Atik 314L+ and channel sensitivity

[swag72]

I understand that our CCD camera's are always going to be more sensitive in one channel than the other. So how do I know which channel my Atik is most sensitive in? And how can I use this information?

[pete_l]

You could either check the datasheet. The sensor is a Sony ICX285AL (same as used in Starlight's H9 / H9C). That will have response curves for the chip, itself.

Alternately, a more practical technique is point your scope at a white star and see what colour corrections you have to make to get a flat, white image. It might be worth defocussing a bit, to get a larger star-disc to sample.

[swag72]

I have the data for a G2V calibration which shows the following ratios

R: 1.11

G: 1

B: 1.04

So does this mean it's more sensitive in the red channel?

[johnrt]

Least sensitive in red I think? Then blue and most sensitive in green?

[Uranium235]

I have the data for a G2V calibration which shows the following ratios

R: 1.11

G: 1

B: 1.04

So does this mean it's more sensitive in the red channel?

Is that a G2V calibration from your setup? To get a true reading you need to do your own testing. If they are your figures then they only reflect the sensitivity of your entire optical train, not just the camera. Its different for everybody, well... apart from those who would have an exactly identical setup

[swag72]

Yes it's a G2V calibration from my setup. This shows that the red is less sensitive? I ask as when I am taking narrowband and I am going to map into HST, then my red channel is the least sensitive and that is the SII data - which is the least abundant!! Is that right?

[steppenwolf]

Based on your G2V figures, I agree that your red is the least sensitive for your total imaging system.

You cannot equate the narrowband figures with the G2V figures - SII for example, which you would use in the Hubble palette as the red channel, just happens to be less abundant than the Ha that you would use for the green channel - this is not to do with the sensor's inherent sensitivity to colour, it is to do with the emission values from the objects that you choose to image.

[riklaunim]

Narrowband doesn't match RGB sensitivity. H-alpha and [s II] are quite close in red/infrared, while [O III] is in green. And narrowband tricolor imaging isn't supposed to give natural/calibrated colors.

QE looks like so:

Max sensitivity is in green with extension into red.

[Uranium235]

The trouble is, Ha and SII are both red - with the SII being the deeper of the two. I think G2V calibrations are more suited to RGB imaging where the difference in the values may be more noticeably different. Maybe the difference in the response between Ha and SII in regard to the 285 chip sensitivity would be of a level that isnt worth worrying about?

[johnrt]

Is this difference in sensitivity that all camera chips have really all that important? So long as I have a decent enough number of subs (I always shoot the same length for each channel) programs such as Pixinsight are flawless in colour calibration and balancing out the histogram.

So should we worry about shooting longer red and blue subs?

[dph1nm]

If you want to do it properly, you should probably match signal-to-noise, not just signal. And that is complicated by the fact that the relative sky brightness is different in the different filters, so a simple G2V calibration won't do.

NigelM