Plenty of synthesizers still use the voice design laid out
by the Minimoog 40 years ago. You know the drill: a few oscillators, a
filter or two, a couple of envelope generators, a couple of LFOs. But
today’s computers are fast enough that instrument designers are free to
seek out new ideas and new sonic horizons, to boldly go . . . you know
the rest. Kaivo is one of that new breed. It has two resonators, but no
filter. It only has one oscillator, but some of the “waveforms”
crossfade among four samples at once. It has a two-dimensional LFO and
the most complex noise source you’ve ever seen. It has no modulation
matrix, just a click-and-drag patch bay. The sounds it produces tend to
be even stranger than that description would suggest. If you’re craving
stock leads, pads, and basses, look elsewhere. Kaivo’s factory patch
list sports names like “Shrapnel Flute,” “Wandering Drum String,” and
Kaivo meshes granular synthesis with physical modeling and
modular sound design. Like Madrona Labs’ first synth, the modeled
analog Aalto, Kaivo is laid out with half a dozen modulation sources in
the top section of the panel, and sound modules in the lower half. The
sound modules are configured in a fixed left-to-right signal path.
The central area is the patch bay. Patch connections are
made by dragging the mouse from one of the outlets of a module in the
upper panel to one of the inlets in the lower panel. The upper modules
also have inlets, so they can modulate one another. Small dials on the
inlets are used to set the amount of modulation.
Multiple sources can be patched to one inlet, but they’ll
all share the same amount setting. The Granulator and Resonator each
have two pitch inputs, however, one linear and one logarithmic, and
several of the modulation sources have output level knobs, so there’s a
reasonable amount of control over modulation depth.
At the left end of the lower panel is the Granulator. This
module produces tones using granular synthesis on one of more than 65
factory waveforms. (An update is planned that will allow user waves to
be imported.) Some of the included selections are actually four separate
waveforms; the granulator can be modulated in both the X and Y
directions, Y modulation producing crossfades between one wave in the
set and another. This may sound confusing, but the animation of the
panel display makes it clear what’s going on. You can program (or
modulate) the amount of grain overlap, the rate of grain production, and
several other things.
Next in line is a lowpass gate, which can function either
as a VCA or as a lowpass filter. Note, however, that this filter is in
line before the resonators, which are where a lot of the creative
tone shaping comes in. As a result, Kaivo’s sound will seldom have a
snappy cutoff when you release your finger from the key: The resonators
will usually continue to ring for a few moments, if not longer.
The first resonator, cleverly named “Resonator,” has a
choice of seven models, including strings, chimes, and springs. You can
program and modulate the pitch, brightness, excitation position,
panning, sustain amount, and nonlinearity. If the Granulator’s pitch is
tracking the keyboard while the Resonator’s pitch isn’t, you’ll tend to
hear quite strong resonant peaks depending on which note you play.
The panning control applies only to the wet output of the
Resonator. Other than that, Kaivo has no user control over the stereo
The second resonator, called “Body,” is monophonic: It
processes the output of all of the voices together. Here you can choose
one of two wooden boxes, a metal plate, or a frame drum. Because these
are two-dimensional surfaces, you can set and modulate both the X and Y
positions where the incoming signal will excite the model. Tone, pitch,
sustain, and nonlinearity can also be modulated.
The output module, at the right end of the lower panel, is
fairly simple. It has a chorus effect (implemented as an on/off
switch), a Tilt dial, which tilts the signal toward the lower or higher
part of the frequency spectrum, and a limiter (also an on/off switch).
Since the resonators sometimes get a little too excited, having a
limiter on-board is highly desirable.
I find that a good way to develop new sounds in Kaivo is
to leave both resonators off, by turning their wet outputs down to zero.
Start by finding a Granulator sound that you like, and then add
resonators to taste. The results can be quite organic, or possibly
suited to a horror movie soundtrack.
Starting at the upper left corner, we have Kaivo’s MIDI
input module. This has outputs for gate (note on/off), pitch (note
number plus pitch-bend), velocity, aftertouch, mod wheel, and two
assignable MIDI control change messages. Pitch-bend depth, glide, the
number of notes of polyphony, and micro-tuning scale can also be set.
The most interesting output from this module is called
“vox.” This transmits the number of the voice that is currently being
activated. If you set Kaivo to three-note polyphony and play a
three-note chord, for instance, the vox output will transmit a value of
1.0 for the first voice, 2.0 for the second voice, and 3.0 for the third
voice. You can use this signal for whatever you like—for instance,
giving each voice its own step sequencer speed. Each voice has its own
sequencer, though all of them share the same panel settings.
The sequencer has 16 steps and graphic editing. It has two
outputs each for gate on/off and the step value, and one of each type
can be delayed by up to eight steps. It might seem odd at first to have
two step-value outputs for a synth with only one oscillator, but by
delaying the step being sent to the Resonator you can produce two
independent pitch lines within a single voice. The output can be
quantized to equal-tempered pitches, and its overall range can be set in
half-steps. Both the number of steps and the offset (starting step) can
be modulated in real time, as can the playback rate.
The LFO has X and Y outputs, which share the frequency and
output level controls. Waveforms include circle, “knights,” and “rain.”
We’re running short of space here, so I’m going to let you download the
demo version and find out for yourself what all that means.
The noise source has a frequency control that goes from
0.1Hz up to 10kHz. It can have up to eight narrow frequency peaks, and
both the rate and level can be modulated. It can produce anything from
subtle tone changes to insane mangling.
The two envelopes are a bit more normal. One is an ADSR
type with a velocity on/off switch and modulation inputs for attack,
decay, and release time. The other is a delay-attack-release type that
can loop, making it an LFO.
I wouldn’t mind seeing a few more modules added to Kaivo’s
arsenal. But while its feature set seems modest at first glance,
there’s a lot of creative sound power lurking in its guts. The
CPU usage meters in my various DAWs showed intense activity when I added
a Kaivo track to a piece, but a glance at the Task Manager in Windows 7
revealed that each instance of Kaivo was being run in a separate
processor, so a quad-processor machine should be able to run up to four
Kaivo instances without needing to do any track freezing. In sum, it’s
powerful, it’s strange, and having a fast computer is a good idea. What
else do you need to know?
PROS: Exotic sound design. Unusual modules. Highly patchable.
CONS: No filters. Only one oscillator. CPU-intensive.
Bottom Line: Fresh sounds for edgy music.
$129 direct | madronalabs.com