Erik -

Your idea sent me to the right parameters, but, in fact, the opposite - for
whatever reason - turned out to be the answer. At a tempo of 100bpm
the ksmps seems to need to be equal to 100 or greater for there to be what
sounds to me like rock solid beat resolution. Thanks for the relationship
information and the quick reply.

d1st

Erik Spjut wrote:
> 
> I don't have time to check out your problem but I am 99.99% certain it will
> disappear if you set kr=sr. The %@$$*&^$%#^ control frames always end
> places other than the exact time you specify UNLESS your duration is an
> EXACT multiple of the control frame period. Hence linseg's, expseg's, and
> any other useful signal ends at the wrong place. My guess is that your
> tempo's give you the exact relationship when things work.