31 replies to this topic

[Borodog]

This is the second of three posts on a processing technique I call the synthetic blue channel method. This first post can be found in the Loony Bin here:

 

https://www.cloudyni...channel-method/

 

I won't go into the details of why this works here, just tell you how to do it and provide some examples. To understand why it works, see the first thread.

 

So you have an achromat that you'd like to image planets with, but it has nasty chromatic aberration. The chromatic aberration produces a blue/violet glow around the planet and turns them yellow and magenta. You can black clip out the blue glow, but the yellow/magenta tone is difficult to deal with. It occurs because the blue channel is out of focus, spreading the blue from the bright highlights around, turning the highlights yellow and turning the shadows blue and the belts and GRS of Jupiter magenta.

 

How can you fix this? Simple. Throw out the blue channel, and replace it with a synthetic one, created from the green and red channels:

 

B_synthetic = 2G - R

 

That's it. Process as normal from there.

 

Some examples.

 

Here are some of my planetary images, imaged with my 1100 EdgeHD. Note that it does not show CA; these are control cases to determine how well the method can reproduce the full color image.

 

Original on the left, synthetic blue channel method on the right.

 

 

 

Note that the reds become a bit more orange on Jupiter. This is pretty clear here because I prefer high saturation images. But the result is not unpleasant.

 

Here's Saturn:

 

Mars:

 

 

Again, these examples do not use ANY of the data from the original blue channel.

 

Next up, actual images from achromats.

[Wdh]

Nice Mike! That explains the new WS version today. I wasn’t sure what the use case was. 

[Borodog]

John Graham provided me this image he took with his wonderful but achromatic 5" Unitron.

 

On the left, original image with chromatic aberration, on the right, with synthetic blue channel method.

 

 

Attached Thumbnails

• 

[Borodog]

Nice Mike! That explains the new WS version today. I wasn’t sure what the use case was. 

Indeed it does!

 

You want to set the red weight to -1, the green to 2, and the blue to 0, although feel free to experiment with other recipes.

[Borodog]

This was imaged last night with my Vixen 102M f/10 achromatic refractor, live-stacked time lapse shot in SharpCap. These are animated PNGs.

 

Original:

 

https://www.cloudyni...8177_511117.png

 

Synthetic blue channel:

 

https://www.cloudyni...8177_563507.png

 

Note that my case was imaged through a William Optic VR-1 filter (VR for Violet Reduction or Violet Reducer), and John's was imaged through a Baader Fringe Killer I believe. For best results you want to use some sort of Fringe Killer/minus violet/violet reduction filter because that eliminates or at least greatly reduces the bleed of defocused blue light into the red and green channels. However, for the planets, because we typically present them in linear light (i.e. we are not applying a strong luminosity stretch/gamma to them), this is not as critical as it would be for for say the Moon or DSO images.

[Borodog]

Note that you can also use this method in cases where the blue channel suffers for other reasons than just chromatic aberration. If the seeing is such that the blue channel is much worse than the green or red channels, this method can recover a higher quality full color image. If your scope suffers from any aberration that results in a soft blue channel, such as a rolled edge or astigmatism, this method will improve the quality of the images.

 

There is no free lunch, of course. The cost here is that we are throwing out the blue signal, which results in a somewhat noisier image. However, the impact of the method on the luminosity is negligible, while the improvement in the colors is dramatic. So paying a small cost in SNR for dramatically improved colors seems like a good trade.

 

Note that you can also do this with mono imaging. You can skip the blue filter entirely and devote that imaging time to G and R instead, saving a spot in your filter wheel, especially since you can also use this method for DSO imaging (more on that in thread number three).

[gstrumol]

Excellent work sir!!  

 

This is amazing stuff.

[Borodog]

Excellent work sir!!

This is amazing stuff.

Thanks. I thought so, but I guess nobody around here is trying to image with an achromat. Shrug.

[Jupiterastro]

That’s too great!!! My refractor (celestron 80dx) has very crazy chromatic aberration looks like this.

And this is how Jupiter looks like through it.

​

[Borodog]

That’s too great!!! My refractor (celestron 80dx) has very crazy chromatic aberration looks like this.

IMG_0074.jpeg

And this is how Jupiter looks like through it.

IMG_0257.jpeg
​ IMG_9993.jpeg

You will definitely want to image through a minus violet filter, as your green and particularly the red channel are also being damaged from defocused blue bleed. It's hard to show the power of the method if you aren't shooting through a minus violet filter, as all of the channels are corrupted. But the red and green Airy patterns indicate that the scope would take great images for its aperture using this method if you shoot through the right filter.

 

[gstrumol]

You can get even better with a little help from GIMP:

 

[Thebonker]

amazing
can you pls help me out with some of my images?

[dcaponeii]

This is the second of three posts on a processing technique I call the synthetic blue channel method. This first post can be found in the Loony Bin here:

 

https://www.cloudyni...channel-method/

 

I won't go into the details of why this works here, just tell you how to do it and provide some examples. To understand why it works, see the first thread.

 

So you have an achromat that you'd like to image planets with, but it has nasty chromatic aberration. The chromatic aberration produces a blue/violet glow around the planet and turns them yellow and magenta. You can black clip out the blue glow, but the yellow/magenta tone is difficult to deal with. It occurs because the blue channel is out of focus, spreading the blue from the bright highlights around, turning the highlights yellow and turning the shadows blue and the belts and GRS of Jupiter magenta.

 

How can you fix this? Simple. Throw out the blue channel, and replace it with a synthetic one, created from the green and red channels:

 

B_synthetic = 2G - R

 

That's it. Process as normal from there.

 

Some examples.

 

Here are some of my planetary images, imaged with my 1100 EdgeHD. Note that it does not show CA; these are control cases to determine how well the method can reproduce the full color image.

 

Original on the left, synthetic blue channel method on the right.

 

Jupiter_synthetic_blue.jpg

 

synthetic_blue_Jupiter.jpg

 

Note that the reds become a bit more orange on Jupiter. This is pretty clear here because I prefer high saturation images. But the result is not unpleasant.

 

Here's Saturn:

synthetic_blue_channel_Saturn.jpg

 

Mars:

 

synthetic_blue_mars.jpg

 

Again, these examples do not use ANY of the data from the original blue channel.

 

Next up, actual images from achromats.

Good work, but I'm not going to lie, I don't see enough difference at those image scales to make the effort.  I don't image with a refractor so it's moot in any case.
 

[dcaponeii]

John Graham provided me this image he took with his wonderful but achromatic 5" Unitron.

 

On the left, original image with chromatic aberration, on the right, with synthetic blue channel method.

Now this pair shows a distinct improvement using the synthetic blue channel method.  Now my question becomes why don't your image pairs in the first panel show any significant difference?

[Borodog]

Now this pair shows a distinct improvement using the synthetic blue channel method. Now my question becomes why don't your image pairs in the first panel show any significant difference?

Because they were imaged with an 11” SCT that does not show CA. The point was to demonstrate how well the method works when you throw out the blue channel entirely, as long as you have good red and green data.

[dcaponeii]

Because they were imaged with an 11” SCT that does not show CA. The point was to demonstrate how well the method works when you throw out the blue channel entirely, as long as you have good red and green data.

Sorry I missed your point.  Not intentional.
 

[Borodog]

amazing
can you pls help me out with some of my images?

A critical piece of the method is that you must have imaged through some sort of fringe killer/minus violet/violet reduction filter. A pale yellow/495 long pass type filter (stacked with an IR cut filter) should also work well. This is because color bleed of defocused blue light into the green and red channels also ruins those channels as well, not just the blue channel. And of course the achromat can't focus IR either, so you have to cut out both ends of the spectrum. You really want clean red and green channels.

[Thebonker]

Oh ok
thanks for the info

[dcaponeii]

A critical piece of the method is that you must have imaged through some sort of fringe killer/minus violet/violet reduction filter. A pale yellow/495 long pass type filter (stacked with an IR cut filter) should also work well. This is because color bleed of defocused blue light into the green and red channels also ruins those channels as well, not just the blue channel. And of course the achromat can't focus IR either, so you have to cut out both ends of the spectrum. You really want clean red and green channels.

OR you can image with a reflecting telescope like God intended (or at least Sir Isaac Newton).  hehe

[Borodog]

A critical piece of the method is that you must have imaged through some sort of fringe killer/minus violet/violet reduction filter. A pale yellow/495 long pass type filter (stacked with an IR cut filter) should also work well. This is because color bleed of defocused blue light into the green and red channels also ruins those channels as well, not just the blue channel. And of course the achromat can't focus IR either, so you have to cut out both ends of the spectrum. You really want clean red and green channels.

I should clarify that only the #8 pale yellow filter has to be stacked with an IR cut filter, but the Baader 495 long pass does not, as it includes an IR cutoff.

[Borodog]

Chromatic aberration correction using the synthetic blue channel method has now been implemented in SharpCap 4.1 as of yesterday's release. It can be found in the right hand pane under preprocessing. I've made a short video explaining how it works and demonstrating with several images, both planetary and deep sky.

 

https://www.youtube....h?v=Hc7Tjcq5WIg

 

Note that because the correction is made in preprocessing, captured SER files are already corrected, and no special treatment is required in Autostakkert or in post. Capture, stack, and post process with an achromat like any other scope.

[gstrumol]

Chromatic aberration correction using the synthetic blue channel method has now been implemented in SharpCap 4.1 as of yesterday's release. It can be found in the right hand pane under preprocessing. I've made a short video explaining how it works and demonstrating with several images, both planetary and deep sky.

 

https://www.youtube....h?v=Hc7Tjcq5WIg

 

Fantastic!! You should be proud   

[John Boudreau]

I should clarify that only the #8 pale yellow filter has to be stacked with an IR cut filter, but the Baader 495 long pass does not, as it includes an IR cutoff.

Actually the Baader 495 doesn't include an IR cut. Neither does the 570 and 610 LP filters in that visual set. The dark and light blue, along with the green filter in that set are bandpass filters and do have a reasonable IR cutoff. Baader's chart is attached.

Attached Thumbnails

• 

Edited by John Boudreau, 31 January 2024 - 11:27 PM.

[Borodog]

Actually the Baader 495 doesn't include an IR cut. Neither does the 570 and 610 LP filters in that visual set. The dark and light blue, along with the green filter in that set are bandpass filters and do have a reasonable IR cutoff. Baader's chart is attached.

Thank you for the correction, John. I specifically read that the 495 LP has an IR cutoff, but I can't recall where.

[Ittaku]

Love this information by the way, even if I don't image with an achro I've always wondered what the equation was for creating a pseudo colour channel for blue. It might just come in handy in RGB mono.