#1  
Old 03-02-2017, 08:54 AM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
galaxies should look noisy?

I have been desperately trying to get enough signal to minimise noise in the smooth regions of nearby galaxies. Maybe that was a futile exercise, since it appears that, once there is enough signal to get rid of read and shot noise, what is left really should still look noisy due to the star structure. Even though the stars are not resolved, they introduce a level of statistical roughness sometimes referred to as "incipient resolution", that looks like random noise, but does not smooth out with more data. This surface brightness structure has been studied in some detail and measured as a way of probing distant stellar populations and as a secondary measure of distance.
explanation in https://ned.ipac.caltech.edu/level5/...Jacoby9_1.html

I assume that the same process applies to the underlying background galaxies in the night sky, so maybe the background sky should never actually look smooth either - perhaps smooth galaxies and smooth sky is not realistic and I have been chasing an unattainable goal thanks for looking. regards Ray.

edit: by the way, the rest of that NED publication is an interesting read

Last edited by Shiraz; 04-02-2017 at 01:48 AM.
Reply With Quote
  #2  
Old 03-02-2017, 09:45 AM
glend (Glen)
Registered User

glend is offline
 
Join Date: Jun 2013
Location: Lake Macquarie
Posts: 7,033
Pretty heavy reading for my old brain but a couple if things occurred fo me: based on King's work it would seem that blue filters would give you better resolution. Secondly, looking at the pixel graphic, would not smaller pixels provide better resolution in general, by reducing the number of stars contributing to individual pixels. This seems contrary to the general assumption that large pixels are better DSOs at longish focal lengths.
Reply With Quote
  #3  
Old 03-02-2017, 11:50 AM
Slawomir's Avatar
Slawomir (Suavi)
Registered User

Slawomir is offline
 
Join Date: Sep 2014
Location: North Queensland
Posts: 3,240
Thank you Ray for bringing this up - very interesting read.

I think stretching data alters it in such way that background noise generally gets amplified, thus some sensible noise reduction can be applied.

Personally I use noise reduction less and less and my affinity with noise-reduced smooth astroimages weakens.
Reply With Quote
  #4  
Old 04-02-2017, 12:03 AM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
Quote:
Originally Posted by glend View Post
Pretty heavy reading for my old brain but a couple if things occurred fo me: based on King's work it would seem that blue filters would give you better resolution. Secondly, looking at the pixel graphic, would not smaller pixels provide better resolution in general, by reducing the number of stars contributing to individual pixels. This seems contrary to the general assumption that large pixels are better DSOs at longish focal lengths.
I read it that King showed how the unresolved star noise was less intrusive in the blue (due to the star population having a wavelength dependence) - made it easier to study globular clusters.
Agree that smaller pixels give better resolution. In the end though, the sky will limit how good it can be and in typical Aus seeing, somewhere between 0.5 and 1 arcsec pixel scale would seem to be about the optimum. Big pixels are needed with long focal length scopes, but they are not inherently better at DSO - there is always a tradeoff between resolution (small) and sensitivity (big).

Quote:
Originally Posted by Slawomir View Post
Personally I use noise reduction less and less and my affinity with noise-reduced smooth astroimages weakens.
am coming to the same conclusion

Last edited by Shiraz; 04-02-2017 at 09:56 AM.
Reply With Quote
  #5  
Old 13-02-2017, 10:39 AM
sil's Avatar
sil (Steve)
Not even a speck of dust

sil is offline
 
Join Date: Jun 2012
Location: Canberra
Posts: 1,474
I think there is a misunderstanding of what noise reduction means. To some people it means elimination/removal when it simply means Noise REDUCTION. The process of reduction is never perfect and in itself it introduces artifacts/noise to the image. Whether you notice it is another matter.

There is also a lot of signal in the "background" regions of an astrophoto that people destroy when they push the background to black. When you bring up the midtones to reveal signal you are also amplifying noise too. Its up to the person processing to determine a suitable cutoff point so they get the signal they want and reduce intruding noise... its a personal taste of the contrast between the two for most images which is why most are useless for photometric analysis.

I've seen and tried so many "optimal" settings rules people declare must be done and have yet to find any that actually work. Instead they may be useful starting points and as I process I make adjustments and measurements to feed back to my starting point settings and its an iterative refinement process to find "optimal-for-me" settings.

What works for you, works for you. Its about exploring options and understanding how they effect your final images. Too many people think there is a secret tool that does it all for you, the only tool is the person processing.
Reply With Quote
  #6  
Old 13-02-2017, 11:34 AM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
thanks Sil. agree with your comments on noise reduction.

However, I have been trying to get enough signal so that I do not use any noise reduction at all - with a lot of signal, the shot noise is inherently low re the signal and noise reduction should not be needed. However, for nearby galaxies, I have been finding that the galaxy itself does not ever get to look smooth, even when there is buckets of signal and a high signal to noise ratio. The linked article points out that this is exactly how it should look - there is a lumpiness in the signal from nearby galaxies that makes them look noisy, even when there is no shot or read noise to speak of. The distribution of the stars within pixels results in signal roughness, even when the stars are not resolved.

this does not apply to far off galaxies or to true smooth extended objects (eg nebulae) - it only applies to nearby galaxies where there is some variation in the number of included bright galactic stars from pixel to pixel - and I suspect that it probably applies to the background sky when going deep and where the numbers of unresolved deep galaxies and dim milky way stars in the pixels will vary.

from the linked article: "An image of an elliptical galaxy with milli-arcsec resolution would look like an enormous globular cluster, but even when the resolution is a thousand times worse, the discreteness of the stars causes measurable bumpiness in its surface brightness. This phenomenon, dubbed ``surface brightness fluctuations'', has been recognized for many years and is sometimes referred to as ``incipient resolution''.
Reply With Quote
  #7  
Old 13-02-2017, 03:23 PM
sil's Avatar
sil (Steve)
Not even a speck of dust

sil is offline
 
Join Date: Jun 2012
Location: Canberra
Posts: 1,474
The maths in that article is a bit beyond me but its totally believable. My own imaging is limited and I can't get to use gear like others here use (and gear I own too). Anyway the lumpiness is definitely a resolution artifact and its there, its not noise to overcome in processing. So when I was imaging with shorter focal lengths I could capture faint fuzzies which sort of looked lumpy or grainy at their small size on the frame (these are things that weren't my primary imaging target that happened to be in the field of view. but the lumps resolved to structure when I imaged at a longer focal length and now it had a similar sized lumpiness at what I thought might be smooth signal. Like looking at a photo of a beach looks smooth, but you go there in person and footprints etc are now obvious but still there is smoothness there until you lie down with a magnifying glass the smoothness gets lumpy again at the limits of the resolution of your eyeballs or camera. its the fractal nature of the universe. So yeah you shouldn't crop your signal to get a smooth background.

I made a shot of the southern cross which almost looks like snow because I pulled up so many dim stars which are really there. Even stellarium etc has limits on the magnitudes it can show you and as the Hubble showed us, no matter how empty a patch of sky looks, there is stuff there. its not a flat even black, so yes I'd expect to perceive some unevenous to all parts of an astrophoto. If you zoom in you can tell the difference between blotchy signal and chroma/lum noise. It helps having a good monitor and being conservative when processing to avoid clipping anything if possible.
Reply With Quote
  #8  
Old 13-02-2017, 04:09 PM
Slawomir's Avatar
Slawomir (Suavi)
Registered User

Slawomir is offline
 
Join Date: Sep 2014
Location: North Queensland
Posts: 3,240
I am wondering about magnitude of these fluctuations in terms of pixel brightness within a galaxy. My guess is that these are minuscule and it is most likely that shot noise (as well as noise from the camera) normally drowns them, and one would possibly need unusually long integration to quantify these fluctuations. I am talking amateur imaging of course.
Reply With Quote
  #9  
Old 13-02-2017, 05:43 PM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
Quote:
Originally Posted by Slawomir View Post
I am wondering about magnitude of these fluctuations in terms of pixel brightness within a galaxy. My guess is that these are minuscule and it is most likely that shot noise (as well as noise from the camera) normally drowns them, and one would possibly need unusually long integration to quantify these fluctuations. I am talking amateur imaging of course.
Not sure Suavi. From reading, SNR of maybe 5 is adequate for the surface brightness fluctuations to be noticeable with a close galaxy - my system will do that quite easily in a few hours with a typical sky. The source papers are a bit heavy for "idle curiosity" reading, but, from a skim over, it seems that the effect is noted in very short (eg 1 minute) images from big scopes at 1 arcsec seeing, so it does not feel like it is stretching thing to assume that we will see similar in multi-hour images at 2 arcseconds. In any event, I started digging around because I ran into "noise" that did not smooth out with more exposure - I now think that I may have already seen this effect on a few galaxy images (this one in particular http://www.astrobin.com/full/276705/0/?real=&mod=). cheers

Last edited by Shiraz; 13-02-2017 at 05:59 PM.
Reply With Quote
  #10  
Old 13-02-2017, 05:47 PM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
Quote:
Originally Posted by sil View Post
The maths in that article is a bit beyond me but its totally believable. My own imaging is limited and I can't get to use gear like others here use (and gear I own too). Anyway the lumpiness is definitely a resolution artifact and its there, its not noise to overcome in processing. So when I was imaging with shorter focal lengths I could capture faint fuzzies which sort of looked lumpy or grainy at their small size on the frame (these are things that weren't my primary imaging target that happened to be in the field of view. but the lumps resolved to structure when I imaged at a longer focal length and now it had a similar sized lumpiness at what I thought might be smooth signal. Like looking at a photo of a beach looks smooth, but you go there in person and footprints etc are now obvious but still there is smoothness there until you lie down with a magnifying glass the smoothness gets lumpy again at the limits of the resolution of your eyeballs or camera. its the fractal nature of the universe. So yeah you shouldn't crop your signal to get a smooth background.

I made a shot of the southern cross which almost looks like snow because I pulled up so many dim stars which are really there. Even stellarium etc has limits on the magnitudes it can show you and as the Hubble showed us, no matter how empty a patch of sky looks, there is stuff there. its not a flat even black, so yes I'd expect to perceive some unevenous to all parts of an astrophoto. If you zoom in you can tell the difference between blotchy signal and chroma/lum noise. It helps having a good monitor and being conservative when processing to avoid clipping anything if possible.
I think we all may have seen the effect - I like your beach sand analogy.

fwiw, from some skim reading, found one paper showing how this effect could allow Hubble to probe galaxies as far away as ~300MLY and another showing how information could be extracted on deep unresolved background galaxies using the statistics of sky surface brightness fluctuations in images from ground-based telescopes.
Reply With Quote
  #11  
Old 13-02-2017, 06:41 PM
Slawomir's Avatar
Slawomir (Suavi)
Registered User

Slawomir is offline
 
Join Date: Sep 2014
Location: North Queensland
Posts: 3,240
Quote:
Originally Posted by Shiraz View Post
Not sure Suavi. From reading, SNR of maybe 5 is adequate for the surface brightness fluctuations to be noticeable with a close galaxy - my system will do that quite easily in a few hours with a typical sky. The source papers are a bit heavy for "idle curiosity" reading, but, from a skim over, it seems that the effect is noted in very short (eg 1 minute) images from big scopes at 1 arcsec seeing, so it does not feel like it is stretching thing to assume that we will see similar in multi-hour images at 2 arcseconds. In any event, I started digging around because I ran into "noise" that did not smooth out with more exposure - I now think that I may have already seen this effect on a few galaxy images (this one in particular http://www.astrobin.com/full/276705/0/?real=&mod=). cheers
I stand corrected. Had no idea about the real magnitude of these variations. Well, in the end I am a narrowband guy
Reply With Quote
  #12  
Old 13-02-2017, 07:44 PM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
Quote:
Originally Posted by Slawomir View Post
I stand corrected. Had no idea about the real magnitude of these variations. Well, in the end I am a narrowband guy
Errr I don't have much idea either, but - with just enough knowledge to be dangerous - it looks like it might explain something that I couldn't understand - that'l do me.
Reply With Quote
Reply

Bookmarks

Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +10. The time is now 03:05 PM.

Powered by vBulletin Version 3.8.7 | Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Advertisement
Testar
Advertisement
Bintel
Advertisement