RiffBox review

["Marc Benigni" <Marc.Benigni@yale.edu>]

Hi all.  I'm new to the list but I've got a couple of old friends here.  I
recently purchased a RiffBox and wrote up a pretty long review for another
forum.  I thought it might be more useful/ interesting to the loopers
delight crowd, so I'm sort of cross posting.  I hope the length of this, my
first contribution, isn't a problem:

This is long, so I'll start with the very short version: well worth the
money if you can imagine having any use for a looper. If the "long version"
gets too far off on tangents etc. I'd be happy to answer specific 
questions.

I haven't had nearly as much time as I'd like to play with the RiffBox, but
I think its about time I posted that review. I'll continue to post ideas 
and
eventually sound clips here, until someone says, "Stop! You're WAY
off-topic!"

First of all, I still love the thing. I have found a couple of 
shortcomings,
but I believe most of them can be easily overcome or worked around.

I tested the unit for a couple of nights in the Vetta's effects loop, then
for about a week as a send from my the Delta 66 on my PC, running Guitar
Rig. I used an external footswitch so I could have the controls available 
on
a desktop (and because I wasn't really in a hurry to be stomping on my 
shiny
new toy. I tested with electric guitars with magnetic and piezo pickups,
clean and distorted, and I tested with an acoustic guitar.

Backline Engineering is a startup company, so packaging and documentation
are simple affairs, but everything is in clean and professional order. The
unit itself is also very old-school, with its 7-segment LEDs etc. Not the
prettiest thing, but it gives you that cool sense that you're playing
something out of the ordinary.

Sound quality is excellent, to the point that I don't have much to say 
about
it. Without giving much thought to levels, loops out sounded
indistinguishable from the signal I was feeding in. On a Sonar bus on the
PC, where levels and routing are very flexible, its a total no-brainer. The
only concern I had was with the Vetta effects loop: its a relatively quiet
signal, and I could not get it to clip the RiffBox's input. This made it a
little bit hard to diagnose problems at first, and (pure speculation here)
may impact the RiffBox's ability to track events.

Which leads us to "events"... definitely the most significant feature, and
probably what Gary's patent is all about. You can program the RiffBox to
count events - notes or chords that reach a given threshhold - as you play.
Looping begins automatically on the first event, and playback begins on the
(n+1)th event, where n is the count you specify. So, theoretically, you can
specify 8 events, play one bar of eighth notes, and RiffBox will repeat 
that
measure. What's cool is that it doesn't try to analyze tempo or anything,
it's just waiting on that next event, so you can play your measure square 
or
swing it very widely and the loop will still work fine.

The challenge is that guitars sustain quite a lot, so if you are playing at
all quickly, you need to be playing evenly, with an intentional staccato, 
in
order for events to count out consistently. Gary's addressed this matter
creatively by allowing you to record a wet guitar signal while triggering
based on a dry one - sort of a sidechain concept. This helps a lot, but 
even
clean guitars are sustainy. Acoustics are more percussive and so fare a bit
better, but all told, at typical tempos (say 80-120 bpm) 8th notes are a
more realistic proposition than 16th notes. (More on this later.)

The issues inherent in counting events aren't so gloomy though, since in
most cases you won't know the number of notes you intend to play in advance
anyway. 9 times out of 10 you'll use a manual mode - you hit the switch, 
and
the RiffBox loops beginning on the very next event. This is really just a
refinement of the typical looper, where you press the switch and looping
begins immediately. But its a BIG refinement. The result is glitchcore
without the glitches. A bit of practice and loops start coming out *very*
clean. And in this scenario you can play fast, legato runs without worry, 
so
long as the last note is distinct from the first, and you hit that pedal
somewhere in between. i.e. if there's a rest at the end of the measure(s),
you're golden.

Once you've got a loop in there, there are many (too many for me to have
tested them in full in a mere week or two), many modes to determine how it
will then behave. All are variations on "stop after n repetitions", "fade
over n repetitions", etc. which, when combined with layering options make
for some cool, musical effects without the tap-dancing typically required.

You can also wire this thing up to a drum machine or sequencer and when 
your
loop starts the drum machine will start, synced up at the correct tempo.
This is a cool feature, but I would love to see a mode that could follow my
tempo as I continued playing!

I've e-mailed a couple of suggestions to Gary, most of them minor, and most
of which can be corrected in firmware at his discretion - things like
changing LED colors to make status more clear etc. One major concern I have
is that I find it pretty easy to accidentally stop a loop and not be able 
to
restart it (because I'm armed to record again), or to corrupt a loop with a
bad layer (to be clear, *my* bad playing LOL) and not be able to revert. I
think Gary's thinking these things through now, and maybe he can post his
thoughts here.

My last and biggest recommendation concerns that whole "guitars are
sustainy" discussion above. Warning, this gets very geeky.  Since I had my
guitar routed through a PC, I realized I might be able to process the "dry"
signal RiffBox was triggering on to help simulate a staccato signal, while
my playing as recorded and looped could remain as expressive as I like. I
still believe this can be done with an expansion algorithm or something. 
But
I didn't happen to have a plug-in that was right for the job. So instead, I
ran my guitar's MIDI output to a softsynth, set that to a very staccato
transient tone, and ran *that* output to the RiffBox. This improved
performance considerably when I tried to achieve consistent event counts. 
It
occurred to me that the RiffBox could probably be programmed to respond to
any MIDI Note On events as if they were event threshholds, and for keyboard
or guitarists with GK pickups the unit would benefit greatly.

In conclusion: well, I guess I started with my conclusion. I think this box
is a very good value for the money, when I consider all the other loopers
I've tried that, for my purposes, proved completely unusable. Hopefully 
Gary
will manage to get some retail distibution. Check one out!