Can someone explain binning please.

[White Rabbit]

I hear the word binning thrown around a lot here and it occurred to me that I dont actually know what it means.

For instance.

"HaLRGB 60 90 40 60 50 all 10 minute subs, Ha and Luminance 1x1 binning, colour 2x2"

This was taken from the picture of the week from the welcome page.

While were about it, what does colour 2x2 mean.

Thanks

[Bolts_Tweed]

Gday

As I understand it binning is what the camera thinks the pixel size is. A 1 x 1 binned image means you are imaging with individual pixels. 2 x 2 means you are combining 4 pixels to work as an equivalent larger pixel with each side measurement equalling 2 pixels.

From memory when my camera arrived the tip was that a 2 x 2 binned image will catch fainter nebulosity, galaxy arms etc with an equivalent loss of resolution

Maybe a simplistic explanation but I am pretty sure this is the extremely basic description of it.

Mark

[rogerg]

Mark has pretty much got it right.

Binning increases sensitivity and "effective pixel size" by effectively combining pixels on the chip, like Mark said. 2x2 means 4 pixels are joined as 1, 3x3 means 9 pixels are joined as one, etc.

Take 2x2: The advantages for astronomy are:
1) the light captured by 4 pixels is combined to 1, so you get 4 times the sensitivity/data in each resulting pixel.

2) the effective pixel size of the chip is increased, so instead of a pixel being 7u x 7u it would be 14u x 14u.

3) the download image size is smaller, so frame rate can be faster. a 2x2 image is litterally smaller in terms of the number of pixels digitised and hence downloaded.

Mark is right that you can use binning to achieve images with fainter nebulosity but really you could have achieved the same at 1x1, but just at a longer exposure time.

Examples of when binning are used in astronomy:

1) for focusing - increased sensitivity means shorter focus exposures are needed and smaller file size means quicker turn-around between focus images. Paritcularly useful on older Parallel Port cameras.

2) for colour - using RGB filters you get less data than a clear filter so longer exposure times would be preferred, binning colour at 2x2 while keeping luminance at 1x1 means you have stronger colour signal to improve colour contrast. Also, the colour channels don't need the high resolution data of the luminance channel (it would be better if all was 1x1 but in reality it makes little difference).

Google for "ccd binning" and you'll find lots of diagrams to help explain it.

Hope that helps.

Roger.

[bloodhound31]

Forgive me for pointing out, but for us guys and girls (me included) who find these questions and answers akin to gold nuggets, if I don't know what binning means, then chances are I don't know what other technical terms mean.

As we become more adept at our hobby, we have to be careful that when we offer advice to beginners, we don't confuse them more by adding more jargon.

Being in the machining and fabrication trade, I am assuming that (7u x 7u, 14u x 14u), you mean microns in measurement of size. I have seen it written as (uM), but not as just (U). It's a small thing I know (pun intended) but has potential to leave people scratching their heads even more..:shrug:

Not saying the for the original poster (OP) that this is the case, more for myself or others who might be learning.:thumbsup:

Thanks for the explanations though. I am grateful for it.:welcome:

Baz.

[rogerg]

Yeap, you're right... it's "microns" .. and my knowledge is haphazard enough that I haven't remembered the correct terminology :)

[telecasterguru]

I also appreciate the information on binning as now I understand what it is.
Just to make sure, I set the camera to the binning setting before I take the image and this will depend on the conditions present and the object to be imaged. Planetary 1x1 and Nebulosity 2x2 maybe? Is binning more important at a dark site or under light polluted skies?

Thanks

Frank

[Bolts_Tweed]

Gday Frank

I dont know if there is a general answer to the question on binning nebs or darl / light skies but I can tell you my experiences.

If you want resolution dont bin. Remember you are reducing the resolution of your camera by a factor of 4 when you 2 x 2 bin. Your 10 megapixel CCD immediately becomes 2.5 megapixels with the pixel size 4 times the priginal size - all equals lower resolution. This starts to get into pixel size ratio or scale as well but I will elave that for more astrophotographically eloquent people than me.

I agree with Roger in that if you want more nebulosity or galaxy detail increase then exposure amd take a heap of em. ( iam not going to get into the difference between 100 x 10 second subs versus 10 x 100 second subs but try it - more fainter neb or galactic data will be in the 100 second images but you need to take a heap to account for the signal to noise problems)

I use my camera for astrophotography and have just started supernova hunting with it around the moon time of the month. I 2 x 2 bin my s.n. images for shorter exposures and faster downloads because I am not trying for 'nice' pictures to stick on a wall. However when i am taking 'nice' astrophotos I usually image at 1x1.

I have 2x2 binned Ha, SII, OII channels for colour that sit under a Luminosity channel taken at 1x1 for detail. There is a bit more mucking around with this but narrowband has so much mucking arund a bit more doesnt matter.

So my advice would be if you are just trying to capture a neb from home or if you taking quick framing snaps anywhere then bin them however if you are trying to resolve the pillars in the Eagle Neb or capture detail in the dust lane in that faint galaxy - dont bin but take more and longer exposures at 1 x 1. Others may disagree but I am getting some results following this approach.

Mark

[Gama]

The sensitivity is actually the same. Your not increasing the QE, but increasing the total output signal by adding all 4 pixel wells together.
Each pixel can still only convert a photon to electrical charge on the same efficiency (QE), but its all combined to make a "Super" pixel with a larger combined charge, that will produce images faster.


Theo

[Bassnut]

Theres more to it than that Theo, if it was just combining pixels for "a larger combined charge" then you could convert bin 1 in software to bin 2 after capture and get the same result, the only "faster" part would be the download time. No other advantage at all, you wouldnt bother, given its a lower res.

The whole point of binning is, its done on chip so that readout noise is lower, (as say with bin2) proportionally 4 times less than with bin1.

[Bolts_Tweed]

Yeah Theo - I agree re sensitivity and I am at fault for inferring it does improve the ability to capture fainter "thingos" - it is just a reference to the instruction manual.

For the guys that dont want acronyms and overly technical talk - Theo is pointing out that the ability of the pixel to capture a quantum (ie a 'bit or packet) of light and convert it to voltage (ie the charge generation) is in no way changed by binning. The ability of the pixel to do this is fixed in the physical properties and performance of the electronics that forms the little bucket that is the pixel.

We teach the undergrads at uni that a CCD is like a series of rows of small buckets (ie pixels) that catch rain (light quanta or packets). By putting 4 buckets together you in no way effect the ability however of each of the individual buckets to catch and retain rain (QE or Quantum Efficiency) by adding them together (binning). It is a good analogy for describing a CCD and I have seen it even being used in introductory text books. It can even be taken further to include dumping the buckets into tanks (serial registry) and then measuring the content of each tank - you can obviously do this quicker (ie faster download speed) if you are dumping 4 buckets at a time instead of 1 at a time. Just another quick one is that each bucket has a capacity (full well capacity). It obviusly is more tech than this (and definately gets beyond me very quickly with regard to most of it) but it may be of use to understand some of the tech talk that goes on about CCDs.

Mark

[Bolts_Tweed]

Sry Fred I didnt see your reply - yep read out noise is also reduced - (maybe it fits as spillage from the buckets to the tank and from the tank to the measurement system in the previous analogy)

Mark

[higginsdj]

Converting in software is not as accurate as doing it on the camera (it increases conversion errors by the same factor as the binning). Binning is most advantageous for photometric work rather than pretty pictures as it effectively increases SNR as opposed to sensitivity.

[Omaroo]

2x2 and 3x3 binning are, in a practical sense, rather good for taking quick snaps (as the image size is smaller - both dimensionally and in bytes) to see if you have your image centred/rotated as you'd like it. Image capture and download times are smaller than with 1x1 set. Once you're happy with your composition, then progress to normal progamming.

[MrB]

It's actually pretty common.
The 'u' is used in place of the correct one where the descender(tail) is on the left, or 'μ' the greek letter 'Mu'

In electronics literature, μF (microfarads) is commonly shortened to just μ too.
hmmm, μtoo.... good name for a band

[White Rabbit]

Thanks for all the replys, I'll print this whole thing off and read it when I have more time.

Thanks
Sandy