[sdiy] Harmonic bandwidth

[Ben Lincoln]

Scott Gravenhorst wrote:
> I don't use Audacity, so I don't know of it's flaws, but I think that the mystery of
> it's code is somewhat problematic here.  The conventional wisdom of not being able to
> detect a difference when only phase is changed (but not changing) is borne of much
> experimental evidence.  
>
> IMO, if digital methods are to be used, something simple that doesn't have any
> capability to adjust itself so that there is no internal mystery.  Perhaps a simple C
> program that generates a .WAV file with the sinewaves for both waveforms?  This way,
> others can look at the code as well as listen to the output.
>   
If I had more free time, I would take you up on that suggestion. 
Audacity *is* open source, but it's also fairly complicated.

I did dig out my old copy of Cool Edit and got a similar effect there. 
The waves I made are not as complicated as the ones in Audacity (no 
multitracking made it a lot more tedious), so the difference is smaller 
but still present (the resonance is where I notice it most):

http://www.eventualdecline.com/AB2.mp3

If it is some sort of digital artifact or common programming error, 
maybe at least it's one that can be put to use intentionally.