Snr

[pvelez]

I am working on an asteroid at the moment, trying to get a decent light curve. Its faint - about mag 15 - so I have increased my exposures and started guiding the shots.

Last night I had exposures 4 minutes long. My RC8 gives 0.68 arcseconds/pixel so I always bin - usually 2x2.

Now I understand that the higher the SNR the better for photometry and that SNR of 50 is ideal while sub-20 is troublesome.

My target varied over time - as you'd expect as it approached the meridian as the airmass decreased. The best I could manage was SNR of 21 and at lower altitudes it was down to around 12.

Here's the question - I had similar SNR at 3x3 and 4x4 binning. Should I expect better SNR at greater binning as I collect more light from my source? Or will the SNR only change as I increase the integration times? My experience is that a 4 minute shot at 3x3 is better than a 3 minute shot at 4x4.

I was calibrating with darks and flats. It was also a lousy night early in the evening with some cloud about.

Any thoughts?

Pete

[Merlin66]

You need more exposure. The shot noise varies as the square root of the total ADU; the bigger the signal, the lower the noise.

[pvelez]

So longer integration times rather than higher binning is the way to go

I had suspected as much. For some reason I had thought that binning could be substituted for longer subs.

Thanks for the tip - looks like I need to g out to 5 minutes to beat the noise down some more.

Pete

[LightningNZ]

Higher binning will work if your stars normally subtend several pixels, however the Kodak 8300 chip seems to not do well at binning according to discussion over at CN. The suspicion is that the horizontal register bins are only a little "deeper" than the regular pixels, meaning that extra charge from binning is lost.

I've never been able to confirm or deny this but it makes sense to me.

Hope this helps,
Cam

[pvelez]

Thanks Cam - it does

It was plain to me the other night that binning made no meaningful difference to SNR

Pete