To bin or not to bin...

[joecoyle]

Hi

These questions have probably been asked /answered a thousand times, but every time I search for “bin” I get posts about binoculars, so I’m going to ask anyway. 

What exactly is binning? I know it is something to do with combining pixels to make super pixels, but what is the aim? Improved sensitivity of the CCD at the expense of image size? To capture faint detail?

Should I be binning my subs? I normally get good 5 minute subs.  If so, which ones? Just RGB or narrowband too? What about Luminance ?

Would binning affect my processing in Pixinsight / do I need to do anything special when stacking images? Does it scale them automatically?

Cheers

Joe 

Ps I use a skywatcher 80ED and an Atik 414EX. 

[carastro]

The main purpose of binning is to speed up imaging time.  You can capture the same amount of data in half the time with 2 x binning, but at the expense of resolution.  Yes you are right, I believe 4 pixels are combined together to form one large one.

Some never bin, but I guess they have copious dark skies, but for those of us who get restricted time to image or get to a darker location, I must admit I do bin most images, but only the colour component.  Luminance should never be binned as it is where you get the detail.  I also don't bin Ha for the same reason. 

Yes the images will be half the size of the non binned images, and they will need re-sizing before you can combine the filters.

Hope this has answered your question.

Carole 

[joecoyle]

Hi Carole,

thats great thanks. I suppose it’s just a question of trying it out now and seeing what happens. 
 

thanks

Joe

[carastro]

By the way, you will need binned darks flats and bias to match your binned data. 

Carole 

[MarkAR]

34 minutes ago, carastro said:

By the way, you will need binned darks flats and bias to match your binned data. 

Carole 

Doh! I need to take more calibration frames for my binned subs.

[joecoyle]

4 hours ago, carastro said:

By the way, you will need binned darks flats and bias to match your binned data. 

Carole 

Oh. Maybe not just a quick try it out then

Flats is a whole new topic - I got a flat box light panel for Christmas 2018. Never once used it.  Don't even know where to begin

[carastro]

Flats are IMO the most important part of imaging as they remove dust and vignetting.  It is really difficult to process an image without flats being applied.

You have an Atik414.  Stick the light panel over the aperture without having moved the camera at all between taking of the images and taking of flats.  Capture VERY short subs getting the histogram approx 1/3 to 1/2 way across from the left.  Or if you have a reading on your software, aim for between 22,000 and 30,000 ADU.

If the light is too bright dim it down with layers of white paper or cloth.  If it is not bright enough, just take a fractionally longer sub.

HTH

Carole 

Edited April 20, 2020 by carastro

[RayD]

42 minutes ago, carastro said:

Flats are IMO the most important part of imaging as they remove dust and vignetting.

Couldn't agree more (well, with the exception of the light frames of course 🙂).

When doing flats and bias frames remember they don't take long to do at all, so take lots of them to make a really decent master of each. 

[joecoyle]

1 hour ago, carastro said:

Flats are IMO the most important part of imaging as they remove dust and vignetting.  It is really difficult to process an image without flats being applied.

You have an Atik414.  Stick the light panel over the aperture without having moved the camera at all between taking of the images and taking of flats.  Capture VERY short subs getting the histogram approx 1/3 to 1/2 way across from the left.  Or if you have a reading on your software, aim for between 22,000 and 30,000 ADU.

If the light is too bright dim it down with layers of white paper or cloth.  If it is not bright enough, just take a fractionally longer sub.

HTH

Carole 

Yeah - I just noticed a nasty great dust doughnut on my Leo triplet! 

So this is my dilemma - last night for example, I set a sequence going (it was 11pm before the triplet came into view, and thought oh, nice clear night, I'll let it run til 3am.)... at the end of my sequences, my mount is parked and camera warmed up. Because of this, I can never take flats because the mount has always moved home... unless you mean don't park the mount/warm the camera, and get up at 3am and take them then?

I value my sleep far too much

Joe

[carastro]

Doesn't matter where the mount is positioned.  

Camera doesn't need to be cooled.

You need to find a time to do the flats if you want good results.  Flats can be done the following day provided you haven't moved the camera/scope relationship (or the dust won't line up).

I once went to an overnighter and had derigged most of my kit but luckily hadn't moved the camera when I remembered I hadn't done my flats.  I packed the telescope and camera intact (very carefully) and when i came home I managed to do the flats.

If you have a permanent set up it is easier.  Some people do flats when the clouds come over for a short while to while away the time.

Also if you have a permanent set up and leave everything set up (camera unmoved) you can re-use the flats.

Flats can be done in the daytime, so long as there is no light leak, I often use the (not too bright) sky for doing flats.  

Carole 

Edited April 20, 2020 by carastro

[RayD]

To underline Carole's point, I am not ready for imaging yet as it isn't dark enough, so I have the roof of the observatory closed and I am running the flats for this sequence as I have just changed my camera.  Easy to do to, takes about 10 minutes tops and leaves plenty of time to get my main sequence ready to run.  You don't have to do them at the end of the sequence, they can be done at the beginning also (or any time if your camera hasn't been moved).

Edited April 20, 2020 by RayD

[joecoyle]

Oh. I thought parking the mount would count as moving the camera too as dust could move during the park. major misconception on my part then d’oh!!! 

i do have a permanent setup. Got a roll off roof shed obsy. :)

I might be more confident about giving flats a go now then! 
 

Thanks Carole & Ray 😀

[carastro]

Permanent set up will be a doddle then once you get used to doing them.  

I did my imaging last night and did my flats this afternoon.  If I am doing another target with the same set up I can use them again. 

Carole 

 

[MattJenko]

Binning should be used when you can afford to lose resolution, to better match pixel size to the available seeing/sky angle. I think it is primarily used to match sensors to focal lengths. An example would be an Atik 460 with 4.54 micron pixels on an RC6 scope of 1360 native focal length. This means that you would be imaging at 0.69 arc secs/pixel, which is pretty unachievable with UK skies. A more suitable resolution would be to bin 2x2 giving a 9.08 micron pixel and a 1.4 arsec/pixel resolution, much more manageable and sensible.

Different sensor types bin in different ways, so a 2x2 binned CCD will actually read out the 4 pixels as a single pixel value, minimising read noise, whereas a CMOS sensor would read out each pixel separately and combine afterwards, giving different characteristics.

One way to take advantage of binning without losing resolution in a final image is to take Luminance subs with no binning and then use 2x2 binning and shorter RGB subs for the colour, which I have done with a fair degree of success in the past.

All the comments on extra flats/darks etc are spot on too.

Cheers

Matt