Camera Spectral Response

← Previous topic: Camera Resolution | Next topic: Lenses →

For many years, we've had only an approximate idea of the spectral sensitivities of the cameras we've been using. This data has come from the camera manual, when available, and has been limited to the sensor and IR filter.

For the polarization system (multi-camera polarization detector), we've made a considerable number of assumptions regarding the spectral content of the signals. We now have some interesting data.

Using an Ocean Optics S2000 spectrometer and a standard desk lamp (incandescent), I've been able to measure the spectral responses of several of the important components of the system, along with the source (sun). Below is the result of the investigation.

The spectrum for the lens is an approximation, and is highly position dependent. (Since it focuses light on the S2000 sensor, it can have a transmittance greater than 100!) The total spectrum is simply the convolution of the individual parts. It is interesting to note that the IR cut filter cuts more UV than the protective UV filter does.

Also interesting, at the CIL it has been determined that the increase in transmittance of a polarizing filter in the IR is accompanied by a decrease in the efficiency of the polarization -- in other words, in the IR a polarizing filter isn't.

Now, what does the solar spectrum look like? The next plot shows the solar spectrum, in arbitrary units relative to 100%, the total transmittance from the previous plot, and the resulting convolution.