[sdiy] Harmonic bandwidth

[Jim Palmer]

> Is it possible that mixing sines like this IS actually creating
> "mathematically" Identical sounds BUT the loudspeaker reacts differently
> when presented with a nice, tidy pseudo-saw (plenty of time to respond)
> and random-phase (it is expected to track the much more convoluted shape
> to give the same sound but, as the wave now effectively 'changes
> direction' more often the 'speaker cone fails to comply ?
> 

i think this is the most likely explanation so far, and i suspected this too
which is why i tried to keep the harmonics away from the freq limits.
there is also the question of crossovers.

> REF 2:-
> If I read this correctly, it disproves the above -unless-
> Any loudspeaker cone will respond differently when given a positive
> transient (voicecoil moves out of magnet field and loses sensitivity)
> compared to a negative transient (coil into field and increases
> sensitivity). This would amount to a non-linear transfer function in the
> audio path.

man i hope not, but this is also a possibility...
maybe the best way to test this is to compare on several speakers,
a/b between both speakers and clips to see if the differences are similar.
i don't have a convenient setup to do this.