Author Topic: Mottles in Red channel  (Read 176 times)

Offline rdryfoos

  • PixInsight Old Hand
  • ****
  • Posts: 334
    • View Profile
Mottles in Red channel
« on: 2019 April 17 08:21:27 »
Does anyone know what causes chromatic mottling in higher signal areas of an RGB (SHO) image?  I say high signal areas to mean non background regions.  I am not referring mottled background.  Take a look at the crop from the Rosette nebula in Hubble palette.  I do not always have this problem--but sometimes.  I tried combining these stacks 2 ways--first without linear fitting to the Ha stack (the brightest) and then after linear fitting to the Ha stack (per the Light Vortex Video that indicates this is the proper way--not the way we were taught at the PI seminar, so I have my doubts).  But--regardless, I get mottles that show up in the red channel either way.  I have inspected all the channels, and it turns out that when scrolling through the RG and B channels when looking at the image, the mottles are in all channels--but they manifest themselves in the red channel (in blue for instance they look like dark regions, and in green, less dark).  But in the simplest terms--it is mottling in the red channel and I have tried everything to remove it as well as to prevent it and nothing seems to work.

Thanks,
Rodd




Offline STEVE333

  • PixInsight Addict
  • ***
  • Posts: 182
    • View Profile
    • sk-images
Re: Mottles in Red channel
« Reply #1 on: 2019 April 17 10:01:50 »
Hi Rodd - Sounds like a frustrating problem.

Just a shot in the dark: Do you dither with your guiding? I know that not dithering can cause streaks which could look like mottling.

Steve
WO Star71 ii
Canon DSLR (modified)
http://www.SteveKing.Pictures/

Offline rdryfoos

  • PixInsight Old Hand
  • ****
  • Posts: 334
    • View Profile
Re: Mottles in Red channel
« Reply #2 on: 2019 April 17 10:06:57 »
Hi Rodd - Sounds like a frustrating problem.

Just a shot in the dark: Do you dither with your guiding? I know that not dithering can cause streaks which could look like mottling.

Steve
  Yes--I dither.  I though it might have something to do with to high a value in DBE (top line of the 3-cant recall the name--default is .5).  But I tried lowering this as far as possible without elimination of points and the mottles were still there.
Rodd

Offline STEVE333

  • PixInsight Addict
  • ***
  • Posts: 182
    • View Profile
    • sk-images
Re: Mottles in Red channel
« Reply #3 on: 2019 April 17 11:39:39 »
Yes--I dither.  I though it might have something to do with to high a value in DBE (top line of the 3-cant recall the name--default is .5).  But I tried lowering this as far as possible without elimination of points and the mottles were still there.
Rodd
Some of the tutorials showing the use of DBE have had that parameter (Tolerance) as high as 5.0 without any bad effects.

I would be surprised if DBE were causing the problem. I have never had any success removing "local" bumps or mottling with this process. I use it primarily for removing my local Light Pollution which is typically a fairly smooth gradient across the image.

Hopefully you can find a way to eliminate the mottling from occurring in your Raw images.

Steve

WO Star71 ii
Canon DSLR (modified)
http://www.SteveKing.Pictures/

Offline rdryfoos

  • PixInsight Old Hand
  • ****
  • Posts: 334
    • View Profile
Re: Mottles in Red channel
« Reply #4 on: 2019 April 17 11:48:33 »
Yes--I dither.  I though it might have something to do with to high a value in DBE (top line of the 3-cant recall the name--default is .5).  But I tried lowering this as far as possible without elimination of points and the mottles were still there.
Rodd
Some of the tutorials showing the use of DBE have had that parameter (Tolerance) as high as 5.0 without any bad effects.

I would be surprised if DBE were causing the problem. I have never had any success removing "local" bumps or mottling with this process. I use it primarily for removing my local Light Pollution which is typically a fairly smooth gradient across the image.

Hopefully you can find a way to eliminate the mottling from occurring in your Raw images.

Steve
  Me too--but I have no idea where to start.
Rodd

Offline ngc1535

  • PixInsight Addict
  • ***
  • Posts: 236
    • View Profile
Re: Mottles in Red channel
« Reply #5 on: 2019 April 17 14:27:49 »
Typically color patches (mottling) comes from upsampled chrominance data that is registered with a luminance image. Another way to see the mottle is through some forms of interpolation (bicubic for example)  usually when registering images. These may not be the issue in your case.

If the issue is of the type of smearing small scale color noise to larger scales- one of the solutions is to minimize the noise in the combined color channels images before any other processing. I have found that MURE Denoise is particularly good at this very thing! It despeckles data... which leads to less color mottling later.

Although the issue in your data isn't in the sky ... it definitely isn't "High" signal-to-noise based on your image. Something that measures (in a linear sense) .2-.3 is probably the minimum for qualifying as a significant signal.

-adam

Offline rdryfoos

  • PixInsight Old Hand
  • ****
  • Posts: 334
    • View Profile
Re: Mottles in Red channel
« Reply #6 on: 2019 April 17 15:05:46 »
Typically color patches (mottling) comes from upsampled chrominance data that is registered with a luminance image. Another way to see the mottle is through some forms of interpolation (bicubic for example)  usually when registering images. These may not be the issue in your case.

If the issue is of the type of smearing small scale color noise to larger scales- one of the solutions is to minimize the noise in the combined color channels images before any other processing. I have found that MURE Denoise is particularly good at this very thing! It despeckles data... which leads to less color mottling later.

Although the issue in your data isn't in the sky ... it definitely isn't "High" signal-to-noise based on your image. Something that measures (in a linear sense) .2-.3 is probably the minimum for qualifying as a significant signal.

-adam
Thanks, though not sure what you mean. I registered the frames to the lowest Fehn sub which I don’t remember what it was but it was not lum obviously. I tried mire seniors as well and it did not eliminate the mottled. The mottled become evident with a saturation boost- just a small one produces, or shows, them

Offline rdryfoos

  • PixInsight Old Hand
  • ****
  • Posts: 334
    • View Profile
Re: Mottles in Red channel
« Reply #7 on: 2019 April 18 12:30:42 »
To provide a bit more information, and to put this is a ascale perspective, here is the link to teh full image

https://www.astrobin.com/362038/T/?real=&nc=user

As is evident, the mottles are at a small scale--which even at full resolution is not terribly noticeable.  But when processing, I often zoom way in to assess my handiwork.  Still trying to determine the possible sources: imbalance in the chroma channels?, LP?  ill advised work flow?   Poor data acquisition skills?
Rodd

Offline STEVE333

  • PixInsight Addict
  • ***
  • Posts: 182
    • View Profile
    • sk-images
Re: Mottles in Red channel
« Reply #8 on: 2019 April 18 13:00:26 »
That is a beautiful image Rodd. Nice details and lovely color.

I really don't see the problem. If you are not happy with the noise, is it possible you just need a bit more data (although 8 hrs isn't bad)?

Steve

WO Star71 ii
Canon DSLR (modified)
http://www.SteveKing.Pictures/

Offline rdryfoos

  • PixInsight Old Hand
  • ****
  • Posts: 334
    • View Profile
Re: Mottles in Red channel
« Reply #9 on: 2019 April 18 13:08:34 »
That is a beautiful image Rodd. Nice details and lovely color.

I really don't see the problem. If you are not happy with the noise, is it possible you just need a bit more data (although 8 hrs isn't bad)?

Steve
  If you look closely at the left hand portion of the image--the red area--you can see the mottling if you really look hard.  Mind you--this is after a literal D-Day scale invasion of everything I could throw at it.  It definitely should not be there.  More data might be the answer, though i don't think so.  I did a rosette with 18 hours--mottles were still there.  Besides, I see grand images of these targets with 3-4 hours without mottles, so there must be a root cause.  It could be sky conditions.  Or optics flaws.  I sent the DSQ 106 back for inspection due to distorted stars.  Maybe the mottles are another artifact of the same flaw. 

Offline STEVE333

  • PixInsight Addict
  • ***
  • Posts: 182
    • View Profile
    • sk-images
Re: Mottles in Red channel
« Reply #10 on: 2019 April 18 15:38:04 »
If you look closely at the left hand portion of the image--the red area--you can see the mottling if you really look hard.  Mind you--this is after a literal D-Day scale invasion of everything I could throw at it.  It definitely should not be there.  More data might be the answer, though i don't think so.  I did a rosette with 18 hours--mottles were still there.  Besides, I see grand images of these targets with 3-4 hours without mottles, so there must be a root cause.  It could be sky conditions.  Or optics flaws.  I sent the DSQ 106 back for inspection due to distorted stars.  Maybe the mottles are another artifact of the same flaw.

I think I see what you are talking about Rodd.

I have quite a bit of light pollution at my location. I mention this, because, the LP causes quite a bit of noise in my images. If I am too aggressive in reducing the noise in the Linear processing I will invariably end up with "blotchy noise" in the final finished image. I don't know if this applies to your situation, but, thought it worth mentioning just in case.

Best of success in finding the source of your mottling.

I hope to see more of your images.

Steve
WO Star71 ii
Canon DSLR (modified)
http://www.SteveKing.Pictures/