OSC noise patterns

[rob]

Hi,

Is there an easy way of removing the 'rivers' noise patterns that are prevalent with one shot colour cameras (both cooled and DSLR) ? In my case the image is of a star cluster which has no surrounding dust or nebulosity. These systems can be very fussy when it comes to darks.

Thanks
Robert

[admin]

Hi,

Is there an easy way of removing the 'rivers' noise patterns that are prevalent with one shot colour cameras (both cooled and DSLR) ? In my case the image is of a star cluster which has no surrounding dust or nebulosity. These systems can be very fussy when it comes to darks.

Thanks
Robert

Hi Robert,

Would it be possible for you to attach an example of what we're dealing with? It would really help!

Thanks,

[rob]

I've attached a small fragment of the image showing the issue.

[rob]

I used the upload attachment but I can't see it - I hope you can ?

[admin]

Sorry, Rob.
Can't see it - could you try clicking the 'Place inline' button after uploading?

I've been trying to think what sort of pattern noise can cause 'rivers' appear in your data - the only thing I can think of that would look like that would be AC noise from a faulty or cheap AC adapter... Even so, a lot of averaging of your dark/bias frames should take care of that?

[rob]

Sorry it's a bit small but the limit is 512kb and that doesn't go far in TIFF land.

NGC7789_rivers_crop.tiff (478.68 KiB) Viewed 11997 times

[gboulton]

Try dropbox.

[rob]

Great idea, never used it before but this is the link to the 36Mb full tiff

https://www.dropbox.com/s/vzns50qc24bed ... ivers.tiff

And a baby version: https://www.dropbox.com/s/ryxca4nrfs0cs ... _crop.tiff

[admin]

Thanks so much for taking the time to upload your data Rob - it really helps!

I'm assuming the data you posted has already been stretched and you're interested in removing the colour noise in the background?

If so, You will probably attain the best results using the Wavelet De-Noise module (preferrably while you are still in Tracking mode, but non-tracking will also work). Bump up Color Detail loss and you should see the colour noise disappear. Play with the scale parameters to home in on the 'size' of the noise blotches/grain (I just usually start off with 100% for all scales).
If you're worried about destroying colour in non-background parts, set brightness mask power to 1.00 and play around with its value. What this does is progressively apply noise reduction to darker parts of your image. The value here is how linear this response is. High values keep progressively more noise and only drops off at the last moment, whereas low values < 1.0 drop off quickly and remove noise even in the brighter parts. A value of 1.0 is 'linear'; the noise reduction ramps up 1:1 as pixels get darker.

Does this answer your question?

[rob]

Hi,

Yes it does. I can get the noise down quite a lot with that method.

Thankyou very much.

Rob