I just used this technique to stretch a the horse head. I used other techniques from other tutorials before and this is by far the best way. Other tutorials I have followed have ruined images that probably could have been much better. Funny thing is that the other tutorials are from people who are pretty well invested in the hobby. Thanks!
Man!!! Thank you SO much for sharing this info and for your good tutorials. I'm re-making all my astro images again; using this method, results just incredible! Thank you again!
So glad I ran across this video. I've watched at least a dozen videos on stretching, from all the usual astrophotography channels, and they're all variations on what boils down to an experienced astrophotographer using instinct and gut feel to adjust curves and levels. This video feels like a completely different approach. The arc-sine hyperbolic stretch demonstrated here has made a huge improvement in my images--and it's repeatable. When I was stretching by the seat of my pants my images came out different every time. With this technique I get much more consistent results and I'm not accidentally screwing up the color balance or clipping blacks and highlights. Great technique. Game changer.
Thanks, Jon, I appreciate this comment. I ran into the same experience when I first started. I never found an astrophotographer that knew what they were doing on RUclips really unless they were PI users. That lead me to creating a logical repeatable workflow that follows image processing techniques from an engineering perspective in PS. I think PI is ultimately the way to go, but I haven’t found a PI technique I cannot implement in PS. Thanks, again.
Your video was referred to me on a FB astrophotography page and I am so glad they shared it with me. I'm new to stretching data but I really like the method and results here compared to what I learned in another video for stretching data in Photoshop. Hoping you have a follow-up to this one to show how you did the remaining post-processing on the image. Thanks!
Yes, i have 2 more follow up videos to this initial video. Plus i have an entire narrowband processing tutorial videos. Just goto my youtube channel page and look under image processing tutorials. You will see all the videos there.
Just came across your video and have to reiterate what everyone else has said - hands down the most easily understood PS processing tutorial I have encountered. Like that it was not too long, and would love to see followups of processing this particular image to its final iteration so as to understand your complete workflow. Subscribed and anxiously awaiting for more like it.
Found this really helpful thanks. I like knowing the theory behind the curve, instead of eyeballing its good to know I can start from a known good initial stretch. I have a recent image of M42 I'm going to reprocess and see how much it changes the outcome.
Everyone clearly is getting great results. I'm getting a polarized mess every time and with every zoom option. I'm using subscription P.s and its all up to date. I know I have every layer set right and blend modes correct, just cant figure out whats going on. Your tutorial is easy to follow. Thanks
@@stevelloyd1374 this process works optimally for really dark skies, or dual, tri or quad narrowband filtering. What is your setup? Not sure I understand what you mean by ‘zoom’ options.
@@astroedsastrophotographych4562 Thanks for your response, I downloaded an 16 bit raw TIFF from Johnathan Eggleton that he shared on the ZWO FB page. The image was taken using a. L-extreme filter. The zoom factor I was referring to was the screen zoom in Photoshop that you mentioned could cause viewing issues. I might try one of my one images and have another play. Is there a way to share screenshot images with you? Thanks.
@@astroedsastrophotographych4562 I been trying out the curves in photoshop cc , but this technique would be the best way. I will try it out on messier 33 and messier 42.
I'm not normally one to leave comments on videos, but the information in this video is awesome. I've just followed along with my very first stacked TIF file from DeepSkyStacker and was amazed at the information I could see. (loving the ASI294, and that bit about balancing its colour cast / RGGB filter was very helpful) So much more detail than just a simple edit in Lightroom and I can now see what you mean about 'star bloat' and 'colour bleaching'. Have now subscribed and plan to watch the rest of this series. Is there a way to follow them along in order of the editing process? I couldn't see if they were numbered in some kind of order. Thanks so much for the info and the effort in making these tutorials. ( lol , 42 comments...)
Thnks, Matt. Glad it has helped. If you look at tue playlist for image processing tutorials I have only 3. This is the first in the series. Next is starnet++ followed by why contrast is important. They all use the same image until a final image is produced. Note that this isn’t a narrowband tutorial the final video just tries to maximize the data thru detail enhancement, mainly focusing on the idea of contrast. Thanks again! 42 seems like your number 😀
Great great video. I've been doping a landscape wide astrophotography, got my tracker few weeks ago, and I wonder if this technique would work well on wide lens shots.... Will try for sure.
It works best in combination with dark skies, dark frames, and flats. If one of those isn’t a part it could look off. You can also substitute dark skies with a narrowband filter. Clear skies.
Nice video. Well explained and demonstrated. One comment, though. Can’t you complete this process in 32 bit mode? Some desirable data deep in shadows is being compromised by the early 16-bit conversion, while it looks like these steps should be available in Photoshop’s 32 bit mode. Normally the 16-bit conversion would be made after completely the initial stretching (i.e. what is done in this video).
Great results! However keep in mind, according to Mark Shelly's site, Photoshop is not capable of doing a hyperbolic arcsin due to restrictions on the plotted points. His workaround was to create functional square roots of the desired curves. Two stretches must be performed back to back so they multiply to a true hyperbolic. Mix and matching the various 10x, 30x, 100x... in order to get the stretch factor you need. Im not sure stretching thrice will give you a true hyperbolic either. Nevertheless, you gave a great intro into Mark's technique. Thanks.
Thanks, appreciate the feedback. I usually do two to three stretches (eye balling it) but good to know two stretches with varying multipliers is the correct approach to a true hyperbolic stretch! Thanks, Stephen! Clear skies.
A how to by you. Really cool. This was the most interesting thing I stumbled upon Today. I love to see how folks work. Nice you quote your source. Forest Tanaka was my biggest influence. Who is yours?
Thanks, Garnett. Honestly, probably my biggest influence is Chuck Ayoub. He has always motivated me to strive for the best and grow, through observation of his example.
If you use a newtonian telescope you’ll get these diffraction spikes due to the spider vanes which hold the secondary mirror. If you’re using a refractor telescope you will not get these. Hope this answer helps.
You're videos have been amazing and like others have said they are the most well explained I've seen, thank you so much! So I'm a newbie and have a 1600mm, how can I incorporate you're tutorials into a mono workflow? Should I simply assign my Ha, OIII and SII to my choice of Channels and pick up your video from there?
ok looks great, but would nice to see your final image as well. I have used the download arcsinh stretches, but not in this method, and especially like your meticulous color balancing. I also just got a triad ultra, so will be using that in a Bortle 9. My question is, if you have PixInsight, and do the standard stretch there...will that give a different result than doing how you did it, especially with the grey multiply/divide layers ? or is this a Photoshop "equivalent " to a stretch in PixInsight? thanks for this video!
Thanks and thanks for watching, Greg. I plan to release more videos on my workflow that will show this image progress. I didn’t want to release a super long video no one would watch, so stay tuned for more. As for pixinsight, I’m not a user myself but do know that there is a separate module for doing arcsinh stretch that Mark Shelly wrote in Pixinsight, difference is you don’t need the grey mult/divide layers you can apply the stretch much simpler in PI. See his website for reference, he talks about it, link in the description. So I believe the standard stretch is not the same but could be wrong.
Thanks for a terrific video. One question from a newbie....Is step1, the initial color balance, necessary?, Since step 2 is to bring all the channels down to the leftmost limit without clipping, wouldnt the histograms get aligned automatically just by doing step2?
My initial thoughts agree with you. What I plan to do is what you suggest and do a empirical comparison between the two and see if there is any difference. If there is I will respond back to this post eventually. Thanks for watching!
Hi Ed I have been using your method and workflow now on all my images, it’s fantastic. Gives so much more depth and colour, hard to believe that it was there waiting to be discovered. One question - on the initial channel mix alignment - I am finding that most images are too dark after stacking and all the colours are right on the left-hand side of the histogram without any differentiation between them. So I have been doing some initial curve stretching so I can separate the colours, usually after a couple of stretches the colour channel differences become apparent and I can start aligning them. Is this the correct approach or is there a better way?
Hi Ray, Usually the only reason my channels are seperated initially is because my OSC has an RGGB sensor and thus green channel pulls ahead. I only align the channels because of that. Are you using a DSLR? Also filters can change this response. Seems odd that you would have to separate tue channels and realign them. If they are already aligned I would think that is where you would start the arcsinh stretch from...
Hi Ed I am using a OSC ZWOASI2600 and a Optolong L-Extreme filter. I find that a couple of slight curves stretches separates the channels enough to get going, no real issues
@@astroedsastrophotographych4562 so when this happens I should proceed straight to the next stage (grey multiply/Grey divide). Should I do any channel alignments after the first arc sine curves stretch if they are out of wack? (Sorry for all the questions, please remember you I am a complete novice!). Rgds Ray
Great tutorial. It has helped with a lot of my targets and I have re-processed them. However, it seems to fail dramatically on my M42 orion images. Any clue to why?
Were you in dark skies, or use a filter? Else it isn’t the technique but rather the light pollution issue. Also I assume you do flats and dark frame correction? Else that would be another issue.
@@astroedsastrophotographych4562 I am in bortal 5. Half of my shots were without L-Pro filter and the rest were. Most were on an ASI533 some with a DSLR and the rest ASI1600MM. Yes Darks and Flats are shot.
@@NitescapeAstro sounds like you were stretching light pollution data at bortle 5. Bortle 3 or below you could probably get away with it or using filters.
Thank you for the video. I am new to Astrophotography and just purchased a monthly subscription to PS. How do I load the unzip the file to PS which is web based? The instructions for loading the file assumes the program is physically on your computer. Thank you for your consideration in this matter
Very interesting! I’m eager to try this on some data, but my images from my canon 6D tend to have a red glow and the red histogram leads the other two, would I want to boost the red in my green and blue channels instead to align the histograms?
Hi Ed. I am using this technique and, after a series of stretches, I can see color in my stars but the nebulosity still looks grayscale. It is a reflection nebula so I am expecting some blue hue. I am zoomed in beyond 67% so that's not the problem.
I'm using an Edge HD8 at f/7, IR/UV filter, ZWO ASI071 MC, and 5 hours total integration. I processed this target last year not using this technique and the color shows just fine. I was going back to this data and re-processing it using the color-preserving technique when the lack of color problem arose. I really like the idea of preserving star color and not blowing them out so I am hoping I can get this color-preserving technique to work!
@@robertstewart1466 i have the same camera, and the same problem. When stretching the multiple channel, the reds of nebula are very pale, not B/W, but very pale. Can't seem to figure out how he gets the normal color when stretching that layer.
OK, I'm with you up through stamp and desaturate the Grey multiply layer. I then stamp and add the Grey divide layer. Change the blend modes to Divide and Multiply. However when I begin stretching (arc sin) the image remains in grey scale. Yours doesn't. I have started this over about 5 times and the same thing happens. Suggestions? I know its something small I'm missing. Thanks for great instructional videos.
Also I would check the image you have prior to the mult and divide layers. Make sure that is the color version and you didn’t set it to grey on accident.
This is a great video, i now have added these to the list of things to use for astroprocessing in PS. One question, I shoot mono with filters.. So, would it be best to stretch before I combine or after in this instance lrgb, and what does this mean for narrowband imaging?.. As far as how to utilize these with false color? Tips or hints?
That was interesting. But my question is, how much sense does this way of stratching make, when I remove the stars with Starnet ++ before stretching? What I understand is that it is about preventing stars from blowing out. So is there an other reson to use this stratching procedure?
Edit: You dont remove the stars with starnet first, you stretch first then use starnet. The whole point is to stretch your data and stars as best as possible without blowing out the stars or over-stretching and bleaching your data. Then you use starnet, then with your starless image continue processing your data and recombine the stars later that you extracted from the first arc hyperbolic sine stretch.
@@kowalskik.1333 you should always stretch first before using starnet, else you will not be able to stretch your data properly afterwards. Even the instructions that come with starnet say this.
@@astroedsastrophotographych4562 OK agree, but my frame comes already pre stretched from APP (Astro Pixel Prozessor) to Photo Shop. Should I make sure APP is not stratching at all?
@@kowalskik.1333 i would compare app stretching to the method I show here, and then use what you think is better on your data. So for this method, don’t put your data thru app.
Thank you for the video! I tried to apply this method on my 1.5h integration of Heart and Soul in Bortle 4-5 sky on 135mm astromodified DSLR with moderate twilight pollution. ArcSinh probably (not sure yet) does a better job at stretching (final result visually looks slightly beter), comparing to simple black and midpoint Levels adjustment. However, the actual introduction of Grey Multiply and Divide method made the final image much worse. The colors look desaturated, and almost like washed out into the grey. I am not sure if this method is just more scientifically correct (more realistic color rendering), while visually less appealing??? Or is my data just not good enough for this method to show its potential? Note that my original stack shows substantial color cast and gradient due to twiligh pollution. Do I need to remove the color cast and gradient first, before applying the grey×÷ method? UPD: Tried this method AFTER removing the background. The result is the same - the image looks bleak, almost colorless, way less vivid and appealing than if no grey×÷ was used. Whats going on?
This technique works optimally on good data. Unfiltered data is horrible and corrupted (as you say bortle 4-5, with a gradient), you are extracting dominant signal which is light pollution it sounds like in your case. You need a good narrowband filter or super dark skies. Your problem has nothing to do with this technique rather your equipment/light pollution issues. Even using background removal techniques will not resolve this issue, because you truly have pixels that are still a mix of light pollution with signal.
Thank you for the reply! Will try to improve my data! Also, will try to apply this technique on some of the beter unstretched data available online to compare the result. Thanks
Hey Ed follow up question on this does it matter in which way you channel mix, for example the way my histogram was aligned, I had to drop the Green in my Blue channel. I'm guessing this is still fine?
I’d say let your result guide you. Really no wrong way unless it degrades your image quality. If that is your only option then you have no choice. If you have another option, I would do a study where you do both ways and see which results you like best or maybe they would even turn out equivalent. 😊
Can this process be completed with an image from L+RGB data? Once stacked in DSS, I normally use curves and levels on each colour and then combine them and then add the Lum. I find this method works great on a DSLR image.
I following this technique alswell i am using an Atik 383L mono, now I did the RGB images using this but What should i do with my Luminance? i mean how to add that to it? or is the multiply or divide layer the actual luminance layer? as i shoot with LRGB filters.
Ah yea but then i shouldnt crop the RGB before feeding it to starnet right? Becouse otherwise the Lum wouldnt align anymore with the RGB isnt it? Or should i add the lum before feeding it all to starnet?
This because of a ps bug. Zoom past 67% and you will see it works. I explained this in the video. You must have skipped around and not watched it all, as I explain all this, after the stretching technique the next step combines all this and you no longer need to zoom in to see correct details.
I use multiple (2 or 3) smaller arcsinh stretches so I don’t over stretch my data, there is a certain point which the data cannot be pushed past without bleaching (pushed too close in luminance to the white point). It gives me finer control in stretching.
Hi there, not to my knowledge. There is this option in PI, I just implemented it in PS. I could code up an action eventually, but if I did I would sell it as my own action.
I tried this and I ended up with a stretched black/white image instead of color. I don't use these keyboard combos and I have no idea which layer you copied over to the top, at the end.
Dear friend, thank you for sharing this. Could it be that in the third Arsin arc 10 I already have more swelling of stars and more saturated around them? I have tried several times and I cannot get the stars to remain unchanged with the initial image
@@astroedsastrophotographych4562 First of all thanks for your answer I use a Nikon D 750 exposure 48 sec focal length 180 mm askar 180 iso 500 50 light 50 darks 30 ways no flat
@@astroedsastrophotographych4562 it's the latest version of Photoshop CC. Yes I followed the instructions. I unzipped and double click to load a sinh and then PS says wrong kind of document. I don't get it. It's and acv file and I made a scratch curve like he said to see where it is being saved. My curve is there but it will not let me download these files in that. I don't know man I'm about over this Astro stuff. Nothinbg ever goes the way it's supposed to.
Not sure, it shoukd work with latest version of photoshop CC, that is what I’m using too. AP is hard and those who truly want to be successful will find away. If you’re not liking this hobby, that is ok, it isn’t for everyone. I enjoy debugging problems until i find solutions, and AP is full of that, it gets boring when everything runs to smoothly. 😊
Hi. Does anyone know if these work on a Mac? I've installed them in what I think is the right place but they don't show up in the Preset Manager to load. Any help much appreciated.
yep, bleaches every time for me. I've tried about everything, different zoom levels, etc. and when I stretch the multiply later, it's just bleached red. ASI071MC Pro, Rosette nebula that should be very red.
I have one image that has the main data towards the middle of the levels adjustment panel. There is also (height wise very small) a lot of data to the left of the main spike. Do you clip only up to the small data or go up to the main spike? It looks kinda like this " _____IIIIII_____"
I do multiple small level adjustments up the small amount of data to the left of the main spike. Continually doing these small levels you will eventually or should see the those two spikes merge into one. At least that is what happens for me.
Super agradecido por lo que nos traes aquí ya lo estoy probando! Mi único problema es que no tengo mucha información de negros a pesar que hice un apilado en dds y use 45 lights de 1 minuto tal vez debería hacer muchos más darks de los que hice o más flats ? Allí si tendría más información de los negros? Gracias por todo
I downloaded the arcsinh files. I put them in the same curves folder mentioned, but when I select Curves in PS, they are not there to select. Am I doing something wrong?
Hi I tried to send you a message on your website contact form. Don't think it went through. Basically I followed your technique using the arcsin10 adjustment but am not getting the same results. I use a DSLR and telephoto lens on a Star tracker and take about 50 to 100 lights and stack them to start my image in photoshop. My question is will your arcsinh10 adjustment process outlined in your Photoshop videos work for my DSLR image or is that for sophisticated telescope cameras that take color images in different channels?
Next question. When you say you’re not getting the same results, what do you mean? Did you shoot the same target as me with the triad ultra filter? If you have light pollution and no narrowband filters your results won’t be limited by my stretching technique but rather by your light pollution.
@@astroedsastrophotographych4562 Thank you! I have tried so many different ways of getting a great image. LOL As a photographer, I find astro a frustrating challenge. I have used the arcsin stretches and I think I did better when I was not guiding. Oh well, the experience of just being out is half the fun. Thank you again.
@AstroEd's Astrophotography Channel I mean the processing. I enjoy shooting but I'm new to the complex processing. I'm definitely down to learn. But so many ways to get to the same end result too.
I have a question…I apply your method to my recent Milky Way photo and it brings up a lot of noise…the imagine have already a lot of light pollution thanks to the places where a I based and if I tried to reduce the light pollution with the dust tool in photoshop and then apply your method it’s brings up a lot of noise…then I tried to not reduce the ight pollution and apply this method but the part of the Milky Way near the horizon comes out to bright…probably I made some mistakes…do you know what I can try? Thanks
![Astrophotography: WHY using Star Removal (Starnet++) is a MUST in your Photoshop [PS] workflow! P1.](http://i.ytimg.com/vi/qvgYO6PTZ28/mqdefault.jpg)
![Astrophotography: WHY using Star Removal (Starnet++) is a MUST in your Photoshop [PS] workflow! P1.](/img/tr.png)
I just used this technique to stretch a the horse head. I used other techniques from other tutorials before and this is by far the best way. Other tutorials I have followed have ruined images that probably could have been much better. Funny thing is that the other tutorials are from people who are pretty well invested in the hobby. Thanks!
Thanks so much for this comment! Excellent to hear someone who has tried it! Glad it has helped!!
this sounds so weird if you werent an astro person
Man!!! Thank you SO much for sharing this info and for your good tutorials. I'm re-making all my astro images again; using this method, results just incredible! Thank you again!
So glad I ran across this video. I've watched at least a dozen videos on stretching, from all the usual astrophotography channels, and they're all variations on what boils down to an experienced astrophotographer using instinct and gut feel to adjust curves and levels. This video feels like a completely different approach. The arc-sine hyperbolic stretch demonstrated here has made a huge improvement in my images--and it's repeatable. When I was stretching by the seat of my pants my images came out different every time. With this technique I get much more consistent results and I'm not accidentally screwing up the color balance or clipping blacks and highlights. Great technique. Game changer.
Thanks, Jon, I appreciate this comment. I ran into the same experience when I first started. I never found an astrophotographer that knew what they were doing on RUclips really unless they were PI users. That lead me to creating a logical repeatable workflow that follows image processing techniques from an engineering perspective in PS. I think PI is ultimately the way to go, but I haven’t found a PI technique I cannot implement in PS. Thanks, again.
Your video was referred to me on a FB astrophotography page and I am so glad they shared it with me. I'm new to stretching data but I really like the method and results here compared to what I learned in another video for stretching data in Photoshop. Hoping you have a follow-up to this one to show how you did the remaining post-processing on the image. Thanks!
Yes, i have 2 more follow up videos to this initial video. Plus i have an entire narrowband processing tutorial videos. Just goto my youtube channel page and look under image processing tutorials. You will see all the videos there.
thanks man. almost wanted to give up astro because i couldnt bring anything useful out of my photos. appreciate helping everyone out
So awesome! Stick with it! More tutorials to come!
Would be awesome if you could do a follow-up to this and let everyone know what your next steps would be in image processing
I do have a follow on video to this. It is the only other processing tutorial video I have and it is the next step I take which is using starnet.
Just came across your video and have to reiterate what everyone else has said - hands down the most easily understood PS processing tutorial I have encountered. Like that it was not too long, and would love to see followups of processing this particular image to its final iteration so as to understand your complete workflow. Subscribed and anxiously awaiting for more like it.
Awesome! Thanks for watching! I plan to release the next in the series within a couple weeks. Thanks again!
Thank you for the tutorial. This is a great trick to have in one's toolbox.
I appreciate your video and will attempt this technique next time I am editing
So awesome to hear! Let me know how it goes next time you edit 😊
Wow, thanks for this video. I also have downloaded the Arcsinh curves as well from the other video.
Thanks for watching. 🙂
Found this really helpful thanks. I like knowing the theory behind the curve, instead of eyeballing its good to know I can start from a known good initial stretch. I have a recent image of M42 I'm going to reprocess and see how much it changes the outcome.
Awesome! Hope M42 reprocess turns out great 😊
Ive just followed this Ed, very well explained and nice to have a different approach to use. Thanks.
Thanks, Ollie!
One of the best tutorials on stretching data in Photoshop!
Thanks, Wido!!
Everyone clearly is getting great results. I'm getting a polarized mess every time and with every zoom option. I'm using subscription P.s and its all up to date. I know I have every layer set right and blend modes correct, just cant figure out whats going on. Your tutorial is easy to follow. Thanks
@@stevelloyd1374 this process works optimally for really dark skies, or dual, tri or quad narrowband filtering. What is your setup? Not sure I understand what you mean by ‘zoom’ options.
@@astroedsastrophotographych4562 Thanks for your response, I downloaded an 16 bit raw TIFF from Johnathan Eggleton that he shared on the ZWO FB page. The image was taken using a. L-extreme filter. The zoom factor I was referring to was the screen zoom in Photoshop that you mentioned could cause viewing issues. I might try one of my one images and have another play.
Is there a way to share screenshot images with you?
Thanks.
Excellent tutorial Thomas! Will give it a try👍
Awesome! I’d be curious to see your results with it! 😊 Thanks!
Easiest video to follow, good job as always!!
Thanks, Josh!
Another technique to try many thanks!!
Absolutely! Thanks for checking it out !
Very nice, i will try this out for some of my deep sky astrophotography.
Great. Let me know how it goes!
@@astroedsastrophotographych4562 I been trying out the curves in photoshop cc , but this technique would be the best way. I will try it out on messier 33 and messier 42.
Cool, good luck!
Great video and nicely presented. The explanations are helpful where needed.
Great to hear 👍, thanks for watching, Marc 🙂
This is a fantastic tutorial. Thanks so much.
Absolutely, thanks for watching!
I'm not normally one to leave comments on videos, but the information in this video is awesome.
I've just followed along with my very first stacked TIF file from DeepSkyStacker and was amazed at the information I could see.
(loving the ASI294, and that bit about balancing its colour cast / RGGB filter was very helpful)
So much more detail than just a simple edit in Lightroom and I can now see what you mean about 'star bloat' and 'colour bleaching'.
Have now subscribed and plan to watch the rest of this series.
Is there a way to follow them along in order of the editing process? I couldn't see if they were numbered in some kind of order.
Thanks so much for the info and the effort in making these tutorials.
( lol , 42 comments...)
Thnks, Matt. Glad it has helped. If you look at tue playlist for image processing tutorials I have only 3. This is the first in the series. Next is starnet++ followed by why contrast is important. They all use the same image until a final image is produced. Note that this isn’t a narrowband tutorial the final video just tries to maximize the data thru detail enhancement, mainly focusing on the idea of contrast. Thanks again! 42 seems like your number 😀
Great great video. I've been doping a landscape wide astrophotography, got my tracker few weeks ago, and I wonder if this technique would work well on wide lens shots.... Will try for sure.
It works best in combination with dark skies, dark frames, and flats. If one of those isn’t a part it could look off. You can also substitute dark skies with a narrowband filter. Clear skies.
Would love to see a tutorial on the best way of doing Globulars !
Not a bad idea
Ed
Great tutorial on how to produce fantastic images using photoshop!
Its great to see your workflow
Cheers mate🦘🦘🦘
Thanks, Ben! I plan to periodically release some image processing techniques I use. Thanks again!
Nice video. Well explained and demonstrated.
One comment, though. Can’t you complete this process in 32 bit mode? Some desirable data deep in shadows is being compromised by the early 16-bit conversion, while it looks like these steps should be available in Photoshop’s 32 bit mode. Normally the 16-bit conversion would be made after completely the initial stretching (i.e. what is done in this video).
Great video , complete new approach , thanks
Thanks. This is an amazing way to stretch data!
Best tutorial yet on how to use this tool properly. I've had it for about a month or two but I've been over using it.
Thanks, Ted!
Thanks again, best video i have seen so far, at the very end of this video did you then flatten the image ?
Thanks for watching! Never flattened the image. Kept it as a non-destructive workflow for future reference.
Nice method thanks for sharing
Absolutely! Thanks for watching!
Great results! However keep in mind, according to Mark Shelly's site, Photoshop is not capable of doing a hyperbolic arcsin due to restrictions on the plotted points. His workaround was to create functional square roots of the desired curves. Two stretches must be performed back to back so they multiply to a true hyperbolic. Mix and matching the various 10x, 30x, 100x... in order to get the stretch factor you need. Im not sure stretching thrice will give you a true hyperbolic either. Nevertheless, you gave a great intro into Mark's technique. Thanks.
Thanks, appreciate the feedback. I usually do two to three stretches (eye balling it) but good to know two stretches with varying multipliers is the correct approach to a true hyperbolic stretch! Thanks, Stephen! Clear skies.
A how to by you. Really cool. This was the most interesting thing I stumbled upon Today. I love to see how folks work. Nice you quote your source. Forest Tanaka was my biggest influence. Who is yours?
Thanks, Garnett. Honestly, probably my biggest influence is Chuck Ayoub. He has always motivated me to strive for the best and grow, through observation of his example.
I’ll give this a try! Looks pretty good.
Are the spikes/rays in the stars from some other processing, or natural out of camera?
If you use a newtonian telescope you’ll get these diffraction spikes due to the spider vanes which hold the secondary mirror. If you’re using a refractor telescope you will not get these. Hope this answer helps.
Brilliant thanks
Awesome stuff!
Thanks, Mike!
This is marvelous! How do you get that channel separated histogram in the right menu?
If you don’t see it, I would try going up to the file menu under windows and select histogram. I am also using the latest version of photoshop CC.
You're videos have been amazing and like others have said they are the most well explained I've seen, thank you so much!
So I'm a newbie and have a 1600mm, how can I incorporate you're tutorials into a mono workflow? Should I simply assign my Ha, OIII and SII to my choice of Channels and pick up your video from there?
I recently released a narrowband tutorial series, check that out, it should address your questions! Thanks and clear skies!
Ok man, you deserve way more subscribers. (I just did) Your stuff is excellent 👍 Tnx for sharing your knowledge and please keep them coming 🤞
Thanks, man! I appreciate that! I do a lot of astro imaging adventures and way less tutorials. But I still have tutorials planned.
ok looks great, but would nice to see your final image as well. I have used the download arcsinh stretches, but not in this method, and especially like your meticulous color balancing. I also just got a triad ultra, so will be using that in a Bortle 9. My question is, if you have PixInsight, and do the standard stretch there...will that give a different result than doing how you did it, especially with the grey multiply/divide layers ? or is this a Photoshop "equivalent " to a stretch in PixInsight? thanks for this video!
Thanks and thanks for watching, Greg. I plan to release more videos on my workflow that will show this image progress. I didn’t want to release a super long video no one would watch, so stay tuned for more. As for pixinsight, I’m not a user myself but do know that there is a separate module for doing arcsinh stretch that Mark Shelly wrote in Pixinsight, difference is you don’t need the grey mult/divide layers you can apply the stretch much simpler in PI. See his website for reference, he talks about it, link in the description. So I believe the standard stretch is not the same but could be wrong.
Thanks for a terrific video. One question from a newbie....Is step1, the initial color balance, necessary?, Since step 2 is to bring all the channels down to the leftmost limit without clipping, wouldnt the histograms get aligned automatically just by doing step2?
My initial thoughts agree with you. What I plan to do is what you suggest and do a empirical comparison between the two and see if there is any difference. If there is I will respond back to this post eventually. Thanks for watching!
Excellent!!
Thanks!
Thanks Astro Ed💫
Absolutely! Thanks for watching!
good video man. thumbs up
Thank you, Mike!
Hi Ed I have been using your method and workflow now on all my images, it’s fantastic. Gives so much more depth and colour, hard to believe that it was there waiting to be discovered. One question - on the initial channel mix alignment - I am finding that most images are too dark after stacking and all the colours are right on the left-hand side of the histogram without any differentiation between them. So I have been doing some initial curve stretching so I can separate the colours, usually after a couple of stretches the colour channel differences become apparent and I can start aligning them. Is this the correct approach or is there a better way?
Hi Ray, Usually the only reason my channels are seperated initially is because my OSC has an RGGB sensor and thus green channel pulls ahead. I only align the channels because of that. Are you using a DSLR? Also filters can change this response. Seems odd that you would have to separate tue channels and realign them. If they are already aligned I would think that is where you would start the arcsinh stretch from...
Thanks Ray for asking that question. I was under the same impression as you were.
Hi Ed I am using a OSC ZWOASI2600 and a Optolong L-Extreme filter. I find that a couple of slight curves stretches separates the channels enough to get going, no real issues
@@rayjohnson7437 Hi Ray, I guess I don’t understand why you are trying to separate the channels?
@@astroedsastrophotographych4562 so when this happens I should proceed straight to the next stage (grey multiply/Grey divide). Should I do any channel alignments after the first arc sine curves stretch if they are out of wack? (Sorry for all the questions, please remember you I am a complete novice!). Rgds Ray
Great tutorial. It has helped with a lot of my targets and I have re-processed them. However, it seems to fail dramatically on my M42 orion images. Any clue to why?
Were you in dark skies, or use a filter? Else it isn’t the technique but rather the light pollution issue. Also I assume you do flats and dark frame correction? Else that would be another issue.
@@astroedsastrophotographych4562 I am in bortal 5. Half of my shots were without L-Pro filter and the rest were. Most were on an ASI533 some with a DSLR and the rest ASI1600MM. Yes Darks and Flats are shot.
@@NitescapeAstro sounds like you were stretching light pollution data at bortle 5. Bortle 3 or below you could probably get away with it or using filters.
you just earned a sub! 👍🏻
Thanks, Stefan!
Anyone know if there is such a preset for GIMP?
Thank you for the video. I am new to Astrophotography and just purchased a monthly subscription to PS. How do I load the unzip the file to PS which is web based? The instructions for loading the file assumes the program is physically on your computer. Thank you for your consideration in this matter
Did you figure it out yet? I have the same problem.
Unfortunately, I have no clue. I am also not affiliated with Adobe PS, so I’m no help here.
Very interesting! I’m eager to try this on some data, but my images from my canon 6D tend to have a red glow and the red histogram leads the other two, would I want to boost the red in my green and blue channels instead to align the histograms?
That is exactly what i would do, to balance the color.
Hi Ed. I am using this technique and, after a series of stretches, I can see color in my stars but the nebulosity still looks grayscale. It is a reflection nebula so I am expecting some blue hue. I am zoomed in beyond 67% so that's not the problem.
Hmmm… let me know your telescope speed and integration time and filters used and camera
I'm using an Edge HD8 at f/7, IR/UV filter, ZWO ASI071 MC, and 5 hours total integration. I processed this target last year not using this technique and the color shows just fine. I was going back to this data and re-processing it using the color-preserving technique when the lack of color problem arose. I really like the idea of preserving star color and not blowing them out so I am hoping I can get this color-preserving technique to work!
@@robertstewart1466 i have the same camera, and the same problem. When stretching the multiple channel, the reds of nebula are very pale, not B/W, but very pale. Can't seem to figure out how he gets the normal color when stretching that layer.
OK, I'm with you up through stamp and desaturate the Grey multiply layer. I then stamp and add the Grey divide layer. Change the blend modes to Divide and Multiply. However when I begin stretching (arc sin) the image remains in grey scale. Yours doesn't. I have started this over about 5 times and the same thing happens. Suggestions? I know its something small I'm missing. Thanks for great instructional videos.
What layer are you applying the arc sin stretch to? It should be the multiply layer.
@@astroedsastrophotographych4562 The multiply layer.
@@pemongillo strange you should see color not greyscale… also make sure your zoomed past 67%, maybe you’re seeing artifacts?
@@astroedsastrophotographych4562 Ok I'll try that
Also I would check the image you have prior to the mult and divide layers. Make sure that is the color version and you didn’t set it to grey on accident.
This is a great video, i now have added these to the list of things to use for astroprocessing in PS. One question, I shoot mono with filters.. So, would it be best to stretch before I combine or after in this instance lrgb, and what does this mean for narrowband imaging?.. As far as how to utilize these with false color? Tips or hints?
Can you help me figure out how to unzip these into the proper folder? I can’t seem to find the right one that makes it appear in my curves presets!
That was interesting. But my question is, how much sense does this way of stratching make, when I remove the stars with Starnet ++ before stretching? What I understand is that it is about preventing stars from blowing out. So is there an other reson to use this stratching procedure?
Edit: You dont remove the stars with starnet first, you stretch first then use starnet.
The whole point is to stretch your data and stars as best as possible without blowing out the stars or over-stretching and bleaching your data. Then you use starnet, then with your starless image continue processing your data and recombine the stars later that you extracted from the first arc hyperbolic sine stretch.
For sure I remove the stars on the very beginning. What else fore I should use Starnet++ ?
@@kowalskik.1333 you should always stretch first before using starnet, else you will not be able to stretch your data properly afterwards. Even the instructions that come with starnet say this.
@@astroedsastrophotographych4562 OK agree, but my frame comes already pre stretched from APP (Astro Pixel Prozessor) to Photo Shop. Should I make sure APP is not stratching at all?
@@kowalskik.1333 i would compare app stretching to the method I show here, and then use what you think is better on your data. So for this method, don’t put your data thru app.
Thank you for the video!
I tried to apply this method on my 1.5h integration of Heart and Soul in Bortle 4-5 sky on 135mm astromodified DSLR with moderate twilight pollution.
ArcSinh probably (not sure yet) does a better job at stretching (final result visually looks slightly beter), comparing to simple black and midpoint Levels adjustment.
However, the actual introduction of Grey Multiply and Divide method made the final image much worse. The colors look desaturated, and almost like washed out into the grey.
I am not sure if this method is just more scientifically correct (more realistic color rendering), while visually less appealing??? Or is my data just not good enough for this method to show its potential?
Note that my original stack shows substantial color cast and gradient due to twiligh pollution. Do I need to remove the color cast and gradient first, before applying the grey×÷ method?
UPD: Tried this method AFTER removing the background. The result is the same - the image looks bleak, almost colorless, way less vivid and appealing than if no grey×÷ was used.
Whats going on?
This technique works optimally on good data. Unfiltered data is horrible and corrupted (as you say bortle 4-5, with a gradient), you are extracting dominant signal which is light pollution it sounds like in your case. You need a good narrowband filter or super dark skies. Your problem has nothing to do with this technique rather your equipment/light pollution issues. Even using background removal techniques will not resolve this issue, because you truly have pixels that are still a mix of light pollution with signal.
Thank you for the reply!
Will try to improve my data!
Also, will try to apply this technique on some of the beter unstretched data available online to compare the result.
Thanks
Would this workflow work for Andromeda Galaxy, I'm using color camera, ASI 2600MC Pro. Thank you.
Yes 👍
Hey Ed follow up question on this does it matter in which way you channel mix, for example the way my histogram was aligned, I had to drop the Green in my Blue channel. I'm guessing this is still fine?
I’d say let your result guide you. Really no wrong way unless it degrades your image quality. If that is your only option then you have no choice. If you have another option, I would do a study where you do both ways and see which results you like best or maybe they would even turn out equivalent. 😊
@@astroedsastrophotographych4562 Yep makes sense Ed pretty much what Ive done just wanted to make sure. Thanks 👍👍
Can this process be completed with an image from L+RGB data? Once stacked in DSS, I normally use curves and levels on each colour and then combine them and then add the Lum. I find this method works great on a DSLR image.
Yes, this method should still work for that.
Will this work with a normal DSLR file to pull out detail?
Yes, if you are not in lots of light pollution or you are using some type of filter to reduce light pollution.
I following this technique alswell i am using an Atik 383L mono, now I did the RGB images using this but What should i do with my Luminance? i mean how to add that to it? or is the multiply or divide layer the actual luminance layer? as i shoot with LRGB filters.
Check out my next few videos specifically show that in the next video, but using a false luminance. It would be the same process. Thanks for watching!
Ah yea but then i shouldnt crop the RGB before feeding it to starnet right? Becouse otherwise the Lum wouldnt align anymore with the RGB isnt it? Or should i add the lum before feeding it all to starnet?
This doesnt seem to work for me, the image looks a bit like a negative picture after stretching. Does this method work in photshop CS6?
This because of a ps bug. Zoom past 67% and you will see it works. I explained this in the video. You must have skipped around and not watched it all, as I explain all this, after the stretching technique the next step combines all this and you no longer need to zoom in to see correct details.
@@astroedsastrophotographych4562 Great thats fixed it thank you.
Can you use this when editing the MW.
I’ve edited mw shots before I arrived at this technique. However I would, and see what results from this technique.
why is stretching done in multiple steps? why not in one que? does it have relevance to precision?
I use multiple (2 or 3) smaller arcsinh stretches so I don’t over stretch my data, there is a certain point which the data cannot be pushed past without bleaching (pushed too close in luminance to the white point). It gives me finer control in stretching.
what version of PS are you using?
Photoshop CC 2020
@@astroedsastrophotographych4562 Probably why I couldn't find the Layers Divide Mode! I'm back on PS2! Any way past that?
Maybe GIMP!
Has anyone made an action yet?
Hi there, not to my knowledge. There is this option in PI, I just implemented it in PS. I could code up an action eventually, but if I did I would sell it as my own action.
@@astroedsastrophotographych4562 I made one myself however could not get it to work on more then one iteration.
I personally wouldn’t recommend an action for this technique. It is usually data dependent.
I tried this and I ended up with a stretched black/white image instead of color.
I don't use these keyboard combos and I have no idea which layer you copied over to the top, at the end.
Keep watching the video and it will make sense.
Did you solve this problem? I always get stretched desaturated image.
Dear friend, thank you for sharing this. Could it be that in the third Arsin arc 10 I already have more swelling of stars and more saturated around them? I have tried several times and I cannot get the stars to remain unchanged with the initial image
Unsure, what is your setup that you used for capturing your data?
@@astroedsastrophotographych4562 First of all thanks for your answer
I use a Nikon D 750 exposure 48 sec focal length 180 mm askar 180 iso 500
50 light 50 darks 30 ways no flat
I'm trying to load the curves Arcsin but Photoshop says it's the wrong kind of document and will not open it. Why?
The files are already unzipped
What version of photoshop? Did you follow the instructions on the website? I had no issues.
@@astroedsastrophotographych4562 it's the latest version of Photoshop CC. Yes I followed the instructions. I unzipped and double click to load a sinh and then PS says wrong kind of document. I don't get it. It's and acv file and I made a scratch curve like he said to see where it is being saved. My curve is there but it will not let me download these files in that. I don't know man I'm about over this Astro stuff. Nothinbg ever goes the way it's supposed to.
Not sure, it shoukd work with latest version of photoshop CC, that is what I’m using too. AP is hard and those who truly want to be successful will find away. If you’re not liking this hobby, that is ok, it isn’t for everyone. I enjoy debugging problems until i find solutions, and AP is full of that, it gets boring when everything runs to smoothly. 😊
Hi. Does anyone know if these work on a Mac? I've installed them in what I think is the right place but they don't show up in the Preset Manager to load. Any help much appreciated.
It works, but it often still bleaches the image for me. I’m probably doing something wrong but I have no idea what.
Are you using an lp filter? Are you in a high bortle sky?
yep, bleaches every time for me. I've tried about everything, different zoom levels, etc. and when I stretch the multiply later, it's just bleached red. ASI071MC Pro, Rosette nebula that should be very red.
I have one image that has the main data towards the middle of the levels adjustment panel. There is also (height wise very small) a lot of data to the left of the main spike. Do you clip only up to the small data or go up to the main spike?
It looks kinda like this " _____IIIIII_____"
I do multiple small level adjustments up the small amount of data to the left of the main spike. Continually doing these small levels you will eventually or should see the those two spikes merge into one. At least that is what happens for me.
Super agradecido por lo que nos traes aquí ya lo estoy probando! Mi único problema es que no tengo mucha información de negros a pesar que hice un apilado en dds y use 45 lights de 1 minuto tal vez debería hacer muchos más darks de los que hice o más flats ? Allí si tendría más información de los negros? Gracias por todo
I downloaded the arcsinh files. I put them in the same curves folder mentioned, but when I select Curves in PS, they are not there to select. Am I doing something wrong?
My Bad! I did ctrl-L for Levels. I am used to doing that first. Thanks for the video, I am now trying this on my first AP image.
Excellent, glad you got it sorted out.
Hi I tried to send you a message on your website contact form. Don't think it went through. Basically I followed your technique using the arcsin10 adjustment but am not getting the same results. I use a DSLR and telephoto lens on a Star tracker and take about 50 to 100 lights and stack them to start my image in photoshop. My question is will your arcsinh10 adjustment process outlined in your Photoshop videos work for my DSLR image or is that for sophisticated telescope cameras that take color images in different channels?
Do you take flats? Sorry I need to check my website email more often. This technique should work regardless of camera.
@@astroedsastrophotographych4562 Yes I take Flats, Dark, and Biases. THanks for the fast response! Keep up the good video teaching!
Yeah no problem. Second question. How do you stack your data, DSS? You use the raw files to stack?
@@astroedsastrophotographych4562 Yes I stack raw files.
Next question. When you say you’re not getting the same results, what do you mean? Did you shoot the same target as me with the triad ultra filter? If you have light pollution and no narrowband filters your results won’t be limited by my stretching technique but rather by your light pollution.
Beautiful, but does not work for me. I will keep trying.
Sorry to hear that, let me know if I can try and help.
@@astroedsastrophotographych4562 Thank you! I have tried so many different ways of getting a great image. LOL As a photographer, I find astro a frustrating challenge. I have used the arcsin stretches and I think I did better when I was not guiding. Oh well, the experience of just being out is half the fun. Thank you again.
@@JonnyPink65 I know the challenge, keep going you will break through! Clear skies John.
Man...this seems complicated for someone who doesn't know much on processing in PS
Astrophotography is complicated. One has to commit to learning.
@AstroEd's Astrophotography Channel I mean the processing. I enjoy shooting but I'm new to the complex processing. I'm definitely down to learn. But so many ways to get to the same end result too.
I have a question…I apply your method to my recent Milky Way photo and it brings up a lot of noise…the imagine have already a lot of light pollution thanks to the places where a I based and if I tried to reduce the light pollution with the dust tool in photoshop and then apply your method it’s brings up a lot of noise…then I tried to not reduce the ight pollution and apply this method but the part of the Milky Way near the horizon comes out to bright…probably I made some mistakes…do you know what I can try? Thanks
Yeah this is a hard situation. In all honesty, don’t shoot in light pollution, it will be a losing battle on a target like the milkyway.