>> + 			leftVol: leftVol * scaledVol // ScaleFactor
>> + 			rightVol: rightVol * scaledVol // ScaleFactor.
> 
> Since this method is sent with Float leftVol and rightVol arguments,
> perhaps it's worth keeping precision by not truncating to integers here
> too.

On a related note, I saw when profiling sound-synthesis code that quite 
some time is spent truncating values and checking for possible overflow 
in sound buffers.

I implemented the following in SoundBuffer:

at: index add: aNumber

	| s |

	s := (self at: index) + aNumber truncated.
	s >  32767 ifTrue: [s :=  32767].  "clipping!"
	s < -32767 ifTrue: [s := -32767].  "clipping!"

	self at: index put: s

and also

at: index mult: aNumber

	| s |

	s := ((self at: index) * aNumber) truncated.
	s >  32767 ifTrue: [s :=  32767].  "clipping!"
	s < -32767 ifTrue: [s := -32767].  "clipping!"

	self at: index put: s

It would be nice if those where available as primitives. I guess they 
would be much faster, and those operations are always required somewhere 
in the chain when working with sound buffers, so I think such primitives 
would be quite useful in general.

Stef