The basic principle for absolute calibration of an instrument to be used for 'in situ' measurements is to calibrate it in the same configuration as the 'in situ " configuration.

So, according to my experience, all the optical devices in front of your instrument have to be included in the calibration procedure.

I hope this short answer could help you.

Didier

Dear Amir

The first and foremost rule when dealing with instruments is that they should be calibrated otherwise we never know whether we are getting a reliable data from the instrument.

Second, to measure the light over a range of angles, cosine correctors are necessary. This goes as per the theoretical concept of light measurement and relates to the fact of uniformity of light distribution over the area of the detector used in the instrument.

I hope this helps.

Aamir,

Not sure if this topic is still of interest to you but, in principle, CC is not mandatory. However, you do need to make sure you use the exact same optical setup in both configurations (calibration & actual experiment) to properly cancel out your solid angle and optics transmission dependency. So if you don't want to use CC in the actual experiment, you shouldn't use it in the calibration either. The question then becomes: how do you do your calibration? If you use a tungsten lamp you could use a pyrometer to determine the temperature of your lamp blade and, upon knowing the emissivity, you can reconstuct the spectrum using planck's law. If however, your radiometric source was calibrated using the CC in the first place (your lamp's reference spectrum was given to you with the CC), then it's a bit more cumbersome...

To be brief: if you have a calibrated source in Irradiance [W/m^2/nm] then you need the CC. If you have a calibrated source in Radiance [W/m^2/sr/nm] then you don't need the CC (but you need to do the measurement in the exact same setup as the calibration).