DeepSkyStacker "noise rings"

[dynasty99]

Here's a strange one for you I don't remember having ran into before, but now it has cropped up on multiple image sets I have tried to process, and I don't know the reason:

I seem to see circular bands of noise in my images straight after they come out of DSS. This banding does not seem to be at all visible on any of my subs, only on the stacked image.  This happens when just using light frames (no darks,flats,biases). 

To this post I attached 3 images:

- a single subframe I use in the stack.. I can't detect any noise banding here
- the stacked image straight out of DSS, from about 30 such subframes
- processed image:  I just do some levels and stretching mostly

Has anyone encountered this before? Maybe it is just because I am very light polluted, and pushing short exposures on faint objects too hard (strain your eyes... prizes for whoever identifies what nebula you're looking at), but maybe I'm setting some bad option in DSS.

Equipment:
Nexstar 6SE AltAz mounted

Nikon D3100 @ prime focus

Thanks

[Alien 13]

I have had this too using the same process as you have described but only visible after a hard stretch in PS I am not sure of the cause but have managed to get rid of it using an artificial flat.

Alan

[FunkyKoval35]

It is a normal. This is record of uneven field illumination due to vignetting caused by the optics and due to partial obstructions in the light path such as dust particles on the sensor. To remove it You need to use flat frame files during stacking procedure.

[Astrosurf]

Do you use photoishop? If so, do get GradientXTerminator. It's brilliant. Plus the colour balance option is a marvel: http://www.rc-astro.com/resources/GradientXTerminator/

Alexxx

[dynasty99]

Thanks for the advice everyone! 

I will try using an artificial flat, a real flat.. and maybe I'll look into GradientXTerminator too, but admittedly I'd prefer a free option

Cheers

[Alien 13]

I must admit the 3rd image is very similar to what i have seen and does look more like a diffraction effect with evenly spaced light/dark rings.

Alan

[uhb1966]

And i would also add some dark frames if possible... can make a big difference

[lensman57]

Here's a strange one for you I don't remember having ran into before, but now it has cropped up on multiple image sets I have tried to process, and I don't know the reason:

I seem to see circular bands of noise in my images straight after they come out of DSS. This banding does not seem to be at all visible on any of my subs, only on the stacked image.  This happens when just using light frames (no darks,flats,biases). 

To this post I attached 3 images:

- a single subframe I use in the stack.. I can't detect any noise banding here

- the stacked image straight out of DSS, from about 30 such subframes

- processed image:  I just do some levels and stretching mostly

Has anyone encountered this before? Maybe it is just because I am very light polluted, and pushing short exposures on faint objects too hard (strain your eyes... prizes for whoever identifies what nebula you're looking at), but maybe I'm setting some bad option in DSS.

Equipment:

Nexstar 6SE AltAz mounted

Nikon D3100 @ prime focus

Thanks

These gradients are normal in low SNR captures. You have to flatten the  background by effective flat fielding or in post process using  software.

A.G

[dph1nm]

This is definitely NOT normal. Looks like quantisation or 8/16 bit issue to me. You can get this sometimes if you try to remove a circular gradient from an 8bit image.

NigelM

[dynasty99]

Hi Nigel - that you say is my feeling too, but I don't know enough about the technicalities of 8/16bit images, rasterization, etc. to debug by problem. As far as I know I stay 16 bit the whole way. Images from a Nikon D3100 (.NEF), processed in DSS and saved as 16bit tiff and opened in PS where they show these rings (the rings are visible also in the stacked image preview in DSS).  I see DSS marks the "Depth" of my files as Gray 16 bit. This seems incorrect, no? They should be one-shot-colour, so RGB or something like that shouldn't they? Maybe related..

I have since tried using an articifial flat, which amplifies the banding, and also a genuine flat, which of course helped toss various dust flecks but didn't help the banding issue.
 

Thanks again for all the input everyone! I really appreciate it.

[dynasty99]

Ok, I think I have found a solution, if not the answer to why this was happening.

• From DSS I take Autosave'd 32bit TIFF: modifying the levels in DSS and doing the "Save Image As.." a 32bit tiff with modifications seems to introduce the rings, OR saving directly as a 16bit tiff seems to introduce the rings.
• In PS I immediately can convert to a 16bit TIFF (I don't understand why this opens the HDR Toning dialogue, but I adjust a few sliders to make things look nice before conversion anyways).
• Now the image has quite a non-flat field (even using my flat frames in DSS), and so I apply an artificial flat (median, gaus blur, median, add noise... my usual recipe), and then the image is relatively flat, and free of banding!

Very glad to have this one resolved  .

It's not much compared to the images people post on here, but here is my first image with this junk resolved... M101 (not images from my original post.. I'll reprocess them another time)

Cheers all,

[Astrosurf]

Well done sorting that out! I've made a mental note of it.

[dph1nm]

Rather curious though. I wonder if there is something about Nikon RAW files that DSS isn't handling quite right - or even a bit sticking on in the camera. A histogram of a single RAW sub would probably dispell the latter. In fact histograms of the various stages are probably what is required to get to the bottom of it!

NIgelM