Any thoughts on Impulse Response and Group Delay testing?

[Mike Sulzer]

Inductors alone do not introduce any phase distortion: the rest of the circuit is just as important. If you connect a pickup to a very high impedance the rest of the circuit consists of the pickup capacitance and resistance. You have a low Q resonant circuit with an f0 of maybe 12 KHz for a PAF type humbucker. If you connect to a high impedance through a cable, you add the cable capacitance in parallel with the pickup C, and get a (usually) much lower resonance. So a study of the phase delay due to the pickup inductance is a study of the effect of the resonance.

The effect of Litz wire, if any, would be at the higher frequencies.

The brain collectively analyzes responses in time of a set of bandpass filters something like 1/3 octave wide, I believe. Only phase differences that change the (detected, that is amplitude versus time) outputs of these filters can be heard. Thus audible phase differences must significantly affect the transient response of the signal enough so that the filter outputs change; this eliminates from consideration certain changes in phase. For example, rearranging the relative phases of the harmonics in a signal, with the changes restricted to say 360 degrees has little effect. On the other hand, using an fft to produce very severe changes that alters the shape of the envelope of the waveform over a time much longer than individual cycles of oscillation has a huge effect. In effect, this type of change delays in time certain ranges of frequencies with respect to other ranges, producing significant differences in the response of the filters in the ear.

[Rick Turner]

And right there is the whole issue of "group delay". This is what I want to know more about. It was certainly an issue...the most important one we looked into when developing the Mama Bear DSP based acoustic guitar modeler. Acoustic guitars are a messy mass of group delay!

[Mike Sulzer]

And right there is the whole issue of "group delay". This is what I want to know more about. It was certainly an issue...the most important one we looked into when developing the Mama Bear DSP based acoustic guitar modeler. Acoustic guitars are a messy mass of group delay!

Group delay, yes; delaying different ranges of frequencies differently. But the delays have to be pretty large before they matter, given how the system in the head works.