Now that I see these are useful for more than just my students, I will. Most of the ones I've made are more for MaxIm DL and for research students, but I will expand!
Great video! I am new to LRGB and have been working with paint.net. I think I can do most of this in paint.net, but am downloading gimp and raw therapee so I can use fits files, too.
I too found this a very helpful video. However, I share other commenters' view that a more objective colour calibration methodology would be ideal, particularly as I am red/green colour blind!
Yes, this is a simple, quick method and is not perfect--certainly won't be as close to true color as a mathematical color calibration method would be! Glad it was helpful on the whole, though. Edit: One thing you can do that's not too complicated is look up one of the stars in your field, find its spectral type, and apply ratios to the red, green, and blue images to represent how bright each should be based on how bright that star is in each of those images. You can also use the sky background brightness levels as an indicator of how bright the red, green, and blue images are compared to each other--MaxIm DL has a tool that lets you do exactly that automatically. That said, I strongly prefer the appearance of our final images when they have undergone the non-linear stretch in each filter to enhance color contrast. We are running another astrophysical imaging class this semester, and last time we did delve into this a little bit--maybe it will be a good opportunity to create some new content regarding this. I know there are tons of sources for this out there, too. I'll dig around and see--does anyone else already know of a good video tutorial for mathematical color balancing in astrophotography?
Hi there, Thank you for providing the video. It is a nice basic explanation on creating colour images from monochrome channels. It is good to see GIMP being used for astronomical image processing. I did not like the 'very liberal' colour adjustments and think there should be more work done in getting the colours right. Colour calibration should not be taken lightly and the method displayed here is very subjective and dependant on the colour settings of the monitor. Proper calibration can be achieved by the colour values of stars with known temperatures. May I also offer a nice way to align the layers over the top of each other. The way I do it is taking the difference between the layers, that way when the image is aligned it will be black. Besides that, the video and the way the content was presented was great. Keep up the good work. Regards, Renato Langersek
You have a great speaking voice. The tutorial sounds like poetry. Suggestion: add how to do a star mask and what it is for. Hopefully this will get me doing something with tri-color imaging. Two years and not one successful image. I find the processing language illiterate. All of the programs are a heap of disorganized tools with no logical association. Thanks!
Thank you for this very useful and informative tutorial. I am learning GIMP and have a long way to go. But there is progress, after just a few hours I'm now able to spell it correctly. At 11:25, and thinking about at my 102 mm refractor, I chuckled when she says 'only have a 14 inch'. One thing I can say has to do with aligning two layers. Set all but two layers (the two you want to align) to non-display (using the eyeball in the layers panel) and set the Mode to Difference. Doing this will cause the displayed image to be darkest when alignment between the two layers is best. Then set one of the aligned layers to non-display and display another layer and repeat the alignment (making sure the Mode is set to Difference).
Do you know how/can you make a video on processing a NB SHO image in GIMP? That would be very helpful as there are not any I can find on RUclips. I tried to process one last week without success and was only able to salvage the Ha on the Cynus Wall. Thanks again.
It seems that the choice of color saturation is subjective, leading to color results that are not a true representation of the nebula - most specifically the use of the non-linear curve tool. Is there a more scientific method of assuring accurate color representation (short of comparing to other one-shot color)? I would seem to think that bringing in each color at identical luminance values would provide a true color representation (although that is what you did and the nebula turned green). So, again, how can we get true representaive color? One way that might work is to match the luminance of each channel to a spectral graph and play with the saturation tool until the luminace values match those in the spectragraph.
@johnbspringer "true representative color" The immediate difficulty is how you define that. Even if your eye had enough resolving power to see the nebula directly, it would look gray because there's not enough light for your eye to see color. Even your idea of a spectrograph is arguably not true color because the spectrograph won't match your eye's response over the visual range, which is decidedly non-linear. And, finally, the many color illusions you can find on the net demonstrate that your brain's interpretation of the visual field plays a large role in the colors you see. In my own astrophotography, I really don't worry about "true" color; I'm not sure you can even quantify to what extent different viewers will see the same colors in the same image, anyway. "Visually appealing" is the best you can aim for. Having said all of that, one interesting technique used by some astrophotographers is to photograph a star field with with stars of the same spectral type as the Sun (G2), on the notion that a G2 star defines "white" (thanks to evolution). Exposures through an RGB filter set thus allow them to calculate how long to expose each filter such that the same G2 star reaches the same brightness in each filter. But even that kind of careful work can yield uncertain results by the time you subtract dark current from the images and balance the sky background of the various colors. And then you probably still wouldn't match what your eye would see if you could just somehow turn up the brightness of the nebula.
@@pseudorandomly I did not know that. Thank you. If I look visually through the most powerfull optical telescope with just my eye, will I be able to see color? When a one-shot color camera picks up color, is that for all practical purposes the approximate color of the nebula? Meaning if our eys had the resolving power, would we see approaximaly the same color as the one-shot color cameras?
@@johnbspringer "If I look visually through the most powerfull optical telescope ..." The short answer to that question is "yes", though you're not going to see full-blown Walt Disney-esque color; you'll be able to detect *some* color, though. Even smaller telescopes allow some (not all) people to see a faint greenish tinge to the Orion Nebula, and some folks (not me!) can see bluish tints in some other nebulae. The interesting thing about photography is that folks do not expect their photographs to exactly match reality; they like their photographs (and their TV) to have a bluer picture than real life. I recall reading that back in the days of film, Kodak created a film that was very close to matching colors as perceived by eye for daylight images, and people complained. In any case, the CCD imager in your camera only approximates what your eye would see, as you say. The CCD chip itself is more sensitive on the red end and less so on the blue end. Whether that's good enough "for all practical purposes" is a judgement call. There are even more difficulties. For some classes of objects, the light emitted falls in very narrow bands of the visual range, rather than broad-spectrum color. Your eye is not uniformly sensitive to these narrow bands, whereas a CCD camera would be much more so. Some objects put out essentially *only* one very narrow-band red color that a CCD camera hardly picks up at all without modification, because it's blocked by a filter covering the CCD chip. (Somewhat ironically, the filter is there to help the other color balance look "more real" by cutting back on some of the red!) And that is not to mention the color response of a printer or a computer monitor or viewscreen used to view the image. Finally, you may recall the internet-famous picture of a dress that some people saw as white and gold while others saw blue and black. There's a picture of a tennis shoe going around now that is either pink and white or green and gray. So if you can't even photograph dresses and tennis shoes in "true" color, faint astronomical objects have almost no chance! You may have thought you asked a simple question! I believe "Dr. Becky" Smethurst, an astrophysicist here you RUclips, has a video about the reality of color in images of astronomical objects. You may find it interesting.
What is the CLEAR image? How can I produce that? Usually we have R G B channel window, and a RGB combined window. I can extract the 3 channel individually in SIRIL but what is that CLEAR image?
We have an Astrodon Luminance filter in our filter wheel, which passes a large range of the color spectrum from a bit of ultraviolet through a good deal of the infrared spectrum. So it is like a combination of all of the brightness not only in red, green, and blue, but also a little bit of UV and IR. It gives a gorgeously high amount of signal-to-noise ratio compared to the color filters.
Whoops--so in other words, we are using 4 filters when we image, not three! We use luminance and photometric red, green and blue. Our red, green, and blue filters are likely different from most astrophotographers' out there since ours are the kind intended for scientific measurement, not specifically for astrophotography or to mimic human eyesight.
I processed my first LRBG in GIMP today with help from your tutorial. Much appreciated.
Great tutorial, very well explained and easy to understand 😊
Exceptionally useful. Great tutorial, 👍
Just created my first Astro image in GIMP. Thank you for putting this together.
Thank you for sharing great tips on LRGB editing with GIMP. Thanks to you, I was able to create full-color image from LRGB images!
This was a very helpful tutorial.
This is helping me during my first week of astrophotography .
Clear instructions after a week of no sleep.
Ty
Thanks, just took my first set of LRGB data and this walked me through the process very nicely.
Great video!
this is one of the best Gimp tutorials. please please make one more. just one please.
Now that I see these are useful for more than just my students, I will. Most of the ones I've made are more for MaxIm DL and for research students, but I will expand!
Big thanks!!!
Great video. Simple and effective.
Great video! I am new to LRGB and have been working with paint.net. I think I can do most of this in paint.net, but am downloading gimp and raw therapee so I can use fits files, too.
thank you thank you
Thank you, that was very helpful!
Very helpful. - Thank you!
I too found this a very helpful video. However, I share other commenters' view that a more objective colour calibration methodology would be ideal, particularly as I am red/green colour blind!
Yes, this is a simple, quick method and is not perfect--certainly won't be as close to true color as a mathematical color calibration method would be! Glad it was helpful on the whole, though. Edit: One thing you can do that's not too complicated is look up one of the stars in your field, find its spectral type, and apply ratios to the red, green, and blue images to represent how bright each should be based on how bright that star is in each of those images. You can also use the sky background brightness levels as an indicator of how bright the red, green, and blue images are compared to each other--MaxIm DL has a tool that lets you do exactly that automatically. That said, I strongly prefer the appearance of our final images when they have undergone the non-linear stretch in each filter to enhance color contrast.
We are running another astrophysical imaging class this semester, and last time we did delve into this a little bit--maybe it will be a good opportunity to create some new content regarding this. I know there are tons of sources for this out there, too. I'll dig around and see--does anyone else already know of a good video tutorial for mathematical color balancing in astrophotography?
Hi there,
Thank you for providing the video. It is a nice basic explanation on creating colour images from monochrome channels. It is good to see GIMP being used for astronomical image processing.
I did not like the 'very liberal' colour adjustments and think there should be more work done in getting the colours right. Colour calibration should not be taken lightly and the method displayed here is very subjective and dependant on the colour settings of the monitor. Proper calibration can be achieved by the colour values of stars with known temperatures.
May I also offer a nice way to align the layers over the top of each other. The way I do it is taking the difference between the layers, that way when the image is aligned it will be black.
Besides that, the video and the way the content was presented was great.
Keep up the good work.
Regards,
Renato Langersek
RenatoVK4TNT How do you do this? Thank you.
@@johnroskruge2752, which part?
RenatoVK4TNT aligning the layers, I used the method in the tutorial. Many thanks.
You have a great speaking voice. The tutorial sounds like poetry.
Suggestion: add how to do a star mask and what it is for.
Hopefully this will get me doing something with tri-color imaging. Two years and not one successful image. I find the processing language illiterate. All of the programs are a heap of disorganized tools with no logical association.
Thanks!
Thank you for this very useful and informative tutorial. I am learning GIMP and have a long way to go. But there is progress, after just a few hours I'm now able to spell it correctly.
At 11:25, and thinking about at my 102 mm refractor, I chuckled when she says 'only have a 14 inch'.
One thing I can say has to do with aligning two layers. Set all but two layers (the two you want to align) to non-display (using the eyeball in the layers panel) and set the Mode to Difference. Doing this will cause the displayed image to be darkest when alignment between the two layers is best. Then set one of the aligned layers to non-display and display another layer and repeat the alignment (making sure the Mode is set to Difference).
Do you know how/can you make a video on processing a NB SHO image in GIMP? That would be very helpful as there are not any I can find on RUclips. I tried to process one last week without success and was only able to salvage the Ha on the Cynus Wall. Thanks again.
It seems that the choice of color saturation is subjective, leading to color results that are not a true representation of the nebula - most specifically the use of the non-linear curve tool. Is there a more scientific method of assuring accurate color representation (short of comparing to other one-shot color)? I would seem to think that bringing in each color at identical luminance values would provide a true color representation (although that is what you did and the nebula turned green). So, again, how can we get true representaive color? One way that might work is to match the luminance of each channel to a spectral graph and play with the saturation tool until the luminace values match those in the spectragraph.
@johnbspringer "true representative color"
The immediate difficulty is how you define that. Even if your eye had enough resolving power to see the nebula directly, it would look gray because there's not enough light for your eye to see color. Even your idea of a spectrograph is arguably not true color because the spectrograph won't match your eye's response over the visual range, which is decidedly non-linear. And, finally, the many color illusions you can find on the net demonstrate that your brain's interpretation of the visual field plays a large role in the colors you see.
In my own astrophotography, I really don't worry about "true" color; I'm not sure you can even quantify to what extent different viewers will see the same colors in the same image, anyway. "Visually appealing" is the best you can aim for.
Having said all of that, one interesting technique used by some astrophotographers is to photograph a star field with with stars of the same spectral type as the Sun (G2), on the notion that a G2 star defines "white" (thanks to evolution). Exposures through an RGB filter set thus allow them to calculate how long to expose each filter such that the same G2 star reaches the same brightness in each filter.
But even that kind of careful work can yield uncertain results by the time you subtract dark current from the images and balance the sky background of the various colors. And then you probably still wouldn't match what your eye would see if you could just somehow turn up the brightness of the nebula.
@@pseudorandomly I did not know that. Thank you. If I look visually through the most powerfull optical telescope with just my eye, will I be able to see color? When a one-shot color camera picks up color, is that for all practical purposes the approximate color of the nebula? Meaning if our eys had the resolving power, would we see approaximaly the same color as the one-shot color cameras?
@@johnbspringer "If I look visually through the most powerfull optical telescope ..."
The short answer to that question is "yes", though you're not going to see full-blown Walt Disney-esque color; you'll be able to detect *some* color, though. Even smaller telescopes allow some (not all) people to see a faint greenish tinge to the Orion Nebula, and some folks (not me!) can see bluish tints in some other nebulae.
The interesting thing about photography is that folks do not expect their photographs to exactly match reality; they like their photographs (and their TV) to have a bluer picture than real life. I recall reading that back in the days of film, Kodak created a film that was very close to matching colors as perceived by eye for daylight images, and people complained.
In any case, the CCD imager in your camera only approximates what your eye would see, as you say. The CCD chip itself is more sensitive on the red end and less so on the blue end. Whether that's good enough "for all practical purposes" is a judgement call.
There are even more difficulties. For some classes of objects, the light emitted falls in very narrow bands of the visual range, rather than broad-spectrum color. Your eye is not uniformly sensitive to these narrow bands, whereas a CCD camera would be much more so. Some objects put out essentially *only* one very narrow-band red color that a CCD camera hardly picks up at all without modification, because it's blocked by a filter covering the CCD chip. (Somewhat ironically, the filter is there to help the other color balance look "more real" by cutting back on some of the red!)
And that is not to mention the color response of a printer or a computer monitor or viewscreen used to view the image.
Finally, you may recall the internet-famous picture of a dress that some people saw as white and gold while others saw blue and black. There's a picture of a tennis shoe going around now that is either pink and white or green and gray. So if you can't even photograph dresses and tennis shoes in "true" color, faint astronomical objects have almost no chance!
You may have thought you asked a simple question!
I believe "Dr. Becky" Smethurst, an astrophysicist here you RUclips, has a video about the reality of color in images of astronomical objects. You may find it interesting.
what is a clear stack image?
What is the CLEAR image? How can I produce that? Usually we have R G B channel window, and a RGB combined window. I can extract the 3 channel individually in SIRIL but what is that CLEAR image?
I created a monochrome channel with SIRIL to use as luminance layer. Seems to works in a way. Thank you lady. The RGB part is enough to me.
We have an Astrodon Luminance filter in our filter wheel, which passes a large range of the color spectrum from a bit of ultraviolet through a good deal of the infrared spectrum. So it is like a combination of all of the brightness not only in red, green, and blue, but also a little bit of UV and IR. It gives a gorgeously high amount of signal-to-noise ratio compared to the color filters.
Whoops--so in other words, we are using 4 filters when we image, not three! We use luminance and photometric red, green and blue. Our red, green, and blue filters are likely different from most astrophotographers' out there since ours are the kind intended for scientific measurement, not specifically for astrophotography or to mimic human eyesight.