[sdiy] Harmonic bandwidth

[Ben Lincoln]

I admit that I am not an expert in the area, but how can that be? I 
understand that if you take a signal and A/B it with A and B having 
different overall phases no human will be able to tell the difference. 
But when you instead change the phase of only some of the harmonics, the 
resulting waveforms are going to be different due to constructive and 
destructive interference.
In order to make sure I wasn't completely wrong, I tried this myself in 
Audacity - I started with two sine waves, 100Hz and 200Hz. In one copy 
of the project, I deleted a tiny piece of the 200Hz wave so that it was 
running at just less than 90 degrees ahead of the phase of the 100Hz 
wave. A/Bing the two it is easy to tell the difference. I tried it with 
some more complicated variations too, and the result was the same.

ASSI wrote:
> The established fact is that you are not able to distinguish two signals 
> that differ only in their harmonic phases by ear as long as those 
> phases are static or quasi-static.  Random phase offsets therefore do 
> nothing to the sound of anything.  Random phase modulation on the other 
> hand does and the result can be described as added noise.
>
>
> Achim.
>