Does anyone know how to calculate quantum efficiency of a sensor? Let's assume a monochrome one for now.
I guess one needs to point it at a light source of a known brightness and then count the number of electrons collected by the CCD in a given time. But this need to be divided by the number of incident photons. But I guess the QE is also wavelength dependent.
So what light source can be used and how can its photon emission rate be calculated?
I don't think it would be that easy Mark. You would need either a calibrated light source that gave a standard output or another 'standard sensor' of known Q/E to compare it against or both.
I suspect your best bet would be a monochromatic LED, but how you'd go about callibrating it I've no idea...
You could always probably use a filter in front of a light source at least that way you'd know that only specific wavelengths would be getting through...
John
Maplins, RS and Farnell all publish specifications for LEDs, occasionally they have the manufacturers spec. also
e.g.
http://www.maplin.co.uk/module.aspx?ITAG=SPEC&ModuleNo=19498&doy=18m6#spec
Is that the sort of info that will help ?
Thanks for all your suggestions.
Another idea ocurred to me whilst cycling to work this morning. We all have a very well calibrated light source readily available - it's called the midday sun. O.K. maybe it's not readily available a lot of the time owing to the weather!
Just need to take a picture with a very stopped down lens at a very high shutter speed to prevent CCD saturation. And repeat the experiment with various (coloured) filters.
There must be loads of useful data available. Just need to adjust for atmospheric absorption depening on its azimuth. I'll let you know how I get on ...