Re: RiffBox review

[Jeff Larson <Jeffrey.Larson@Sun.COM>]

Marc Benigni wrote:
> As for acting as a sync slave, not entirely sure what you're expectations
> are.  It doesn't support any time stretching or anything along those 
>lines.
> It will start and stop  based on external sequencers or audio triggers to
> let you loop within the confines of an established tempo, but I haven't
> really had any cause to test a setup like that yet.

The manual says that the loop time can be defined with MIDI start (FA)
and stop (FC) messages.  But nowhere does it say that it responds to
MIDI clocks.  It will SEND them to a drum machine but it does not
receive them.  This will allow you to tap the loop length with a MIDI
footswitch, but if you were controlling this with a sequencer,
the sequencer and the loop will gradually go out of sync since
they are not sharing the same clock.  How noticeable this is depends
on the amount of feedback however.

Setting the length with MIDI stop is odd too.  In order
to have the sequencer and the looper be in sync, you'd have to start
the sequencer (which starts the loop), then stop the sequencer
(which ends the loop and starts it playing).  But now the sequencer
is stopped, so you have to start it again and there is no way
to do that exactly in sync with the start of the loop.  To work around
this you'd have to have a MIDI track in the sequencer that sent start/stop
messages to the looper at the right times rather than using
the sequencer's own transport controls, workable but hard to control.

To support slave sync properly, at minimum you have to be able to derive
the tempo from incoming MIDI clocks, then when pressing
the footswitch to start/end the loop, quantize the actual
start/end to a "beat" as defined by the MIDI clocks.  Beyond
that, it would be nice to have a "beat multiplier" setting so you
could quantize the loop boundary to a multiple of a beat,
such as a 4 beat bar.  With that, then you just leave the sequencer
running and set up beat quantized loops at any time.

Jeff