Combining unbinned data with binned data, same channel

[ManuelJ]

Hi,

I'm thinking about capturing the same data binned and unbinned, to get nice SNR in the faint nebulosity without erasing the detail in the good SNR data. The only thing I can think of it's a SNR mask, but that doesn't work well, because the stars will get bigger because of the unbinned image.

A star mask will do the trick, but creating a perfect star mask is something outside my knowledge.

Any ideas?.

Regards,
Manuel.

[Carlos Milovic]

You may try a wavelets decomposition. In a nutshell, replace the first two layers of the binned image (upsampled to the unbinned size) with the first two layers of the other image.

How to do that:
- Delete the first two layers on the binned image (previously upsampled and registered with the unbinned image).
- Clone the unbinned image, and delete the first two layers on the cloned image.
- Apply this PixelMath expression to the clone: "unbinned - clone + 0.5". Disable rescaling, and replace the image.
This will create a high pass image, where you have only the information of the first two layers. Now we have to add this data to the binned image.
- PixelMath: "binned + clone". Enable rescaling.

You may try also using a mask derived from the luminance, to target the high SNR zone more, and thus preventing too much noise to contaminate your background. I would use a blured luminance as mask, with the histogram adjusted.

[ManuelJ]

Good point, Carlos. I'll try that.

Thanks,
Manuel.