#1  
Old 15-12-2015, 05:35 AM
konstantinos75
Registered User

konstantinos75 is offline
 
Join Date: Oct 2014
Location: Athens,Greece
Posts: 60
To Bin or not to Bin

Hi

I would like to raise this interesting topic and hear the experts opinion about this.

After some research I have done, when someone is doing astrophotography with a CCD camera attached to the telescope, it is advisable to

* bin 2x2 the exposures with R,G,B filters; this will increase the SNR of the final RGB image; it saves you a lot of time, as exposure time is reduced to half comparing to binning 1x1. For example, if with bin 1x1 you need 5 minutes exposure, with bin 2x2 you need 2.5 minutes to get the same result.

* bin 2x2 is better be done from dark skies for taking many exposures in less time.
From light polluted skies is better go with bin 1x1.

* if you need to combine the R,G,B frames with the L frame, you must not use bin 2x2 for the R,G,B frames and bin 1x1 for L frame because in this way you will loose in resolution of the final LRGB image. Instead use bin 1x1 for all L,R,G,B frames.

* for narrowband filters you must bin 2x2 for Ha, SII, OIII to get the best result in less exposure time. But in case you need to combine these filters with L you must go with the bin 1x1 rule for all filters.

* if the image scale of your imaging device is less than 0.5 arcsec/pixel (for example you have a long focal length telescope) is better bin 2x2 all the L,R,G,B,narrowband frames.

I would like to see if these assumptions can be justified and if someone has some arguments about this.

Thank you
Konstantinos

Last edited by konstantinos75; 15-12-2015 at 05:56 AM.
Reply With Quote
  #2  
Old 15-12-2015, 07:52 AM
Slawomir's Avatar
Slawomir (Suavi)
Registered User

Slawomir is offline
 
Join Date: Sep 2014
Location: North Queensland
Posts: 3,240
Quote:
Originally Posted by konstantinos75 View Post
* for narrowband filters you must bin 2x2 for Ha, SII, OIII to get the best result in less exposure time. But in case you need to combine these filters with L you must go with the bin 1x1 rule for all filters.
Hi Konstantinos,

I am no expert so I will be very brief

From my experience in narrowband imaging from a light polluted area, it is best to use as narrow filters as possible and expose 1x1 for as long as practical, if one is after faint detail and quality data.

Regards
S
Reply With Quote
  #3  
Old 15-12-2015, 08:12 AM
Paul Haese's Avatar
Paul Haese
Registered User

Paul Haese is offline
 
Join Date: Jan 2009
Location: Adelaide
Posts: 9,937
I don't use binning at all anymore. I tried using it for a couple of years but found it did not work for my needs. I have the luxury of darks skies so I don't need to hurry the data collection.

One benefit of using unbinned data is being able to easily combine my RGB data with narrow band data.
Reply With Quote
  #4  
Old 15-12-2015, 08:43 AM
LewisM's Avatar
LewisM
Novichok test rabbit

LewisM is offline
 
Join Date: Aug 2012
Location: Somewhere in the cosmos...
Posts: 10,388
Never did bin, and I am always in a hurry. I don't like the image size typically of binned images - an 8300 binned image is too small IMHO. I have experimented with binning the fits after acquisition, and whilst it seems to boost some things, I feel it loses more, especially in fidelity.

So, even though I am perpetually rushing and under severe LP, I don't bin.
Reply With Quote
  #5  
Old 15-12-2015, 11:08 AM
PRejto's Avatar
PRejto (Peter)
Registered User

PRejto is offline
 
Join Date: Jan 2011
Location: Rylstone, NSW, Australia
Posts: 1,397
The luminance signal needs to have full resolution (i.e., unbinned), but colour data can be lower resolution and theoretically the eye will take the detail from the luminance.

Up until recently I didn't have the best luck binning colour 2x2, but I also didn't try all that hard. I recently bought a Trius color camera and have used that data to combine with high quality luminance. In some ways one could argue that the colour data from the Trius-c is not even as good as a mono Trius binned 2x2 given that out of an array of 4 pix there is only 1 red, 1 blue and 2 greens. However, I think this result looks pretty good:

http://www.astrobin.com/full/232099/0/

By the way, CCDStack has no problem at all combining binned images with un-binned images. It does so automatically.

Peter
Reply With Quote
  #6  
Old 15-12-2015, 03:22 PM
konstantinos75
Registered User

konstantinos75 is offline
 
Join Date: Oct 2014
Location: Athens,Greece
Posts: 60
Hi

Check out these experimented results by Sara Wager

http://swagastro.weebly.com/to-bin-or-not-to-bin.html

According to her conclusion: "Based on this analysis, for me the binning argument makes sense. It makes my imaging setup 2.93x more sensitive, so I will collect more faint data. For me this is important. There is a downside of course........ My sensor size of 3380 x 2704 has just halved. But by increasing the resolution of the images in Photoshop to 300DPI, I don't see that as a major problem."

So I would be particularly interested if someone else has followed the experiment path and has some data facts to present. A picture is 1000 words according to an old Chinese proverb

Thank you for your contribution to this topic
Reply With Quote
  #7  
Old 15-12-2015, 05:26 PM
ericwbenson (Eric)
Registered User

ericwbenson is offline
 
Join Date: Sep 2009
Location: Adelaide, Australia
Posts: 209
Quote:
Originally Posted by konstantinos75 View Post
Hi

Check out these experimented results by Sara Wager

http://swagastro.weebly.com/to-bin-or-not-to-bin.html

According to her conclusion: "Based on this analysis, for me the binning argument makes sense. It makes my imaging setup 2.93x more sensitive, so I will collect more faint data. For me this is important. There is a downside of course........ My sensor size of 3380 x 2704 has just halved. But by increasing the resolution of the images in Photoshop to 300DPI, I don't see that as a major problem."

So I would be particularly interested if someone else has followed the experiment path and has some data facts to present. A picture is 1000 words according to an old Chinese proverb

Thank you for your contribution to this topic
Hello,
This is issue has been covered many times, short version there is no easy answer...

IMO the analysis in the link is flawed, for one thing it compares raw ADU counts (and calls this sensitivity?) in small pixels versus large 2x2 superpixels, it should be comparing SNR to SNR. Then of course you have the problem of spatial normalization, the small pixels may implicitly have more information via higher spatial resolution (hence higher useful signal) but it depends on the image scale/seeing/optics etc...except not in their case since the 1x1 image scale is 0.47"/pixel, very possibly oversampled, so perhaps they should bin all the time an not just for color subs Although really I can't tell much difference from the side by side binned versus unbinned shots.

Other points, not all sensors benefit from binning (the read noise just goes up by x4 for some CCDs). With sensors where it can be advantageous, binning only affects the subexposure length, there is no magic gain in SNR for the same total integration time. In light polluted areas, binning usually does NOT make sense since you reach sky limited in such a short time anyways (and you spend more time downloading than integrating). For narrowband it does make sense to bin since the sky flux is so low, but it is ironic since most are trying to maximize resolution in that case!

Best
EB
Reply With Quote
  #8  
Old 15-12-2015, 06:52 PM
gregbradley's Avatar
gregbradley
Registered User

gregbradley is offline
 
Join Date: Feb 2006
Location: Sydney
Posts: 17,877
I agree the argument is not that clear cut.
It depends on the sensor, your seeing, the quality of your optics and the aperture of your scope. Your available imaging time is also a huge factor.
For example the popular KAF8300 does not do proper 2x2 binning as it has a hardware limitation that limits it to a 50% increase not 4X.
I don't know for sure but I imagine if your camera is only 2 megapixels binning makes .5 megapixels and that is pretty low res. So you need enough pixels to make it work also.

I find I get excellent 2x2 binned Ha images from my CDK and 16803. It goes super deep super fast.

But I imagine that same camera on my Honders binned may not be so impressive as 12 inches aperture and F3.8 does not really need 2x2 to capture the signal.

The little Sony sensors I think have a good case for binning at times. Like hers, the 3.69 micron pixels would be too small for most people's scopes and seeing. So binning makes sense and no wonder there is an improvement.

The bottom line is if you have a question about whether its good or not for you is to try it out and observe the differences.

My experience also is that binning 2x2 tends to mean rounder stars for some reason. But they are a bit larger when I flick from 1x1 to 2x2 images of the same object.

2x2 makes sense if you want to do deep and you have enough aperture to retain high resolution.

It also makes sense if your available imaging time is limited due to weather, having to setup each night or clouds/moon. In which case any image is better than no image.

Greg.
Reply With Quote
  #9  
Old 15-12-2015, 08:27 PM
alpal's Avatar
alpal
Registered User

alpal is offline
 
Join Date: Jan 2012
Location: Melbourne
Posts: 3,602
Here is a picture showing the difference in binning.
I found that I had to bin at 2x2 & expose for 900 seconds to get any of the spiral arms to show up on NGC 300.
I use a KAF8300 sensor with a 10" f4 Newt.

The final LRGB picture is here:
https://www.flickr.com/photos/247194...in/photostream


cheers
Allan
Attached Thumbnails
Click for full-size image (Binning comparison NGC 300_3.jpg)
153.1 KB61 views
Reply With Quote
  #10  
Old 15-12-2015, 08:52 PM
ericwbenson (Eric)
Registered User

ericwbenson is offline
 
Join Date: Sep 2009
Location: Adelaide, Australia
Posts: 209
Quote:
Originally Posted by alpal View Post
I use a KAF8300 sensor with a 10" f4 Newt.
Hi Allan,
Out of curiosity have you by chance calculated your "optimum" subexposure length (e.g. using the online calculators)? One would think that at ~1.1"/pixel unbinned your sampling would be fine for most locations.
Regards,
EB
Reply With Quote
  #11  
Old 15-12-2015, 09:01 PM
alpal's Avatar
alpal
Registered User

alpal is offline
 
Join Date: Jan 2012
Location: Melbourne
Posts: 3,602
Quote:
Originally Posted by ericwbenson View Post
Hi Allan,
Out of curiosity have you by chance calculated your "optimum" subexposure length (e.g. using the online calculators)? One would think that at ~1.1"/pixel unbinned your sampling would be fine for most locations.
Regards,
EB

Hi Eric,
No - I don't use those calculators.
I just try it out & see what happens.

Obviously in this case I tried a 600 second exposure at 1x1
& it didn't show the galaxy arms.
A 600s 2x2 exposure showed a fair bit more
so then I did a 900 second 2x2 exposure where I started to see the galaxy arms
so I took 6 luminance frames in the limited time before the clouds came over.
NGC 300 is a difficult target from a light polluted location.
I could probably do 2 minute exposures on Orion & get even the faintest detail.
BTW - I use a 2" Astronomic CLS-CCD filter in front of all other filters
owing to the sodium lights all down the street where I live.

cheers
Allan
Reply With Quote
  #12  
Old 16-12-2015, 06:58 AM
konstantinos75
Registered User

konstantinos75 is offline
 
Join Date: Oct 2014
Location: Athens,Greece
Posts: 60
Hi

Thank you all for your valuable input regarding this interesting topic.

Let me add some references from bibliography about binning.

(1)

"In LRGB cases where this occurs, taking an unfiltered luminance frame using full sensor resolution and creating the individual RGB images in a binned 2×2 pixel mode will improve the SNR."

Reference:
Creating and Enhancing Digital Astro Images
Grant Privett
Patrick Moore’s Practical Astronomy Series ISSN 1617-7185
Springer Science Business Media,2007


(2)

"Obtain your luminance frames when the skies are most steady and obtain your binned RGB frames when skies are more turbulent.
"you can acquire the clear exposures binned 2 × 2 to improve your signalto-noise ratio in the dim areas."

"If you are using an H-alpha filter, consider binning your H-alpha channel 2 × 2 if needed to speed up light collection."
"Furthermore, most monochrome cameras are easy to bin 2 × 2, 3 × 3, or even 9 × 9 to allow very short exposures of deep sky objects to check centering and framing, which can be critical when trying to locate your target on a small CCD chip."
"You can gather light more efficiently if you bin your red, green, and blue channels to collect color information more rapidly while acquiring luminance information at full resolution. Because color data can be acquired with lower resolution, the color exposures in an LRGB sequence can be binned 2 × 2, accumulating in 15 min the same signal-to-noise as a 1-h color exposure unbinned."

"For very dim objects, you can obtain your clear luminance exposures binned 2 × 2 and your RGB exposures binned 3 × 3."

Reference:
The 100 Best Targets for Astrophotography
Ruben Kier
© Springer Science+Business Media, LLC 2009
Patrick Moore’s Practical Astronomy Series

(3)

"This is especially true for the commonly used strategy of binning color data with the ultimate goal of combining
it with unbinned luminance data, which may be an acceptable approach for a well sampled image but can lead to loss of valuable color information when the image is already on the verge of being critically undersampled, as is typical of
many wide field images."
"However for dim galaxies and nebulae binning the detector can yield significant time savings by lowering the read noise and making the camera act as if it is more sensitive."
"A final word about binned data. Many think that binning 2 Χ 2 automatically doubles SNR, but in reality the benefits of binning are reduced as light pollution increases."
"Even though RGB data is more forgiving than Lum data for LRGB imaging, and the reduced resolution with 2 Χ 2 binning won’t be noticeable in most LRGB images, we still shoot everything unbinned."

Reference:
The Patrick Moore Practical Astronomy Series
Lessons from the Masters Current Concepts in Astronomical Image Processing
Robert Gendler Editor
© Springer Science+Business Media New York 2013

(4)
"Luminance and color layers binned 1x1 (no binning). This provides a sharp image and good color. Highly recommended,
as it will provide the sharpest detail and excellent color. Requires additional imaging time relative to 2x2 binned color components."

"Luminance 1x1 and color layers 2x2, combined in MaxIm DL. Star images are very slightly larger when FWHM is measured, but still good."
"The main advantage of binned color components is that they take less time to acquire. If time is limited, use 1x1 binning for your luminance image and 2x2 binning for your color images."
"For higher resolution and longer exposure times, use 1x1 binning for all images."
"The ability to process the luminance image independently of the color components is a significant advantage of LRGB imaging versus RGB imaging. However, RGB images tend to have richer color, so you should choose the color combination method based on what is more important to you: color (RGB) or details (LRGB). If you want the best of both
worlds, either shoot your LRGB with the color images binned 1x1 instead of the more usual 2x2, or take longer and more RGB images to reduce noise and improve color accuracy."

Reference:
The New CCD Astronomy
Ron Wodaski
New Astronomy Press
http://www.newastro.com
© 2002 Ron Wodaski

(5)
"The argument proposes that for any given overall imaging time, a better result is obtained by using RGB at 1×1 binning than L (1×1) and RGB (2×2)."
"Binning is however, a useful technique to improve the quality of weak signals, not only for color exposures but also when used for expediency during focusing and plate solving."

Reference:
The Astrophotography Manual
A Practical and Scientific Approach to Deep Space Imaging
Chris Woodhouse
© 2016 Chris Woodhouse



So My understanding is that if someone has the possibility to go for 1x1 binning for all filters this is the choice to go.
2x2 or 3x3 binning is preferable for very dim objects.
Reply With Quote
  #13  
Old 16-12-2015, 09:04 AM
gregbradley's Avatar
gregbradley
Registered User

gregbradley is offline
 
Join Date: Feb 2006
Location: Sydney
Posts: 17,877
Another approach is a hybrid luminance of both 1x1 and 2x2 using the 2x2 to bring more depth and the 1x1 for more sharpness and detail.
Unfiltered luminance will give you bloated stars. Not real bad but noticeable. I have a clear filter which has no UV/IR coating like luminance normally does and that's what it does. I think you do gain a bit more sensitivity from using it but not that great.

Greg.
Reply With Quote
  #14  
Old 17-12-2015, 09:26 AM
ags_'s Avatar
ags_ (Phil)
Registered User

ags_ is offline
 
Join Date: Sep 2007
Location: Rubyvale Qld
Posts: 417
I only ever binned CH4 on a 14" sct never RGB. Using my 20" I now don't need to bin CH4.
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 02:08 AM.

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