CMOS for Photometry

[Wayne Hawley]

I am starting to take images of asteroids and have obtained my observatory code. I use an ATIK Horizon Series 1 CMOS camera on a LX200 Classic 10 inch telescope. 
Software is AstroArt 7 for imaging and Astrometrica is used to process the images. 
The ATIK Horizon has a facility to change the gain setting which I believe is supposed to enable a higher SNR if the gain is increased. 
Does anyone have experience of the use of this camera or the ZWO ASI1600 which uses the same chip and also has variable gain. 

Wayne

[Dr Andrew Smith]

Have a look at this http://www.astrosurf.com/buil/atik_vs_zwo/

Regards Andrew 

[Richard Miles]

Thanks Andrew.

I see that Buil comments: “For normal camera use it is mandatory to select the « Even illumination » option (accessible also from the ASCOM interface). If you set this option on, the gradient disappears – in this situation the Atik Horizon camera works properly ” Wayne’s images have a vertical gradient to the background, especially when flatfields are generated, which have short exposure times. So turning off this strange progressive shutter is absolutely essential if photometry is the purpose!

[Wayne Hawley]

Thanks Andrew I’ve read the test report and will recheck all my settings. Currently I have both Even Illumination and the Pad data options switched off. 

[James Screech]

I use the ZWO ASI1600MM for variable star photometry, you need to keep the gain low as increasing it will reduce the signal to noise ratio. I mainly use a gain of 0 and at most 80, you basically need to capture as many photons as possible without saturating the sensor. Increasing the gain does not increase the number of photons captured, it reduces them, that in turn means that the SNR is less.

[Dr Andrew Smith]

I am not sure I agree with this. Yes increasing the gain reduces the well depth but thus can be compensated for by take more shorter exposures and or defocusing.

Increasing the gain reduces both the read noise and the quantisation  noise.  

There is a balance to be had trading off these factors. I use a gain of 200 (ASI gain) for my spectroscopy.

Regards Andrew 

[James Screech]

I’ve always found that increased gain leads to a larger scatter in the photometry readings and larger errors reported by the photometric software. The error (sigma) increases in greater proportion as the signal (photons) decreases (sigma being the square root of the signal) and this is not counteracted by the reduction in read noise. At least that’s what I suspect, I’m no mathematician though.

[Dr Andrew Smith]

You should convert ADU to e using the (inverse) gain in e/ADU for a real measure of S/N. This is probably what your software is not taking into account. The gain does not change the number of photons captured (assuming it’s not saturated ) .

You can quite easily measure the gain in the right units see here for example http://www.astrosurf.com/buil/isis/noise/result.htm

Regards Andrew 

[James Screech]

But for a given magnitude, to avoid saturation you need a short exposure, therefore fewer photons and a so a lower SNR.

[Dr Andrew Smith]

Yes but given the low read noise you can stack multiple short exposures to get the same S/N ratio.  This is the normal strategy with CMOS cameras.

Regards Andrew 

[Dr Paul Leyland]

You could always just suck it and see. It shouldn’t take more than one night to take dozens of exposures of a relatively bright star at a variety of settings. Then process them and see what works best for your equipment.

[James Screech]

Yes, stacking would counter act that, but isn’t something that I generally do with photometry. Though having said that I have done it on some DSLR photometry to improve the SNR in the past.

[Author]

Posts