"Correcting the non-linearities"? That kind of language makes me nervous. Equalization is correcting linear features like frequency and phase response, via a linear operation (multiplication in frequency domain, or convolution in time domain; or what is the actual implementation in practical audio equalizers? in hardware you can do an RLC network with a certain transfer function, which we're hoping has few nonlinearities) The nonlinearities are exhibited in features like harmonic distortion and resonances, which are not being directly compensated for. Or am I missing something here?