Hope this is not daft

[alan potts]

Question, I have a 183mc which has pretty small pixels 2.4um I think. Now I would like to use this on my 800mm scope but at 1x1 bin rate they are really too small. If I use it as 2x2 binned will this work? I want to try M1 and at least get something of reasonable size on the capture.

Alan

[Merlin66]

Alan,

You can’t bin CMOS sensors. The best would be to expose and then bin in software.

 

[ollypenrice]

Over sampling with modern CMOS cameras is not the issue it was with CCD. You won't resolve down to your numerical sampling rate but you should still get a good result. Just software bin the captures.

Olly

Crossed with Merlin. 😁

Edited November 21, 2020 by ollypenrice

[wimvb]

You CAN bin MONO cmos sensors which will give you a slightly increased snr, but not to the same extent as ccd. That's because the binning is done in the camera software or in the driver and not during sensor read out. Colour sensors have the colour filter array that prevents 2x2 binning. When you deBayer during stacking, you can use super pixel mode, which takes a group of 2x2 RGGB pixels to make one RGB pixel without any interpolation. That won't give you the same advantage as camera or driver binning, but it's definitely something to experiment with, though.

[alan potts]

Thanks for all the answers, I was rather led by the APT software showing 1x1 2x2- 4x4 binning and just thought it was all cameras. I will give the 183mc a try on the APO, sure I will get something.

Alan

[alan potts]

2 hours ago, ollypenrice said:

Over sampling with modern CMOS cameras is not the issue it was with CCD. You won't resolve down to your numerical sampling rate but you should still get a good result. Just software bin the captures.

Olly

Crossed with Merlin. 😁

Olly can you suggest a software for this, I only have PS.

Alan

[ollypenrice]

1 hour ago, alan potts said:

Olly can you suggest a software for this, I only have PS.

Alan

Does your stacking software offer this option? Some do. Failing that, just resizing in Ps to the size you feel the data will support will have a similar effect though Vlaiv advises binning for the best SN ratio. The big mosaic captured by Yves recently, and which I post processed, had been resampled downwards at the stacking stage in APP. It was incredibly clean. Because I can hardware bin I haven't done this myself.

Olly

Edited November 21, 2020 by ollypenrice

[alan potts]

3 minutes ago, ollypenrice said:

Does your stacking software offer this option? Some do. Failing that, just resizing in Ps to the size you feel the data will support will have a similar effect though Vlaiv advises binning for the best SN ratio. The big mosaic captured by Yves recently, and which I post processed, had been resampled downwards at the stacking stage in APP. It was incredibly clean. Because I can hardware bin I haven't done this myself.

Olly

I will need to read up on that one, never tried it before. I always thought Resize was about making the image and file size smaller. I am going to try the 183mc on the 800mm, I have used it with reducer at 635mm and it gave nice results so it will be usable I would have thought. I was also wondering if I can use the 183mc as a guide camera through the 330mmBorg for a better guide maybe.

Alan

[ollypenrice]

15 minutes ago, alan potts said:

I will need to read up on that one, never tried it before. I always thought Resize was about making the image and file size smaller.

Alan

It is but, to do that, information from more than one pixel must be compressed onto one. It's routine to find that, if you don't have enough data to present an image at full size (meaning one camera pixel getting one screen pixel), you may still find it looks good at 66% or so. In fact you really do need a deep dataset to show at full size and sometimes you might make a decision not to try but to accept a smaller size. I think folks will often do this when making a mosaic, for instance, because the final image is going to end up huge in any case. The net effect of resampling downwards is similar to binning.

Olly

[MartinB]

Just to clarify CMOS vs CCD binning.  Combining the output of 4 pixels i.e 2x2 binning you achieve an increase in SNR in the same way as if you had stacked those 4 pixels.  CCD cameras can read 4 combined pixels which reduces the read noise by a factor of 4.  CMOS cameras can't do this since each pixel has to be read before combination so there is no reduction in the read noise of the binned pixels.  This is less of an issue for CMOS cameras given that they already have low read noise compared with CCDs.  A low read noise allows for shorter exposure times but also increases the dynamic range (linear full well / read noise = dynamic range). 

The decision regarding when to bin isn't straight forward and seeing considerations need to be considered.  It is easy to software bin an over sampled image, much harder to unbin an undersampled one!  Best to experiment  to see how the camera and scope set up performs in different seeing conditions.