[sdiy] "sample accurate light swing"

[Colin f]

> -----Original Message-----
> From: Synth-diy [mailto:synth-diy-bounces at dropmix.xs4all.nl] On Behalf Of
> Richie Burnett
> Sent: 24 October 2016 15:25
> To: Adam Inglis <21pointy at tpg.com.au>; synth-diy DIY <synth-
> diy at dropmix.xs4all.nl>
> Subject: Re: [sdiy] "sample accurate light swing"
> 
> My personal opinions:
> 
> > So presumably this
> > a) was a deliberate design decision, and
> 
> Yes, most likely.

I think it more likely it's just a consequence of the buffer size of the
"ACB" sound generation.
If it was done on purpose, they should have gone with 2ms.
 
> > b) contributes in a positive way to the feel/groove, ?
> 
> It make it authentic.  So it's positive if you want it to sound like the
original
> machine.

It could well be too small a variation to have any effect.
The human auditory system uses inter-aural delay as a source of spatial
information - i.e. if an impulse arrives at your left ear ~1ms before the
right ear, you perceive it as being sourced directly to your left, because
the path round your head takes roughly 1ms to travel at the speed of sound.
Try it out in an audio editor :)
Such timing differences, along with spectral markers superimposed by the
shape of your outer ears, are critical for our 3-dimensional location of
sound, and the reason why binaural audio is so much more immersive than
stereo.
The ~1ms threshold sets the limit on our ability to detect a timing
variation in rhythmic sounds.
The 2ms slop in the old X0Xs is just enough to be perceived.

Cheers,
Colin f