Niccolò Belli wrote:
I did one more test with a energy saving lamp which is very close to the pc, so the level should be very high.
No, the level is still very low - peak 0.7 cd/m^2. You want somewhere like a minimum of 10, preferably more. I don't think the spectrum is accurate because it is being lost in the noise. You could try doing an emission measurement instead (-e, no ambient adaptor), as well as getting a lot closer to the lamp. Graeme Gill.