This is like Van Gogh, after composing Starry Night, asking the crowd, "What do ya think of those colors?". To which the crowd replied, "Wow.....". Great job, Cuiv.
Hey Cuiv! Super excited that your super excited about this alternative process! Thanks for the plug and positive comments! I would love to hear more about your experience. Thanks again!!
Thank you!! It's a really cool process although thanks to the comments I now realize that the pixel math at the end didn't do exactly what I thought it did (basically combines the red channel from SII, green channel from Ha and blue channel from OIII) so it really is a way to calibrate those channels, it's super interesting though :)
@CuivTheLazyGeek yes its definitely an organic way of getting to a combined color starting point that isn't overwhelmed with a green bias. After processing dozens of images the most I've had to deal with is a strong magenta cast which is easily removed. The majority of the time the combined color out put is something I wanted but couldn't really achieve with scnr or normalization. I'll have a Colorized HOO video out soon as that works just as well. Thanks again!
colour pallets are all definitely subjective. i have been trying for really dark and moody images over the last 12 months. i think they look great but general feedback is that people prefer the more traditional Hubble pallet. one of the things I love about this hobby is seeing people's personal interpretations. is it scientifically accurate or helpful? probably not but it does draw people in and interest them in ways that hopefully they start looking out into the universe rather than just their phones. stop looking down and start looking up :)
Who's going to be the first YTuber that explains what its doing when you combine 3 x RGB images in this way, which is taking 9 channels and combining them to make a three channel result, it feels like this is a happy accident more than anything else as no one is mentioning what the PixelMath is actually doing.
I'm actually not even sure that's what PixelMath does - I get the feeling it may be extracting the Blue channel of the image assigned to B, etc. which makes it less of a happy accident (and would make sense since I've quickly tested the technique on multiple other narrowband images - so less of a happy accident, but maybe more of it doing something that we don't expect!). But you're absolutely right to point out the weirdness of that final step!
@@CuivTheLazyGeek Right, but it doing something we don't expect is kind of the definition of a happy accident in this context right? :) I did some quick testing myself and can confirm that all it is doing is pulling out the relevant individual channels, so for the Red it takes only the Red from the SII image, Green channel from the Ha Image and Blue Channel from the OIII. That means the processing into the three distinct RGB images isn't really doing a lot or rather not doing as much as people think. The takeaway is that it is actually doing what you might expect I guess :).
@@CuivTheLazyGeek ah ok you already figured out how it works. the result is kind of expected because by coloring the individual data to specific colors will automaticly tweak the luminance for each color channel. i am looking forward to test this on my older images. since we are covered here in clouds for months now i have plenty of time to experiment with some older data...
Those halos though!!! Edit: Good remarks: what does the PixelMath actually do?! Does it extract each channel from the colorized image before combining? Is this all a happy coincidence?? What are your thoughts on this technique? Make sure to check Entering into Space's video as well (because I'm doing it wrong and he's doing it so much better!)!! All links (also to data) in the description! Support me on Patreon!: www.patreon.com/cuivlazygeek I have a new Agena Astro affiliate link (and they have stuff in stock!): bit.ly/3Om0hNG
Looks great. I’ve got some data for this so I’m going to give it a try. EZ processing suite is no longer supported by Dark but there’s a fix on Cloudy Nights for those who want to continue using the EZ Soft stretch. Clear skies.
Flame nebula looks particularly amazing with this technique. With proper masking, we could apply different coloring to different regions. That's awesome!
Thanks, Cuiv, this is perfect timing! I just ordered one of the new Antlia ALP-T Hb/SII dual narrowband filters to use with my OSC camera. I am so excited that I will soon have (using both ALP-T filters) OIII, HA, SII, AND Hb data to play with! I am hoping you will get one of these filters, too and show us how it is done!
i just recreated this effect in photoshop. from the o3 image, only the blue channel is copied to the final image. from the s2 only the red channel, from the ha only the green channel. thats how the algorithm works with transfering rgb data to a greyscale. when matching a rgb-image to a color channel, lets say to blue, only the blue data from the rgb-image is used as a greyscale. red and green will be thrown away. in the end the magic is just getting the right stretch for each channel which is sort of automated with this worklow. very cool. thank you very much for sharing Cuiv!
I dont understand the Pixelmath at the end. Doesnt it only use the corresponding colour channel for the final Image. So most of the colourization in the red, green and blue pictures gets lost?
i just checked how pixelmath works with this kind of adressing the colour channels. My assumption was right. All the curves and adjustments in the individual olour images that are not related to the colour channel used in the combined image are useless.
Hmmm maybe it's a happy accident only - it's weird though because it does give really good results on Steve's Soul Nebula, and I also tried it on Monkey head nebula and California nebula and it gave me results I like...
Thanks Cuive, it is very interesting technique, I will try it ASAP. I fear that it would not work in the same way for each target. I can see the image from Steve (Entering into the space) and yours do look differently. I wonder if I can get those deep colors as Steve did. Thanks for the video, great video!
Completely agree - I think that final PixelMath step is far dodgier than expected. And the whole thing might be a happy coincidence (although I've tested it on other images and I did like the result)
Hi Cuiv. Great video! I use Hyperstar on my EdgeHD 11 and I have 2” Baader Ultra High speed narrowband filters that I purchased before you discovered that some were defective. I have no way to test them, and no way to know if the amount of signal I see is too low or if the targets just don’t have much if that signal in them! I wish I lived in Tokyo and could pop round to let you have a look at them with your spectrometer!!!! What do you recommend I do? I suspect that if I contact Baader they will tell me to go away!
As a very much novice OSC shooter and Pixinsight user I'm always fascinated (and confused) with what's possible in the hands of yourself, Steve and the many other talented astrophotographers on here and elsewhere. Thanks for including your data which is a great and generous thing to do for us learners. I was wondering if you used your Rising Cam OSC or a different mono camera? Thank you and clear skies!
Nice job as always! I've been trying to port this workflow over to GIMP, we don't all have pix, with some success. In that workflow I combine the colorized layers in addition mode so that each colorized layer makes its own contribution to each channel of the result. I believe that is how pixelmath would do it too. Watching the vid though I did not see you remove the mask from the Sii colorized image. If that happened it could explain the unexpected green in the final combination due to a loss of red contribution from the Sii that would have changed the green to gold. Either way it's a vibrant beautiful image.
Cuiv, so glad to see you’ve discovered Steve from Entering Into Space! I’m a huge fan of his Pix tutorials, and I find his videos very… relaxing I guess you’d say? Love to have a beer with that guy (and have him process my data). I finally got a bit of data last night on the Pelican after weeks of clouds. I’m going to give it a go with these techniques. Your enthusiasm is SO inspiring. I sure did miss you while you were on your break. Did you buy that AM5 yet? 😊
Hi Cuiv, I love your videos. Could someone with an OSC do this same technique by splitting RGB channels and just using curves on each channel to play with colors and then recombining? Also why did you use multiplication instead of addition with pixelmath to combine the stars? Thanks !!
Great tutorial. It should be possible to colour using pixel math to the correct colours. That way one should only need to create the masks and then apply the pixel math for each colour. That would save some time and still allow for tweaking the colours later.
@@CuivTheLazyGeek you should just try to use Microsoft Paint and being lazy just do video on that give up on it. Not setting a good example to us trying so hard.
Cuiv, as you have acknowledged in the comments, the content of the video is entirely wrong. Why is the video still up? I see people following this method - people who dont know much about pixelmath or image processing - and they are incredibly confused when this technique does not work the way you described it in the video. I dont understand why you would leave the content up, there isnt even a pinned comment in the description (???) I like your content, but people do come to you for advice, especially when you find out you gave out the wrong advice, please at least make the effort to stop it.
This is like Van Gogh, after composing Starry Night, asking the crowd, "What do ya think of those colors?". To which the crowd replied, "Wow.....". Great job, Cuiv.
Hey Cuiv! Super excited that your super excited about this alternative process! Thanks for the plug and positive comments! I would love to hear more about your experience. Thanks again!!
Looks like a candidate for your interviews?
@goman28 that would be cool! I literally just finished one 😉
Thank you!! It's a really cool process although thanks to the comments I now realize that the pixel math at the end didn't do exactly what I thought it did (basically combines the red channel from SII, green channel from Ha and blue channel from OIII) so it really is a way to calibrate those channels, it's super interesting though :)
@CuivTheLazyGeek yes its definitely an organic way of getting to a combined color starting point that isn't overwhelmed with a green bias. After processing dozens of images the most I've had to deal with is a strong magenta cast which is easily removed. The majority of the time the combined color out put is something I wanted but couldn't really achieve with scnr or normalization. I'll have a Colorized HOO video out soon as that works just as well. Thanks again!
colour pallets are all definitely subjective. i have been trying for really dark and moody images over the last 12 months. i think they look great but general feedback is that people prefer the more traditional Hubble pallet. one of the things I love about this hobby is seeing people's personal interpretations. is it scientifically accurate or helpful? probably not but it does draw people in and interest them in ways that hopefully they start looking out into the universe rather than just their phones. stop looking down and start looking up :)
That's very true Andrew! In the end we can all mess around with the tools provided to us and have fun!
This guy is amazing! He deserves to be supported. Thanks for all your tutorials. Be blessed.
I appreciate that!
Thanks, Cuiv. I have been really struggling with processing SHO data. This is a life saver.
Who's going to be the first YTuber that explains what its doing when you combine 3 x RGB images in this way, which is taking 9 channels and combining them to make a three channel result, it feels like this is a happy accident more than anything else as no one is mentioning what the PixelMath is actually doing.
I'm actually not even sure that's what PixelMath does - I get the feeling it may be extracting the Blue channel of the image assigned to B, etc. which makes it less of a happy accident (and would make sense since I've quickly tested the technique on multiple other narrowband images - so less of a happy accident, but maybe more of it doing something that we don't expect!). But you're absolutely right to point out the weirdness of that final step!
@@CuivTheLazyGeek Right, but it doing something we don't expect is kind of the definition of a happy accident in this context right? :) I did some quick testing myself and can confirm that all it is doing is pulling out the relevant individual channels, so for the Red it takes only the Red from the SII image, Green channel from the Ha Image and Blue Channel from the OIII. That means the processing into the three distinct RGB images isn't really doing a lot or rather not doing as much as people think. The takeaway is that it is actually doing what you might expect I guess :).
@@CuivTheLazyGeek ah ok you already figured out how it works. the result is kind of expected because by coloring the individual data to specific colors will automaticly tweak the luminance for each color channel. i am looking forward to test this on my older images. since we are covered here in clouds for months now i have plenty of time to experiment with some older data...
This looks like it is a lot of fun and bring new life to this hobby! Thanks for the awesome video and Steve for the awesome technique! Clear Skies!
Thanks Dave!
Those halos though!!! Edit: Good remarks: what does the PixelMath actually do?! Does it extract each channel from the colorized image before combining? Is this all a happy coincidence??
What are your thoughts on this technique? Make sure to check Entering into Space's video as well (because I'm doing it wrong and he's doing it so much better!)!! All links (also to data) in the description!
Support me on Patreon!: www.patreon.com/cuivlazygeek
I have a new Agena Astro affiliate link (and they have stuff in stock!): bit.ly/3Om0hNG
Awesome Cuiv I actually really like that dark green! I will definitely give this a go. Thanks
Thanks Simon!!
Looks great. I’ve got some data for this so I’m going to give it a try. EZ processing suite is no longer supported by Dark but there’s a fix on Cloudy Nights for those who want to continue using the EZ Soft stretch. Clear skies.
Flame nebula looks particularly amazing with this technique. With proper masking, we could apply different coloring to different regions. That's awesome!
Nice, I'll try it!
Thanks, Cuiv, this is perfect timing! I just ordered one of the new Antlia ALP-T Hb/SII dual narrowband filters to use with my OSC camera. I am so excited that I will soon have (using both ALP-T filters) OIII, HA, SII, AND Hb data to play with!
I am hoping you will get one of these filters, too and show us how it is done!
I'm waiting for the highspeed version of the SII + Hb one :)
@@CuivTheLazyGeek OK, I didn’t realize the high speed version hadn’t been released yet. 🙂
I really love this image of the horse head and flame!!!!❤🎉
Thank you!!!
Great job showing the technique again...I wanted to see the HSO combo, and if the green changed a lot!
Huh good idea - I need to mess around with that last step some more, see what it actually does and what can be achieved!
Kind of reminds me of the different colors I got with the Rosette. Nice, can't wait to try it.
Have fun!
Great technique - I will try it - I struggle to get Hubble style images that I like. Liked your final image.
It's definitely an interesting technique!
i just recreated this effect in photoshop. from the o3 image, only the blue channel is copied to the final image. from the s2 only the red channel, from the ha only the green channel. thats how the algorithm works with transfering rgb data to a greyscale. when matching a rgb-image to a color channel, lets say to blue, only the blue data from the rgb-image is used as a greyscale. red and green will be thrown away. in the end the magic is just getting the right stretch for each channel which is sort of automated with this worklow. very cool. thank you very much for sharing Cuiv!
I hadn't realized it worked that way, but it makes sense!
I dont understand the Pixelmath at the end. Doesnt it only use the corresponding colour channel for the final Image. So most of the colourization in the red, green and blue pictures gets lost?
i just checked how pixelmath works with this kind of adressing the colour channels. My assumption was right. All the curves and adjustments in the individual olour images that are not related to the colour channel used in the combined image are useless.
Hmmm maybe it's a happy accident only - it's weird though because it does give really good results on Steve's Soul Nebula, and I also tried it on Monkey head nebula and California nebula and it gave me results I like...
Thanks Cuive, it is very interesting technique, I will try it ASAP. I fear that it would not work in the same way for each target. I can see the image from Steve (Entering into the space) and yours do look differently. I wonder if I can get those deep colors as Steve did. Thanks for the video, great video!
Completely agree - I think that final PixelMath step is far dodgier than expected. And the whole thing might be a happy coincidence (although I've tested it on other images and I did like the result)
Hi Cuiv. Great video! I use Hyperstar on my EdgeHD 11 and I have 2” Baader Ultra High speed narrowband filters that I purchased before you discovered that some were defective. I have no way to test them, and no way to know if the amount of signal I see is too low or if the targets just don’t have much if that signal in them! I wish I lived in Tokyo and could pop round to let you have a look at them with your spectrometer!!!! What do you recommend I do? I suspect that if I contact Baader they will tell me to go away!
As a very much novice OSC shooter and Pixinsight user I'm always fascinated (and confused) with what's possible in the hands of yourself, Steve and the many other talented astrophotographers on here and elsewhere. Thanks for including your data which is a great and generous thing to do for us learners. I was wondering if you used your Rising Cam OSC or a different mono camera? Thank you and clear skies!
At that time I was using the monochrome Risingcam!
@@CuivTheLazyGeek Thanks Cuiv.
Welll that's very interesting, might have to give it a try!
It's worth a try at least!
Nice job as always!
I've been trying to port this workflow over to GIMP, we don't all have pix, with some success. In that workflow I combine the colorized layers in addition mode so that each colorized layer makes its own contribution to each channel of the result. I believe that is how pixelmath would do it too. Watching the vid though I did not see you remove the mask from the Sii colorized image. If that happened it could explain the unexpected green in the final combination due to a loss of red contribution from the Sii that would have changed the green to gold. Either way it's a vibrant beautiful image.
Cuiv, so glad to see you’ve discovered Steve from Entering Into Space! I’m a huge fan of his Pix tutorials, and I find his videos very… relaxing I guess you’d say? Love to have a beer with that guy (and have him process my data). I finally got a bit of data last night on the Pelican after weeks of clouds. I’m going to give it a go with these techniques. Your enthusiasm is SO inspiring. I sure did miss you while you were on your break. Did you buy that AM5 yet? 😊
Thanks Jacqueline - yep he does have a really laid back style :) And no I haven't bought the AM5 yet :p
Can someome explain the logic behind applying noise removal before stretching, since the later is non-destructive? Am I missing something here?
I think it it looks fantastic
Thank you!
Hi Cuiv, I love your videos. Could someone with an OSC do this same technique by splitting RGB channels and just using curves on each channel to play with colors and then recombining? Also why did you use multiplication instead of addition with pixelmath to combine the stars? Thanks !!
Late comment. I took some images of SH2-112 with Antilia C1 and C2 filters. Best way to combine?
Nice edit!
Thanks!
I like the SHO images much more than the red ones we always see...
Although the red ones of truer to life :)
Great tutorial. It should be possible to colour using pixel math to the correct colours. That way one should only need to create the masks and then apply the pixel math for each colour. That would save some time and still allow for tweaking the colours later.
That's true - also I'm worried about what PixelMath actually does at the very end
I am gonna try this workflow in PI as soon as I can get my first mono camera 🤤
Hahaha no need to force yourself to get that camera :p
@@CuivTheLazyGeek Right now I am using a 10+ years old DSLR. It's time to upgrade.
I’ve been doing this for years! I used GIMP for it! C’mon Cuiv. Thanks as ever.
Hahaha well, PixInsight can do it too, although the final PixelMath still feels dodgy :)
@@CuivTheLazyGeek you should just try to use Microsoft Paint and being lazy just do video on that give up on it. Not setting a good example to us trying so hard.
I was wondering... How would you apply this technique having only H and O data?
Now that is a good question - not sure yet!
Cuiv, as you have acknowledged in the comments, the content of the video is entirely wrong. Why is the video still up? I see people following this method - people who dont know much about pixelmath or image processing - and they are incredibly confused when this technique does not work the way you described it in the video. I dont understand why you would leave the content up, there isnt even a pinned comment in the description (???) I like your content, but people do come to you for advice, especially when you find out you gave out the wrong advice, please at least make the effort to stop it.
Don't over streach yourself.
Whaaat?
Cool technique, but I hate your blue halos…..sorry.
That's fair! Although they're not really related to the technique :) That's on the Baader filters!