Which model type is more accurate in Spectrasys? I have replaced some Spectrasys Non Linear Amp models with S parameters data for the amp in a simulation. I was suprised that the S parameters give a worse (higher) 2nd harmonic response compared to the Non linear amp model. I think that non linear amp model is the correct one to use as it takes into account compression etc, which S parameter information doesn't, but it appears to give a more realistic response (ever the optimist!). What are your views?