What are these red/green/blue streaks?

[imakebeer]

I was out imaging M31 Andromeda Galaxy last night - Canon 450D + BackyardEOS mounted on a SW150PDS + HEQ5, 30 x 60s exposures at ISO 400.....

Now the battery in the DSLR died before I was finished so I only took light frames, I didn't get a change to do any dark frames.

I've run the light frames I have through Sequator to stack them and now in GIMP to play around some more - from Sequator I can see there are these red, green and blue streaks. I've doctored the image below in GIMP to make them stand out. They look like streaks of coloured rain running from top right to bottom left. Any idea what is causing them?

[ONIKKINEN]

Did you guide or dither? This is walking noise, and dithering is the cure. Guiding alleviates it somewhat but dithering is the real fix.

You can dither without guiding too, just move both of the axis a good 20+ pixels every few shots. Ideally once per 5min, but every dither helps in the end if you cant be bothered to do that every 5 minutes.

Edited December 8, 2022 by ONIKKINEN

[Rallemikken]

13 minutes ago, imakebeer said:

Any idea what is causing them?

I'd say they are hot or dead pixels in the sensor. Recognize the colors. This is why we take darks. When the stacking software arranges the frames according to the stars these pixels will move accordinly, and often forms patterns on the stacked image. Streaks in this case, can take any shape, but will be the same everywhere.

[Padraic M]

I think they are likely to be dead pixels in the camera - hot pixels in this case, where the pixel is 'always on' or 'always bright'. You always get them on DSLR sensors, and I'm sure other sensors too, and you may get more as the sensor ages. Your calibration frames will sort them out.

If you don't have calibration frames for this particular run, you could try stacking using 'Sigma Clip' algorithm, it's pretty good at removing artefacts like this. Otherwise, Gimp Clone or Heal is your friend!

Dithering is important too, but even better if you don't mix them into the stack in the first place (by calibration).

[imakebeer]

Thanks @Rallemikken @Padraic M 👍 Looking through the light frames I can see the dead pixels now, they are fixed in the frame as the target moves a little. I guess sequator has made them trail when I stacked - I'll see if there is a setting to deal with them. As I mentioned, had my battery not died I'd have done some dark frames too.

@ONIKKINEN I don't quite understand the question. The mount was tracking although it is unguided at the moment (on my to do list!). I've heard of dithering - I've see the option in Autostakkert when processing planetary videos but beyond this I don't know what it is. Something else to learn!

[ONIKKINEN]

4 minutes ago, imakebeer said:

Thanks @Rallemikken @Padraic M 👍 Looking through the light frames I can see the dead pixels now, they are fixed in the frame as the target moves a little. I guess sequator has made them trail when I stacked - I'll see if there is a setting to deal with them. As I mentioned, had my battery not died I'd have done some dark frames too.

@ONIKKINEN I don't quite understand the question. The mount was tracking although it is unguided at the moment (on my to do list!). I've heard of dithering - I've see the option in Autostakkert when processing planetary videos but beyond this I don't know what it is. Something else to learn!

Dithering is a random motion that is done between frames, which will make hot pixels like these ones disappear with rejection stacking. Yours are very pronounced because of no dithering and so they appear as a line (because of polar alignment error induced drift and or other sources of misaligbment). If you are guiding you can set the guider to do this automatically between frames, if not youll have to do it yourself. Some software also allow a "direct dither" without guiding.

Others have suggested darks, which will also remove hot pixels but this is not a substitute for dithers. Darks also should not be used with canon cameras as they do internal calibration of their own that results in 2048 median pixel value regardless of dark exposure length or sensor temperature (so the darks are pointless and only add noise).

So in short, darks remove hot pixels, but you will still need to dither to remove walking pattern noise completely. Dithering does both, so dithering is the ultimate fix.

[pipnina]

11 minutes ago, imakebeer said:

Thanks @Rallemikken @Padraic M 👍 Looking through the light frames I can see the dead pixels now, they are fixed in the frame as the target moves a little. I guess sequator has made them trail when I stacked - I'll see if there is a setting to deal with them. As I mentioned, had my battery not died I'd have done some dark frames too.

If your stacking program has an averaging mode called "sigma clipping" or similar, try that with a sigma of 2 or less, I found it very good at removing these hot pixel trails.

Good luck!

[imakebeer]

1 hour ago, pipnina said:

If your stacking program has an averaging mode called "sigma clipping" or similar, try that with a sigma of 2 or less, I found it very good at removing these hot pixel trails.

Good luck!

Yeah, I've just run it through Sequator (like DSS except user friendly and easy!) again and changed the Composition mode from Accumulation to Select Best Pixels. Seems to have done the job 👍

Edit: I went back to sequator and "Composition mode = Accumulation" + "Remove dynamic noises = On" seemed to do a better job

Edited December 8, 2022 by imakebeer
Added details

[imakebeer]

Thanks again for your advice folks 👍 If you're interested, I've posted the finished pic here - M31 isn't great in the end but I'm pleased with M42 😁

[TheThing]

Well, I think they are pretty good!  I've never done any deep sky imaging but, I'd be 'over the moon' with those images.

[newbie alert]

It's not walking noise in the traditional sense... It's hot/ medium/ cold dead pixels walking with the tracking inaccuracies