Juergen Lilien wrote:
But I can't understand the following point. In 2-2. they claim to have used an Eye One pro (spectrometer) as the measuring device, but the sensitivity (RGB) graphs look more like the response curves of a colorimeter (CIE 1931 standard observer color matching function). I thought these reponse curves are only typical for colorimeters, so what am I getting wrong here?
The aim is always to arrive at tri-stimulus values, it's just that colorimeters and spectrometers use two different mechanisms. A spectrometer measures the spectrum and then computes the filter values. A colorimeter uses physical filters. They have different inaccuracies. Making physical filters that match the standard observer curves accurately is hard (ie. expensive), and low cost instruments don't do this very well, hence the need for a CRT/LCD switch, and the problems with wide gamut monitors. Computing the filters allows for the greater flexibility of using different observer models. Graeme Gill.