Equipment used:
- Lumix G85
- Lumix 100-300 f4-f5.6
- SA-GTI
Image:
- 10 sec, f/6.3, 6400 ISO
- 1200 Lights
- 50 Biases, Darks, Flats
Stacked using Siril and edited in GIMP
As you can see, there is a lot of noise, even with heavy editing and noise reduction in siril as well as in GIMP, and I'm out of ideas of what could cause it. I tried different exposures (from 10s to 30s), different ISOs (1600-6400). I tried to manually dither by moving the mount through the App every 30 minutes or so. I also used a lens dew heater at different temperatures ranging from 15°C to -5°C with no difference in noise.
I got a successful result a few months ago with the Orion nebula, the only differences were the exposure of 5 seconds and a Omegon Mount MiniTrack LX3 Essentials mount which couldn't be polar aligned very well. I can't imagine that the difference from 5 to 10 seconds exposure did all the difference.
Can anyone help me?

8 Comments
Zer0_F0x@lemmy.world · 6 pts · 231d
Try longer exposures, your mount can do 30 sec easily with a 300mm lens, even without guiding.
Also, go through your subs and remove anything that looks out of the ordinary. Better to have 600 good subs than 1200 with bad ones in the mix. Look for clouds and condensation mostly.
This sort of noise increases the closer to the horizon the target gets. Stay above 30 degrees if possible.
Check if your white balance is set to auto, if it is set it to something fixed.
Lastly, judging by the end result you should try a less aggressive stretch, especially for the background. You can make a mask of the nebulous part, stretch it a lot and then use it as a luminance and saturation mask over a copy layer that has a neutralized, darker background.
There's lots of good signal in there, give it a go
Linsensuppe@feddit.org · 1 pts · 231d
I tried imaging with 30 seconds exposure recently on the Andromeda Galaxy and got this result with almost identical noise:
This was 315 subs at 30 seconds so around 2.5 hours of exposure.
So just increasing the exposure did not do a difference.
Zer0_F0x@lemmy.world · 2 pts · 231d
Check each channel individually. Are the noise levels and shapes different for red, green and blue?
Linsensuppe@feddit.org · 1 pts · 231d
In the final .fit the Red channel has, as expected, the fewest signal, but also the most noise in the dark areas of the image.
In the Raw subs I couldn't make out a channel with more noise than another.
vastunknown@lemmy.world · 2 pts · 216d
Late to the party but wanted to help:
What you are seeing is a fixed pattern noise that is actually being reinforced by your many, short subs. This is called "walking noise", especially common with cameras of this class when used for AP due to their sensor noise characteristics.
You must dither more frequently. For non-dedicated astronomers camera, you will also want to dither a fairly large amount. Dithering every 30 minutes is effectively not dithering at all.
Things you should aim for to help fix your image:
lefty7283@lemmy.world · 1 pts · 231d
It could also be that you're not dithering often enough. Dithering every 30 mins means you're taking 180 frames before each dither and that'll really let the walking/raining noise build up. you could try doing 30" subs and then dithering every ~10 minutes if that's practical? For me personally I dither after a max of 5 subs (or dither after a sequence of something like LLLLRRGGBB) , but I know I'm also taking longer exposures and have the benefit of automation.
Personally I've never used a GTi or Lumix camera, but if you're able to hook them to a computer with NINA, there is an option to have it dither automatically even without an autoguider
Linsensuppe@feddit.org · 1 pts · 230d
AFAIK NINA doesn't work with my camera, but since my mount is within Wi-Fi reach of me, I can move the GTI through the app slightly from inside. My question with that is, how much do I have to/should I dither? Currently, I move the stars so that there is a significant jump on the SLR screen, so I would guess it's a few hundred pixels in x and y direction.
lefty7283@lemmy.world · 2 pts · 230d
Few hundred pixels is definitely enough, and kinda overkill (mine goes a few dozen). I think the frequency of the dithers is probably the main thing that could help with the noise