Nice tutorial. I originally started using Histogram Transformation on my images but when GHS was released I started to use that. While it can produce some very exact and pleasing stretching, lately I've found that I'm spending way too much time fiddling with the options and still not getting a perfectly satisfying image...so I'm now moving back to Histogram Transformation and I'm finding it's more than satisfactory, especially when combined with Curves.
I think you'll find the Histogram Transformation tool to be the best choice. It gives you manual control and you can tweak it, see the outcomes as you go, and refine it on the Curves tool. Keep an eye out for the Curves video that I am going to make soon. It is the other half of the equation.
This tip is a game-changer! While I usually rely on GHS, this method takes image stretching to a whole new level of refinement. I'm truly impressed, and I can't thank you enough for sharing this amazing tip with me. Great job!
Great video and I support one comment that it takes a bit of time with GHS. I plan to compare both, but certainly your simple method looks like it could solve a lot of my headaches - "fiddling until I get it just right" Thank you and Clear Skies!
This is the original at least to me method of stretching an image, from way back when, before GHS was devised. GHS can be tricky but it is a much more refined way to stretch the histogram, giving you much greater control. Also I don’t understand why you need a separate compositor application after all Pixel Maths is all that any application is doing behind the scenes, whether it’s Photoshop, Pixinsight or Afinity. Can you explain this?
Manually stretching the histogram through the HT tool allows you to have precise manual control. Over time, you will learn to recognize relevant data vs optional data and/or noise on the histogram and can control what you want to incorporate. You can choose whether to stretch to keep all the information or reduce for contrast. You can see the outcomes immediately. It's all in your control. The curves tool takes this further and gives you additional refined control over the light curve. Automation isn't always better; sometimes it's just automated. PixelMath is okay for recompositing plates back together but it essentially just does something like a screen or overlay composite. With Affinity Photo (and there are probably other good tools for compositing plates back together) you can see the outcomes live as you sample each compositing method. This can lead to some interesting and occasionally very nice results you wouldn't have otherwise gotten. In future videos, I will cover these points further.
@@neverfox Sorry, I thought you were referring to that GHS script for PixelMath in the top right of my desktop, third icon down from the right, visible in most any of my editing videos. I tried the script and didn't care for the outcomes. Haven't used it in ages. I dabbled a bit with the GHS tool you're referring to but felt it was unnecessary. The HT tool stretch isn't finicky; it's use is straightforward and it can be made in a few seconds. Good old HT just seems a more elegant tool for the job to me.
nice tip! But I was hoping you're going to address misaligned individual R, G, B histograms too? Shouldn't the have 1. equal or very similar black points and 2. equal or similar x values of their histogram peaks in order to avoid color bias? As a colorblind astrophotographer I always struggle on visually evaluating if there is a color tint in the background...😄
I wish I could help you here, but not being color blind I don't think I could give you good experience-based advice. There are so many other elements to color that also need to be adjusted, such as saturation, vibrance and temperature. Where the black point begins is just one part of that equation.
Did you run the colour calibration & SCNR on the whole image and then remove the stars or run a separate colour calibration & SCNR once the stars where removed?
It's been so long since I made this video, I don't recall the exact procedure. My usual process with OSC data was to plate solve then run SPCC to color calibrate, then remove the stars. However, I've come to dislike SPCC and don't use it at all anymore.
whenever i import stacked tiffs into affinity they get a non destructive gamma correction that gets applied to the exported image, yours seemed to import without that, did you do something different?
Do you mean the gamma correction is done in PI? At first guess, I am thinking you may not be operating in the same color space between PI and AP. I find sRGB is the most consistent. Make sure both apps are setup for the ICC color space. I also turn off any form of compression when saving tifs to avoid degradation, but I doubt that affects color space.
@@SKYST0RY When imported into Affinity the ICC Color Space washes out the colors of the image. It can be turned off but it seems to be applied regardless after exporting
@@shiny_pantz This issue is absolutely new to me. I've never heard of it before. Next question is are you working with images of the same bit size between the apps. If you are working with a 16 bit image in PI are you saving it as a 16 bit image and opening it as a 16 bit image. I really think this must be an issue of pallete, though.
@@SKYST0RY It actually seems to even happen to raw exposures with no processing. After opening a fits file, Affinity will apply the SRGB color profile and then debayer the image. It then applies a levels and curves adjustment which I delete. The image appears washed out, as if a gamma adjustment is being done behind the scenes. I can modify the Display Transform in the 32-Bit Preview window from ICC Display Transform to Unmanaged and this changes it only visually in the workspace. Upon exporting the ICC Display Transform is applied and results in a washed out image. Looking online the only information I found was to edit the image with that transform in mind but that doesn’t sound right, I have no idea what is being done to the image.
@@shiny_pantz Are you trying to open using the astrophotography tool? I just use the ordinary open option so the image data goes directly into Affinity without adjustment.
GHS is just a pain...you end up producing a curve that you could get in curves straight away. I also start with auto stretching the histogram by dragging the STF on it.
Exactly! I cannot figure any worthwhile reason to go about using GHS. It is just reinventing the wheel, or, in this case, the curves tool, which is a much more elegant and precise tool for the same task.
I like what you show but you are a bit too fast for my experience level. Do you have any tutorial for editing OSC images with pixinsight and infinity photo. I have trouble with photoshop
whenever i import stacked tiffs into affinity they get a non destructive gamma correction that gets applied to the exported image, yours seemed to import without that, did you do something different?
I have never had this issue. I tried looking around the settings to see what's happening but can't find anything on it. Are you using 32 bit? That gives more export options so you may want to check that.
![Seungmin "그렇게, 천천히, 우리(As we are)" | [Stray Kids : SKZ-PLAYER]](http://i.ytimg.com/vi/kAzmhLHePqU/mqdefault.jpg)
Nice tutorial. I originally started using Histogram Transformation on my images but when GHS was released I started to use that. While it can produce some very exact and pleasing stretching, lately I've found that I'm spending way too much time fiddling with the options and still not getting a perfectly satisfying image...so I'm now moving back to Histogram Transformation and I'm finding it's more than satisfactory, especially when combined with Curves.
I think you'll find the Histogram Transformation tool to be the best choice. It gives you manual control and you can tweak it, see the outcomes as you go, and refine it on the Curves tool. Keep an eye out for the Curves video that I am going to make soon. It is the other half of the equation.
This tip is a game-changer! While I usually rely on GHS, this method takes image stretching to a whole new level of refinement. I'm truly impressed, and I can't thank you enough for sharing this amazing tip with me. Great job!
I am glad you found it useful. It definitely makes the whole process a lot easier and puts you back in control.
2024. I recall those first few years of RUclips. It was so clean, fun spirited, new and joyful.
The internet can be full of toxicity. I've long since decided just to mute trolls and move on without giving them further thought.
Great video and I support one comment that it takes a bit of time with GHS. I plan to compare both, but certainly your simple method looks like it could solve a lot of my headaches - "fiddling until I get it just right" Thank you and Clear Skies!
It's the curves where you want to do your fiddling, and when you learn curves the fiddling goes very fast.
This is the original at least to me method of stretching an image, from way back when, before GHS was devised. GHS can be tricky but it is a much more refined way to stretch the histogram, giving you much greater control. Also I don’t understand why you need a separate compositor application after all Pixel Maths is all that any application is doing behind the scenes, whether it’s Photoshop, Pixinsight or Afinity. Can you explain this?
Manually stretching the histogram through the HT tool allows you to have precise manual control. Over time, you will learn to recognize relevant data vs optional data and/or noise on the histogram and can control what you want to incorporate. You can choose whether to stretch to keep all the information or reduce for contrast. You can see the outcomes immediately. It's all in your control. The curves tool takes this further and gives you additional refined control over the light curve. Automation isn't always better; sometimes it's just automated.
PixelMath is okay for recompositing plates back together but it essentially just does something like a screen or overlay composite. With Affinity Photo (and there are probably other good tools for compositing plates back together) you can see the outcomes live as you sample each compositing method. This can lead to some interesting and occasionally very nice results you wouldn't have otherwise gotten.
In future videos, I will cover these points further.
@@neverfox Sorry, I thought you were referring to that GHS script for PixelMath in the top right of my desktop, third icon down from the right, visible in most any of my editing videos. I tried the script and didn't care for the outcomes. Haven't used it in ages. I dabbled a bit with the GHS tool you're referring to but felt it was unnecessary. The HT tool stretch isn't finicky; it's use is straightforward and it can be made in a few seconds. Good old HT just seems a more elegant tool for the job to me.
nice tip! But I was hoping you're going to address misaligned individual R, G, B histograms too? Shouldn't the have
1. equal or very similar black points and
2. equal or similar x values of their histogram peaks in order to avoid color bias? As a colorblind astrophotographer I always struggle on visually evaluating if there is a color tint in the background...😄
I wish I could help you here, but not being color blind I don't think I could give you good experience-based advice. There are so many other elements to color that also need to be adjusted, such as saturation, vibrance and temperature. Where the black point begins is just one part of that equation.
Did you run the colour calibration & SCNR on the whole image and then remove the stars or run a separate colour calibration & SCNR once the stars where removed?
It's been so long since I made this video, I don't recall the exact procedure. My usual process with OSC data was to plate solve then run SPCC to color calibrate, then remove the stars. However, I've come to dislike SPCC and don't use it at all anymore.
Sometimes I really like the hyper saturated stuff, they remind me of the photos printed in books from the 70's which I am very nostalgic for.
That makes perfect sense. An emotional connection is a perfectly valid reason to enjoy a style.
whenever i import stacked tiffs into affinity they get a non destructive gamma correction that gets applied to the exported image, yours seemed to import without that, did you do something different?
Do you mean the gamma correction is done in PI? At first guess, I am thinking you may not be operating in the same color space between PI and AP. I find sRGB is the most consistent. Make sure both apps are setup for the ICC color space. I also turn off any form of compression when saving tifs to avoid degradation, but I doubt that affects color space.
@@SKYST0RY When imported into Affinity the ICC Color Space washes out the colors of the image. It can be turned off but it seems to be applied regardless after exporting
@@shiny_pantz This issue is absolutely new to me. I've never heard of it before. Next question is are you working with images of the same bit size between the apps. If you are working with a 16 bit image in PI are you saving it as a 16 bit image and opening it as a 16 bit image. I really think this must be an issue of pallete, though.
@@SKYST0RY It actually seems to even happen to raw exposures with no processing. After opening a fits file, Affinity will apply the SRGB color profile and then debayer the image. It then applies a levels and curves adjustment which I delete. The image appears washed out, as if a gamma adjustment is being done behind the scenes. I can modify the Display Transform in the 32-Bit Preview window from ICC Display Transform to Unmanaged and this changes it only visually in the workspace. Upon exporting the ICC Display Transform is applied and results in a washed out image. Looking online the only information I found was to edit the image with that transform in mind but that doesn’t sound right, I have no idea what is being done to the image.
@@shiny_pantz Are you trying to open using the astrophotography tool? I just use the ordinary open option so the image data goes directly into Affinity without adjustment.
GHS is just a pain...you end up producing a curve that you could get in curves straight away. I also start with auto stretching the histogram by dragging the STF on it.
Exactly! I cannot figure any worthwhile reason to go about using GHS. It is just reinventing the wheel, or, in this case, the curves tool, which is a much more elegant and precise tool for the same task.
I'm wondering why when I do Histograms I dont see the color curve at all. I even zoom in and out...why is this ?
Are you shooting with a color camera? If so, you need to separate the color channels.
@ I figured out why. I didn’t select the image to edit correctly.
What software are you using?
PixInsight and Affinity Photo.
Interesting.
I like what you show but you are a bit too fast for my experience level. Do you have any tutorial for editing OSC images with pixinsight and infinity photo. I have trouble with photoshop
Most of the videos from 2023 are utilizing PixInsight and Affinity Photo.
🤯
whenever i import stacked tiffs into affinity they get a non destructive gamma correction that gets applied to the exported image, yours seemed to import without that, did you do something different?
I have never had this issue. I tried looking around the settings to see what's happening but can't find anything on it. Are you using 32 bit? That gives more export options so you may want to check that.