Binning ?? when would be good ?

[beamish]

there you have it !

I'm trying to figure out when would be a good time to Bin, the Lu has the option of binning 640x480 x2, x3, x4 plus other options.

I would really like some advice on what the consequences/pros& cons are please.

I have seen in in posts relating to DSO's that rgb's have been binned but the L hasn't-are there similar situations relating to planetary imaging RGB where binning may be useful ?

Many thanks ,Karlo

[steppenwolf]

The only advantage of binning is when taking R, G and B channel images.

Binning the RGB channel images has the advantage of collecting more light in a shorter period of time but at a lower resolution but this lower resolution doesn't matter.The detail in an image is held in the Luminance (L) channel so this should be taken unbinned. However, the colour derived from the R, G and B channel data is only 'painted' on top of this 'detail' layer so its clarity is not so important.

This means that deeper images can be taken in a shorter period of time. This is more important in DSO imaging than in planetary imaging.

[beamish]

XLNT , many thanks Steve.

No point doing it on planets then ?

[steppenwolf]

I am not a planetary imager but because of the short exposures required to image the planets, I cannot imagine any advantage in binning for them.

[RobH]

I agree with Steve that binning RGB for planetary images shouldn't be neccesary.

Binning RGB on long DSO images though....

it is quicker to get the data, and the lack of resolution when used with an unbinned luminance layer isn't a big issue, but it does have an effect.

This isn't noticeable on extended objects such as nebulae, but on objects where there is a lot of colour detail, such as galaxies, images with binned colour data aren't quite as crisp.

I've moved away from binning these days for this reason...I'd prefer to spend a bit longer and get full resolution detail for everything.

It's only a very slight difference, but it is noticeable if you're a real anorak like me

Cheers

Rob

[blinky]

So would it be a good idea with a one shot colour (thinking my new QHY8 here) to say bin the colour and the unbin an Ha layer to use as Lum?

[beamish]

Cheers guys, so once I move to a cooled cam for DSO's ( could be a while ) then it may be worth considering, but in the meantime ignore it

[martin_h]

So would it be a good idea with a one shot colour (thinking my new QHY8 here) to say bin the colour and the unbin an Ha layer to use as Lum?

That wont work - when you bin a one shot colour you get black and white..OSC uses 4 pixels for the colour 1 red, 1 blue, and 2 green, binning 2x2 joins all those together to form 1 super pixel so the bayer matrix is distroyed.

[steppenwolf]

I've moved away from binning these days for this reason...I'd prefer to spend a bit longer and get full resolution detail for everything.

It's only a very slight difference, but it is noticeable if you're a real anorak like me

No not an anorak - the best quality images need the best quality data and that takes time and patience! As it happens, the only binning I EVER do is when I am focussing as I want a bright image quickly and I want to download it quickly too - a binned 2 x 2 image is 1/4 the size of an unbinned image so it downloads quickly!

[ollypenrice]

Like Rob, I have pretty much given up on binning. It can give you quirky stars and loss of colour detail. Above all, don't bin if aiming to blend in some Ha. The difference in star sizes is too great.

Olly

[syd_malicious]

The one thing not mentioned about binning is the image scale. ie arc seconds per pixel.

The prevailing theory about this is quite simple really, however not all imagers agree . I will probably give a bad description and google will find a better one but here goes...........

Each pixel of a CCD will cover a certain area of sky, this is measured in arc seconds per pixel. The determinants of this are the focal length at which you are imaging and the pixel size. A good way to calculate this is to use CCD Calc.

The optimal resolution for imaging is determined by the quality of the sky at a particular time/elevation/location etc. In Arizona or up a mountain in Hawaii you may be able to image at the sub arc second level. Realistically in the UK the best sky quality we hope for is 2 - 4 arc sec quality (I think).

You need to actually image (approximately) at half the seeing quality. So if the sky is 2 arc pixel per sec you need to actually set the camera to measure 1 arc sec per pixel. For this scenario imaging at 0.5 is oversampling and at over 2 is undersampling.

Oversampling makes guiding unnecessarily difficult and can lead to distorted stars because of guiding/scintillation issues. Undersampling can lead to boxy or pixelated stars.

Physiologically your eyes perceive detail mainly (or better) in mono. So you need the best resolution in mono (ie luminance) and less so in colour.

Therefore the best option is:-

luminance - half the seeing quality. So for the UK this would be 1 - 2 arc sec per pixel.

colour - perhaps 50 - 100% greater which is 2 - 3 arc sec per pixel.

Once you know the value of arc sec per pixel you can actually determine the best binning value.

For my 14" Meade SC (3556 focal length, guiding is err interesting at times ) and Artemis 11002 (9 micron pixels) I bin at 2 for luminance and 3 for colour though I could probably go up to 4x bining for colour.

Hope that seems logical and I am sure there are better descriptions around

Paul

[dph1nm]

The technical advantage from on-chip (hardware) binning CCDs is that you only get one dose of read-noise instead of e.g. 4 when 2x2 binning. If your are read-noise limited then this is a considerable advantage in terms of depth. This does not apply if you are binning in software.

NigelM