaker than the sensor can collect. Essentially, what you want to avoid is being limited by quantization. Quantization in images shows up as stair stepping and is often called posterization. Samples limited by quantization are amplitude correlated, and so the law of large numbers breaks. You end up measuring the nonlinearity of the sensor, not the underlying signal. For work I’ve done in other fields, having at least an average of 4 but ideally 8-10 should be enough to eliminate the quantization linearity related issues. Note this was in a very different field, and we may have been less concerned about extremely weak signals, and more concerned about preserving dynamic range, so YMMV
I just finished up the 17 part astro tutorials. What a job well done. I'm just starting out and these videos were easy to follow and understand. Keep making more tutorials:)
Thanks a lot Andy! Your comment really means a lot to me. I'm so glad that you and other folks find this tutorial series helpful. I'll continue with the hardware chapter and stoically work my way though procedures (3) and software and processing (4).... It's a lot of work but on the same time it's so much fun! Anyway: clear skies!!
You are very welcome! Thanks for your encouraging comment. This episode was a joy to create - all the coding and thinking about the underlying principles. Have you seen the talk in the video description? It's worth a click! Clear skies!!
I now have a much better understanding of how all the data we gather in our photos is processed, thankyou. It amazes me how the little processor in our cameras can collect data for an image far greater than we can see with our eyes when looking through a telescope lens.
Thanks Andy! Yea our eyes are quite limited when it comes to deep space astro stuff :-) but why should we be equipped with DSO cameras in the first place? Hehe... Yep, the sensors are great but it's actually the whole acquisition process that really enables us to gather data. Long exposures - data addition - post process error correction... That all comes together. It's such a cool thing to do!!! Clear skies!
Great series, really educational on the subject. I am still on the process of doing research before getting my first scope and this series was a godsend. I actually went back to every single video and hit like, I never do that 😄
AND HERE COMES THE 500th. LIKE 👍😊. Yes, I am the 500th one.😀 AMAZING!!! You have explained most of the details in the easiest and most comprehensive way possible. And thanks for creating that Python script to demonstrate the mathematics of stacking the pictures. But at the same time, wondering about why such a less likes - just 499, so far!!! I wish, I can give you more likes 😄. Thanks for creating this video and sharing. BTW, subscribed ✌
I am a new entry into astrophotography and really learned from this series. I am not sure if you are still actively following the comments, and if you do, can you please provide or upload the Python codes you mentioned in this video? Will deeply appreciate the gesture. Thankyou
I actually do read all my comments :-) It's pleasure to help out and to get in contact with folks from all over the world!! Python-Code can be found here: catchingphotons.de/webspace/astro/stacking_catchingphotons.zip Have fun but be warned: The code is crude and hastily cobbled together :-D Anyway: Clear skies!!
Thank you very much! "Stacking" was my favorite video to produce so far because of the interesting math and coding behind this episode, really loved it. Clear skies!
Time stamp 9:45, take a look at the stars at your top left quarter, there is one withe star among one smaller yellow-ish to the right.... the little one (the yellow one), just move its position or is just me?... maybe the pictures where taken in very different days....
Hey Carlos! I think it's down to the rotational angel between the two images. I stacked all of the images together (de-rotated) and there were no aberrations. Those stars are way to far away to visualize the movement with my equipment is this period of time. Clear skies!!
I am a bit in doubt on your noise explanation, since your explanation means it doesn't matter if we take 120 exposures of 1 min or 120 * 60 of 1 sec exposures, since in both ways we can stack and apply large numbers law, while it seems that reality having 120 * 60 of 1 seconds exposure time will not work for very faint objects as good as 120 of 1 min exposures. I was always sure that noise in the image is caused by internal matrix constant noise level and thus large numbers won't work as effective if in every image your signal/noise ration is small, since you bring more noise than structure to the image.
You are right and in the short exposures will be a lot of noise. But for the perfect sensor the noise is external, caused by background or other factors. Important is, that this noise is random and the signal is not. So indeed the rule of large numbers applies and on average there will be more information from the target and the background will average out to a constant value that we can subtract from the image. Inpatient is that the noise is totally stochastical. There is noise however that has a pattern and comes from different sensor sources. For this noise we need either correction frames (darks or bias) or good dithering as this makes pattern noise stochastical (by randomly moving it around) and so the first point is valid again. Cheers!!
Hey Elaine, we talked a bit about that in the latest live session. Detach the camera and close the cover - than take the darks at different gain and temp settings to build a dark library. Clear skies!
Maybe you try a very simple start tracker? They'll allow you to increase the exposure time to tens of seconds and this will make all the difference. Cheers!
How do I know how much signal I have collected? I have seen people in Reddit post 15 second subs of Orion nebula and then get the running man nebula in it too. How?
You need to take the aperture of your scope, the integration time, the sensitivity of your camera, the seeing conditions, the air mass (angle of observation) and much more thighs into consideration. All those factors contribute to the actual numbers of photons you detect ~ the amount of information you connect about the observed object. You can take hundreds of 15" frames and "stack" aka "add" all the information together and unveil the nebula - or you take a few longer exposures: theoretically it doesn't matter. Reveal faint objects with a single short exposure? Unlikely - or at least only with a big aperture. Clear skies!!
I look at some pictures and think it doesn't really look like that, it's too bright. I only take 1/2 second exposures, when I do at all. Maybe I'll get into photography later, but for now I just like looking at the stars. I have other hobbies to go overboard with.😅
Thanks for the comment. Brightness though is actually not the intended goal of stacking. Stacking is there to add data. But this normally results in reducing noise, bring out faint details normally lost to the surrounding darkness of the object and increase contrast within the object itself. Thereby stacking is one of the most important tools of modern Astrophotography. Clear skies!!
Hi there, Can I shoot a video instead of photos? Can this video (videos?) been stacked as well? Is there any video of using a dslr to “catch photons”? Thanks for the valuable information. Nik
Stacking is used in both DSO (galaxies, nebula) AND planetary imaging. In DSO you stack multiple long exposures taken with DSLRs or any dedicated camera. Every sub-frame will be minutes long and together they contain very nuanced information about faint targets you then can stretch to reveal the beauty. In planetary you use small "webcams" because the planets are small anyway (so why bother with large sensors). You then take thousands of ultra-short sub-frames to gather information about the planet. But mathematically it's the same game. Cheers!
@@catchingphotons thanks. To be honest I wouldn't expect you to come back to me so quick! I haven't got any of those goto mounts (beginner). Just trying to focus unsuccesfully until today with my evostar90 on an AZ mount and a canon dslr. Quite confused with the extension ring and where to attach the camera, on the barlow or directly to the scope? I need time to try things! Stacking is another mountain to pass, as my laptop is an old hp with limited possibilities. You're inspired me and my 9yo son I have to tell. Thanks!
"we divide by the number of images to avoid brightness"- what if we want more brightness. Images like hubble's, or say a dark dessert, we do infact want more exposure so that a)we get less noise and more dynamic range and b) to get a brighter image. Then why to divide by number of images when what we want is brighter image. Also, can we get that brightness back by multiplying each pixel by number of images? I really need this for my application.
btw, loved your video. Thank you so much for making it. It helped me a lot. I do have some doubts, if possible if you have a website where I can find a mail ID to get in touch, kindly share if possible. Or, I will just try to ask them here
Hey Mate! "we divide by the number of images to avoid brightness"- what if we want more brightness." -> No. What you actually need is more information. You want to stretch the image in post process and there you need very smooth gradients between every faint detail. Thereby you need to expose hours to gather this information. Never the less. One camera pixel can only store XYZ "photons" before "filling up" (so to say). And by dividing the pixel values by the number of frames you simply drop that number. As you mentioned: it's a technical thing. No information is lost thereby. You can simply multiply the values to get the old values again. To multiply some values and keep others: basically thats stretching the image. Rewatch the video until the "roll the dice program". The program factors the values town to not overcome a given maximum. But that's technical. My screen ends there - that's all. The information value - aka the smoothness of the distribution - will increase ether way. It's just that some people say that stacking means to average out noise - but that's not the point. Stacking means to add information. The averaging is a technical process to keep the values in a given area - nothing else. Greetings !!! Chris
@@catchingphotons Thank you so much for your reply! I am fairly new to this field, although I need thorough knowledge of it for my application. I'll study more on this and surely get back to you. Clear skies!
@@charliejuliet1596 You are very welcome! Make sure to watch this video of the maker of SharpCap: ruclips.net/video/3RH93UvP358/видео.html He really knows what he is talking about :-) Greetings Chris
You can produce stunning images without any mount. Check my Andromeda just using a standard smartphone: catchingphotons.de/wp-content/uploads/2019/11/2019-08-07_M31_Andromeda_G6_iso1600_20sec_15C_WEB.jpeg and ruclips.net/video/iuMZG-SyDCU/видео.html Clear skies!
I have been binge watching these videos all day. I am just getting into the hobby and have found these videos to be incredibly useful. I am surprised that this series doesn't have more views!
The pacing of these videos is just perfect. Not so fast that beginners like me can't follow nor too slow where a newbie might get bored of it. Thank you for these videos. This series deserves more recognition in the community. Subbed as I would love to see more of these videos!
On another note, could you please make some tutorial style videos on using stacking softwares like Deep Sky Stacker (and what the several settings within the software mean). Maybe a video on astrophoto editing (preferably in free softwares like GIMP)? Hell maybe even a extended series just on editing. I am finding the learning curve on the editing process a bit too steep and could definitely use some guidance.
Thank you for your very encouraging comment Amol! This means a lot to me!
I follow a prefixed curriculum - all your points will be covered but it will take me some time. I personally are a big fan of all open source and free to use programs, so GIMP will play a role, too! Deep Sky Stacker and all the other necessary software-parts will be covered, too. Stay tuned!
Clear skies!
-Chris
Brilliant. Knowing why you do something is so much better than just knowing what to do. Great presentation skills and preparation
Thanks Matt, glad you liked it!
Thanks again Chris here’s to you!☕️
Thanks :-D
From sampling theory, I can say this: you need to have enough noise to use averaging to grab signals that are weaker than the sensor can collect. Essentially, what you want to avoid is being limited by quantization.
Quantization in images shows up as stair stepping and is often called posterization. Samples limited by quantization are amplitude correlated, and so the law of large numbers breaks. You end up measuring the nonlinearity of the sensor, not the underlying signal.
For work I’ve done in other fields, having at least an average of 4 but ideally 8-10 should be enough to eliminate the quantization linearity related issues. Note this was in a very different field, and we may have been less concerned about extremely weak signals, and more concerned about preserving dynamic range, so YMMV
Interesting!! I love looking at topics from different perspectives. Electronics, information theory, statistics... Awesome topic 🙂👍
This is incredibly well explained! Well done!
Thanks Pete!! Your comment is very much appreciated! Clear skies!!
Fantastic video thank you! Great explanation and great demonstrations 👍
Thanks bro! Glad you liked it!
Great video if you are mathematically inclined
Thanks ;-) Yep, there is some math behind everything we see...
Just finished watching all the videos in the astro tutorial playlist. Awesome content 👏
Thanks Mete for your encouraging feedback! Highly appreciated 😊
Clear skies!!
I just finished up the 17 part astro tutorials. What a job well done. I'm just starting out and these videos were easy to follow and understand. Keep making more tutorials:)
Thanks a lot Andy! Your comment really means a lot to me. I'm so glad that you and other folks find this tutorial series helpful. I'll continue with the hardware chapter and stoically work my way though procedures (3) and software and processing (4).... It's a lot of work but on the same time it's so much fun!
Anyway: clear skies!!
Fantastic video series! I watched all 18 of them. Gave me sooo much information I needed to get started. Thank You!
Thanks so much for your encouraging words!! Clear skies my friend :-)
Very clarifying. It's clear a lot of work was put into this video! But with impressive results ! Thanks a lot !
You are very welcome! Thanks for your encouraging comment. This episode was a joy to create - all the coding and thinking about the underlying principles. Have you seen the talk in the video description? It's worth a click!
Clear skies!!
I now have a much better understanding of how all the data we gather in our photos is processed, thankyou. It amazes me how the little processor in our cameras can collect data for an image far greater than we can see with our eyes when looking through a telescope lens.
Thanks Andy!
Yea our eyes are quite limited when it comes to deep space astro stuff :-) but why should we be equipped with DSO cameras in the first place? Hehe...
Yep, the sensors are great but it's actually the whole acquisition process that really enables us to gather data. Long exposures - data addition - post process error correction... That all comes together. It's such a cool thing to do!!! Clear skies!
i was waiting for a new video from you thnx :)
Thanks for your encouraging comment! Appreciate it! Greetings Chris
Great series, really educational on the subject. I am still on the process of doing research before getting my first scope and this series was a godsend. I actually went back to every single video and hit like, I never do that 😄
Thank you Dimitris for your encouraging words! Means a lot. Hope your astro-journey goes well!
Clear skies!!
Thanks for great explanation, dude!
You are very welcome! Thanks for your encouraging comment! Clear skies :-)
AND HERE COMES THE 500th. LIKE 👍😊. Yes, I am the 500th one.😀
AMAZING!!! You have explained most of the details in the easiest and most comprehensive way possible. And thanks for creating that Python script to demonstrate the mathematics of stacking the pictures. But at the same time, wondering about why such a less likes - just 499, so far!!! I wish, I can give you more likes 😄. Thanks for creating this video and sharing. BTW, subscribed ✌
Oh thank you very much for your kind words! They mean a lot to me! Cheers and clear skies!
Just discovered your channel. EXCELLENT information here.
Thank you for this encouraging comment!
Very nice video, you earned my subscription! =)
Thank you so much Edoardo!
Thanks, much appreciated
You are very welcome!!
Thank you for this video! I'm late to the game on this one...but very valuable information!
Thanks Jesse for this encouraging comment! Creating this particular video was a lot of fun (coding!)! :-) Clear skies!
excellent! just bingewatched all 17. this one was the only one I had to view x 2. wanted to know the definition of "sub"
Thanks a lot! Yea sub and light-frame or single frame are very much interchangeable.
Excellent explanation! and thank you for the recommendations.
Thanks for the encouraging comment! Clear skies!
Great video mate👍
Thanks Davide! Clear skies!
I am a new entry into astrophotography and really learned from this series. I am not sure if you are still actively following the comments, and if you do, can you please provide or upload the Python codes you mentioned in this video? Will deeply appreciate the gesture. Thankyou
I actually do read all my comments :-) It's pleasure to help out and to get in contact with folks from all over the world!!
Python-Code can be found here: catchingphotons.de/webspace/astro/stacking_catchingphotons.zip
Have fun but be warned: The code is crude and hastily cobbled together :-D
Anyway: Clear skies!!
vow!!!! Thanks a million photons!!!
Love your explanations! I learned a lot! Thank you 👌
Thanks Cindy! Glad the video was of any help! Clear skies!!
A brilliant video that really, clearly explains what stacking is, why we do it and how it works. Great work.
Thank you very much! "Stacking" was my favorite video to produce so far because of the interesting math and coding behind this episode, really loved it. Clear skies!
Time stamp 9:45, take a look at the stars at your top left quarter, there is one withe star among one smaller yellow-ish to the right.... the little one (the yellow one), just move its position or is just me?... maybe the pictures where taken in very different days....
Hey Carlos! I think it's down to the rotational angel between the two images. I stacked all of the images together (de-rotated) and there were no aberrations.
Those stars are way to far away to visualize the movement with my equipment is this period of time.
Clear skies!!
Nice and concise. Makes things a bit clearer.... see what I did there😂
Thanks for the comment! Clear skies to you!
@@catchingphotons Been waiting for clear skies since I got a laser collimator for my Newtonian just before Yule. I think it’s cursed 🤣😂🤣😂
I am a bit in doubt on your noise explanation, since your explanation means it doesn't matter if we take 120 exposures of 1 min or 120 * 60 of 1 sec exposures, since in both ways we can stack and apply large numbers law, while it seems that reality having 120 * 60 of 1 seconds exposure time will not work for very faint objects as good as 120 of 1 min exposures. I was always sure that noise in the image is caused by internal matrix constant noise level and thus large numbers won't work as effective if in every image your signal/noise ration is small, since you bring more noise than structure to the image.
You are right and in the short exposures will be a lot of noise. But for the perfect sensor the noise is external, caused by background or other factors. Important is, that this noise is random and the signal is not. So indeed the rule of large numbers applies and on average there will be more information from the target and the background will average out to a constant value that we can subtract from the image. Inpatient is that the noise is totally stochastical.
There is noise however that has a pattern and comes from different sensor sources. For this noise we need either correction frames (darks or bias) or good dithering as this makes pattern noise stochastical (by randomly moving it around) and so the first point is valid again.
Cheers!!
Amazing video! You earned a sub.
Thanks mate!
Hi Chris, great tut as usual, are there 1.13 => 1.16 episodes ? As I can’t locate them here
In the making... :-) Greetings -Chris
Good explanation. But How do you shoot dark frames and bias frames for RASA when cameras is in the front?
Hey Elaine, we talked a bit about that in the latest live session. Detach the camera and close the cover - than take the darks at different gain and temp settings to build a dark library.
Clear skies!
I dont have a motor so my exposures can be only ⅒ of a second so my photo will have only a minute of exposure
Maybe you try a very simple start tracker? They'll allow you to increase the exposure time to tens of seconds and this will make all the difference. Cheers!
Dobson 10" can be motorized?
How do I know how much signal I have collected? I have seen people in Reddit post 15 second subs of Orion nebula and then get the running man nebula in it too. How?
You need to take the aperture of your scope, the integration time, the sensitivity of your camera, the seeing conditions, the air mass (angle of observation) and much more thighs into consideration. All those factors contribute to the actual numbers of photons you detect ~ the amount of information you connect about the observed object.
You can take hundreds of 15" frames and "stack" aka "add" all the information together and unveil the nebula - or you take a few longer exposures: theoretically it doesn't matter.
Reveal faint objects with a single short exposure? Unlikely - or at least only with a big aperture.
Clear skies!!
Gem
;-)
I look at some pictures and think it doesn't really look like that, it's too bright. I only take 1/2 second exposures, when I do at all. Maybe I'll get into photography later, but for now I just like looking at the stars. I have other hobbies to go overboard with.😅
Thanks for the comment. Brightness though is actually not the intended goal of stacking. Stacking is there to add data. But this normally results in reducing noise, bring out faint details normally lost to the surrounding darkness of the object and increase contrast within the object itself. Thereby stacking is one of the most important tools of modern Astrophotography.
Clear skies!!
Hi there,
Can I shoot a video instead of photos? Can this video (videos?) been stacked as well?
Is there any video of using a dslr to “catch photons”?
Thanks for the valuable information.
Nik
Stacking is used in both DSO (galaxies, nebula) AND planetary imaging.
In DSO you stack multiple long exposures taken with DSLRs or any dedicated camera. Every sub-frame will be minutes long and together they contain very nuanced information about faint targets you then can stretch to reveal the beauty.
In planetary you use small "webcams" because the planets are small anyway (so why bother with large sensors). You then take thousands of ultra-short sub-frames to gather information about the planet.
But mathematically it's the same game.
Cheers!
@@catchingphotons thanks. To be honest I wouldn't expect you to come back to me so quick! I haven't got any of those goto mounts (beginner). Just trying to focus unsuccesfully until today with my evostar90 on an AZ mount and a canon dslr. Quite confused with the extension ring and where to attach the camera, on the barlow or directly to the scope? I need time to try things! Stacking is another mountain to pass, as my laptop is an old hp with limited possibilities. You're inspired me and my 9yo son I have to tell. Thanks!
"we divide by the number of images to avoid brightness"- what if we want more brightness. Images like hubble's, or say a dark dessert, we do infact want more exposure so that a)we get less noise and more dynamic range and b) to get a brighter image. Then why to divide by number of images when what we want is brighter image. Also, can we get that brightness back by multiplying each pixel by number of images? I really need this for my application.
btw, loved your video. Thank you so much for making it. It helped me a lot. I do have some doubts, if possible if you have a website where I can find a mail ID to get in touch, kindly share if possible. Or, I will just try to ask them here
catchingphotons.de or use my business mail linked in the channels info-description.
Greetings and clear skies!
Chris
Hey Mate!
"we divide by the number of images to avoid brightness"- what if we want more brightness."
-> No. What you actually need is more information. You want to stretch the image in post process and there you need very smooth gradients between every faint detail. Thereby you need to expose hours to gather this information. Never the less. One camera pixel can only store XYZ "photons" before "filling up" (so to say). And by dividing the pixel values by the number of frames you simply drop that number. As you mentioned: it's a technical thing. No information is lost thereby. You can simply multiply the values to get the old values again. To multiply some values and keep others: basically thats stretching the image.
Rewatch the video until the "roll the dice program". The program factors the values town to not overcome a given maximum. But that's technical. My screen ends there - that's all. The information value - aka the smoothness of the distribution - will increase ether way.
It's just that some people say that stacking means to average out noise - but that's not the point. Stacking means to add information. The averaging is a technical process to keep the values in a given area - nothing else.
Greetings !!!
Chris
@@catchingphotons Thank you so much for your reply! I am fairly new to this field, although I need thorough knowledge of it for my application. I'll study more on this and surely get back to you. Clear skies!
@@charliejuliet1596 You are very welcome!
Make sure to watch this video of the maker of SharpCap:
ruclips.net/video/3RH93UvP358/видео.html
He really knows what he is talking about :-) Greetings
Chris
“Witch” is spelled as “which”
Thanks for pointing out typos. Clear skies!
Me, without any tracking mount... :o
You can produce stunning images without any mount. Check my Andromeda just using a standard smartphone:
catchingphotons.de/wp-content/uploads/2019/11/2019-08-07_M31_Andromeda_G6_iso1600_20sec_15C_WEB.jpeg
and
ruclips.net/video/iuMZG-SyDCU/видео.html
Clear skies!
"witch...witch...witch"
...craft!