A great technique Shawn and thanks for sharing. If you check the 'Rescale result' option in PixelMath, you can simplify the expression as just 'lum + lum_USM + lum_MLT'. I also find it works even better on starless images otherwise it can sharpen and emphasise stars a bit too much depending on the scope of the star mask.
Hey Shawn I've been a heavy PI user for five years now and I say I always learn new techniques from a limited handful of RUclipsrs like yourself who contribute and share a wealth of knowledge on PI. This video was a great addition to my PI arsenal of techniques. I tried this method on five of my recent sets of image data and the sharpness was outstanding! Thanks for the vids Richard
Excellent tutorial as usual. I appreciate some of the basic procedure comments, i.e., "You can drag the triangle into the image....or.....click the square.....", etc., as it reinforces my knowledge of the PI controls. I am generally already familiar with what you are discussing but it confirms that I have an accurate understanding of the processes. I like your pace and complete discussion of the steps. Thank you.
Excellent topic for a tutorial. I recently got the ASI2600MM and ASI2600MCP to allow me to image chrominance FAST on the OSC camera, then image luminance on the Mono camera to increase detail. Good timing!
Try deconvolution instead. Those stars are not so stood after unsharp mask. Or at least mask the stars beforehand…they don’t need to be sharpened ;-) Another good idea for an image like this (delicate filaments) is to try LHE…the added contrast gives the illusion of additional detail. Finally, a masked low pass filter works extremely well as long as its targeted (not global). PS layers make this quite a bit easier than in PI. In PS, duplicate the image layer (near final LRGB, for example), use filter low pass, change blending to soften or overlay to see the effect….it will be too drastic as a global change, so fine tune opacity and mask areas where it’s not wanted (try an inverted layer mask). Clear skies!
Your videos are providing invaluable insight into Pixinsight, Shawn. Many thanks. BTW, took delivery of a Starfield Optics Gear 80 yesterday; along with an L-eXtreme filter; so look forward to your impressions of the scope as we go forward. First light likely tomorrow night.
Ha. Well this was fun. The MLT settings didn't work for my image, but they did when I deleted the manual settings. I have to say, I would have to do this a number of times to be able from memory to "Send in the Clones" to the extent required. Whew!
Great video Shawn. Out of curiosity, had you used PixelMath to add the three Lum but without the fractions and checked the Rescale Result button. Would it have used 1/3 of each Lum? Asking for a friend.
Thanks Shawn. I was glad you mentioned sharpening should be done near the end, I have been trying to do it at the beginning all this time until very recently I tried it near the end. I like this method you described here as I have used all of these tools you used except the unsharpen mask. For me the biggest thing is going to be keeping track of the masks created. - Cheers Kurt
Hi Shawn. I'm getting addicted to these videos of your's. On January 30, 2021 you did a tutorial on adding luminance to an RGB image. In that one you treated RGB and luminance separately through DBE, noise reduction, and stretching. For this OSC image you wait till after stretch to extract the LUM. Would it not be better to do the extraction BEFORE MLT noise reduction so as to keep the structure of the Lum image as sharp as possible? Secondly, with Starnet and StarXterminator, would it be better to remove the stars, then do the Lum sharpening and lastly add the stars back at the end?
Thanks for the great vid Shawn! I stumbled on the chaotic nebula page a while ago and have been using the technique on my last couple targets. You didn’t include the masked stretch along with exponential transform and HDR transformation steps that proceed the sharpening, or the dark structure enhance step that comes after and am interested in your thoughts on those. For myself, the stretch technique sometimes but not always works better and same with the dark structure, but I’ve yet to get something from the HDR process that I wanted to use. Cheers, Scott
Great tutorial, as always Shawn! When PixInsight burst into the amateur astroimaging scene, the software seemed to be so different to what I had used...and so complex....that I immediately scurried back into my comfort zone & didn't give it the time of day after that However, your tutorials have me feeling far less intimidated. So, once I have upgraded my computer, I will be taking the PixInsight plunge & see what I can do with the tons of old data I have yet to fully process! Thanks again Shawn! p.s. How are things progressing with the Apex super-reducer?
I went through this on my NGC 6960 image and it does help but its easy to get lost in the steps. It would be nice if this could be automated like the EZ Scripts.
Hi Sean, this is a really good sharpening method, thanks for sharing it. My only concern is when you apply the sharpen mask the starts tend to get clipped (become too bright). Have you tried this without the stars?
Some examples of Pixinsight images seem to have more details than would be possbile givien the unprocessed original image data. is Pixinsight synthetically enhancing images using image data scraped from larger telescope databases?
When we extract a luminance layer from an RGB image with the toolbar icon, we are not really removing data but creating a new image representing lightness, right? Then if we modify the L later and add it back, are we now not adding more data to the image? I have found that with full field nebula, that adding the L back makes the whole LRGB image a bit too bright. Is there a way to add back these L details without brightening too much? Many thanks (loaded question?).
It is technically a lightness image. I suppose you could look at it as adding more data but I think of it as modifying that lightness component of the image. I've not experienced any brightening unless a histo or curves transform is applied. Not sure of another way to integrate it not having had issues. Never explored that.
In the LRGB combo tool the lightness transfer function seems to help with that for me. Anything over .5 reduces the brightness, under increases it. The saturation slider can be used similarly to boost the color a bit too. Scott
Hey Shawn, great tutorial, I’ve not done sharpening this way, I’ll definitely try that thanks! What I have done though, similar in as much as I extracted the luminance from my OSC data and added it to the LRGB combination tool, switched of rgb as you did and added it to the L channel, then using the saturation slider, I moved it down to 3.1 from 5, and applied it 3 times to my colour image and the transformation was awesome! I’ve just put a video on my RUclips channel yesterday showing the difference in my two images, check out Simon’s Astro and look at my last vid😀 Thanks Shawn Clear Skies
Decon can be be done as well. It is however not sharpening, often confused. Decon would be done before stretching. Sharpening is done after, near the end of the workflow. I had thought of protecting the stars but wanted to demonstrate this method as close to its workflow as possible. Cheers.
@@VisibledarkAstro a solid Decon in PI often obviates the need for sharpening…especially with unsharp mask which is fairly crude. Rarely does a global change apply to all parts of an image. So in my own processing stars are treated differently and separately from the filaments in the Veil area, for example. Be well.
@@deepskywest3633 agreed. Decon can yield good results. And unsharp mask can be crude if over done. Decon isn't sharpening though. Two different processes. As I said I was just demonstrating this method which used unsharp mask. This method can be modified further or not used at all. Cheers.
A great technique Shawn and thanks for sharing. If you check the 'Rescale result' option in PixelMath, you can simplify the expression as just 'lum + lum_USM + lum_MLT'. I also find it works even better on starless images otherwise it can sharpen and emphasise stars a bit too much depending on the scope of the star mask.
Another good PI video Shawn. Thanks.
Hey Shawn
I've been a heavy PI user for five years now and I say I always learn new techniques from a limited handful of RUclipsrs like yourself who contribute and share a wealth of knowledge on PI. This video was a great addition to my PI arsenal of techniques. I tried this method on five of my recent sets of image data and the sharpness was outstanding!
Thanks for the vids
Richard
Great video Shawn. I watched-while-processing my image to produce my luminance layer. Worked a treat!!!
Hey Shawn - just an FYI: I have used this method on my last few images and it works great in my opinion. Cheers Kurt
That's great! Thanks for the update.
WOW! I'm very new to PI and I've become very invested in your videos, Shawn. This is fantastic, thank you. Onward and upward from here. Cheers
Thanks Shawn, for another informational video. Not only did I learn a new sharpening technique, I also learned about the auto clip in HT!
Me too!
Thank you Shawn , another excellent and very helpful presentation.
I just used this on my Xmas tree cluster and I am so happy. Shawn your awesome man!
Excellent tutorial as usual. I appreciate some of the basic procedure comments, i.e., "You can drag the triangle into the image....or.....click the square.....", etc., as it reinforces my knowledge of the PI controls. I am generally already familiar with what you are discussing but it confirms that I have an accurate understanding of the processes. I like your pace and complete discussion of the steps. Thank you.
Your tutorials never get old. I watched it today and it will definitely help me as I progress my way in learning PixInSight. Thanks for your tips.
Super helpful on my image of M76 that I upscaled 4x. Thank you!
Excellent topic for a tutorial. I recently got the ASI2600MM and ASI2600MCP to allow me to image chrominance FAST on the OSC camera, then image luminance on the Mono camera to increase detail. Good timing!
Very nice! That will work great. You can apply this method to your mono data and then add into the RGB. Thanks for watching and the comment.
Try deconvolution instead. Those stars are not so stood after unsharp mask. Or at least mask the stars beforehand…they don’t need to be sharpened ;-)
Another good idea for an image like this (delicate filaments) is to try LHE…the added contrast gives the illusion of additional detail.
Finally, a masked low pass filter works extremely well as long as its targeted (not global). PS layers make this quite a bit easier than in PI. In PS, duplicate the image layer (near final LRGB, for example), use filter low pass, change blending to soften or overlay to see the effect….it will be too drastic as a global change, so fine tune opacity and mask areas where it’s not wanted (try an inverted layer mask).
Clear skies!
Your videos are providing invaluable insight into Pixinsight, Shawn. Many thanks. BTW, took delivery of a Starfield Optics Gear 80 yesterday; along with an L-eXtreme filter; so look forward to your impressions of the scope as we go forward. First light likely tomorrow night.
That's great Douglas! Let me know how it goes for you. Will be a nice setup for you to work with.
Thanks for the tutorial Shawn. Yet another great lesson for those of use who image with OSC’s!
Ha. Well this was fun. The MLT settings didn't work for my image, but they did when I deleted the manual settings. I have to say, I would have to do this a number of times to be able from memory to "Send in the Clones" to the extent required. Whew!
Really enjoyed this video, a keeper. Informative and easy to follow.
another trick to add to the book! Thanks Shawn!
Great video! Thanks Shawn.
You forgot to show which Process/Script, etc you used to extract the Luminance data.
Wow! Nice technique!
Fantastic job, as always
Very well explained!
Wow combine two sharpening methods ….great.
Another great tutorial Shawn, cheers!
Thanks Cheryl! Appreciate you watching and the comment.
Thanks again Shawn.
Very well explained tutorial! I will try this technique. Thank you!
Glad it was helpful! Thanks for watching and commenting. Cheers.
Great tutorial, thanks.
Great video Shawn. Out of curiosity, had you used PixelMath to add the three Lum but without the fractions and checked the Rescale Result button. Would it have used 1/3 of each Lum? Asking for a friend.
Thanks for this tutorial! Going to give it a try.
Please do!
Very interesting! I will test this process. Thanks!
nice and good tutorial thanks shawn :-)
Thanks Shawn. Great tutorial.
Thanks Shawn. I was glad you mentioned sharpening should be done near the end, I have been trying to do it at the beginning all this time until very recently I tried it near the end. I like this method you described here as I have used all of these tools you used except the unsharpen mask. For me the biggest thing is going to be keeping track of the masks created. - Cheers Kurt
Shawn ~ what algorithm do you use in the MLT step, does it matter? Great technique and will redo a few past images to see the improvement.
Great bud 👌🏻
G’day Shawn
Great information mate
Looking forward to applying this
Cheers
Thanks 👍 Let me know how it goes Ben. Hope you're getting some clear skies down there!
Hey Shawn
Finally had last weekend & last 2 nights in a row and hopefully tonight also
Cheers
Excellent thanks!
Hi Shawn. I'm getting addicted to these videos of your's. On January 30, 2021 you did a tutorial on adding luminance to an RGB image. In that one you treated RGB and luminance separately through DBE, noise reduction, and stretching. For this OSC image you wait till after stretch to extract the LUM. Would it not be better to do the extraction BEFORE MLT noise reduction so as to keep the structure of the Lum image as sharp as possible? Secondly, with Starnet and StarXterminator, would it be better to remove the stars, then do the Lum sharpening and lastly add the stars back at the end?
Thanks for the great vid Shawn! I stumbled on the chaotic nebula page a while ago and have been using the technique on my last couple targets. You didn’t include the masked stretch along with exponential transform and HDR transformation steps that proceed the sharpening, or the dark structure enhance step that comes after and am interested in your thoughts on those. For myself, the stretch technique sometimes but not always works better and same with the dark structure, but I’ve yet to get something from the HDR process that I wanted to use. Cheers, Scott
Great tutorial, as always Shawn! When PixInsight burst into the amateur astroimaging scene, the software seemed to be so different to what I had used...and so complex....that I immediately scurried back into my comfort zone & didn't give it the time of day after that However, your tutorials have me feeling far less intimidated. So, once I have upgraded my computer, I will be taking the PixInsight plunge & see what I can do with the tons of old data I have yet to fully process! Thanks again Shawn!
p.s. How are things progressing with the Apex super-reducer?
I went through this on my NGC 6960 image and it does help but its easy to get lost in the steps. It would be nice if this could be automated like the EZ Scripts.
Hi any idea what was the reason for combining USM and LMT lum with the original one?
Why not only USM + LMT?
nice Shawn thanks!
nice tutorial
Hi Sean, this is a really good sharpening method, thanks for sharing it. My only concern is when you apply the sharpen mask the starts tend to get clipped (become too bright). Have you tried this without the stars?
Some examples of Pixinsight images seem to have more details than would be possbile givien the unprocessed original image data. is Pixinsight synthetically enhancing images using image data scraped from larger telescope databases?
I think RUclips ate any difference. When you were flicking back and forth at the and I could see nothing changing.
When we extract a luminance layer from an RGB image with the toolbar icon, we are not really removing data but creating a new image representing lightness, right? Then if we modify the L later and add it back, are we now not adding more data to the image? I have found that with full field nebula, that adding the L back makes the whole LRGB image a bit too bright. Is there a way to add back these L details without brightening too much? Many thanks (loaded question?).
It is technically a lightness image. I suppose you could look at it as adding more data but I think of it as modifying that lightness component of the image. I've not experienced any brightening unless a histo or curves transform is applied. Not sure of another way to integrate it not having had issues. Never explored that.
In the LRGB combo tool the lightness transfer function seems to help with that for me. Anything over .5 reduces the brightness, under increases it. The saturation slider can be used similarly to boost the color a bit too.
Scott
A little bit too fast but very interesting. Thanks Shawn
I thought I was going slower! :) Well slower than I normally would anyhow.
Interesting process but it did essentially nothing on my image.
Hey Shawn, great tutorial, I’ve not done sharpening this way, I’ll definitely try that thanks! What I have done though, similar in as much as I extracted the luminance from my OSC data and added it to the LRGB combination tool, switched of rgb as you did and added it to the L channel, then using the saturation slider, I moved it down to 3.1 from 5, and applied it 3 times to my colour image and the transformation was awesome! I’ve just put a video on my RUclips channel yesterday showing the difference in my two images, check out Simon’s Astro and look at my last vid😀 Thanks Shawn
Clear Skies
Link to the source?
Here it is chaoticnebula.com/pixinsight-sharpening/ Also can be found by googling "chaotic nebula"
Or, remove stars, sharpen all else and the put the stars back.
Decon can be be done as well. It is however not sharpening, often confused. Decon would be done before stretching. Sharpening is done after, near the end of the workflow. I had thought of protecting the stars but wanted to demonstrate this method as close to its workflow as possible. Cheers.
@@VisibledarkAstro a solid Decon in PI often obviates the need for sharpening…especially with unsharp mask which is fairly crude. Rarely does a global change apply to all parts of an image. So in my own processing stars are treated differently and separately from the filaments in the Veil area, for example. Be well.
@@deepskywest3633 agreed. Decon can yield good results. And unsharp mask can be crude if over done. Decon isn't sharpening though. Two different processes. As I said I was just demonstrating this method which used unsharp mask. This method can be modified further or not used at all. Cheers.
@@VisibledarkAstro nice job. Clear skies.
This was super confusing. You were making copy after copy after copy. Impossible to follow along.