Native Instruments’ Reaktor is arguably one of the most
powerful software tools ever to grace the drives of countless producers;
its ever-increasing array of sophisticated synths and effects is
conclusive proof. From Razor’s ultra-intuitive additive synthesis
workflow to Monark’s uncanny similarity to a real Moog, Reaktor is a
The latest Reaktor-based product is called Molekular, and
though it’s not a synthesizer per se, it packs enough MIDI-controlled
and truly musical features to warrant the attention of tech-savvy
keyboardists and producers everywhere. Molekular is a completely new
take on effects processing that has more in common with giant modular
synths than any other product out there, and that alone is cause for
putting it under the Keyboard microscope.
Molekular is so deep and complex that it’s worth taking a few steps
back to look at the big picture before diving into all of its features.
For starters, you have four simultaneous DSP modules for processing
sounds. Each module specializes in a different area of signal
processing. There’s an FFT-centric module, a time-based effect module, a
modulation effect module, and a distortion module—each with its own
collection of exotic processing modes. The arrangement of these modules
can be organized into a dizzying array of both serial and parallel
configurations, with extraordinary routing flexibility. And that’s just
the signal flow side of things.
On the modulation side, Molekular includes four LFOs, four
step sequencers, three triggered envelopes, an envelope follower, and
four “logic” processors that combine various modulation inputs
(including MIDI messages) to create even more sophisticated results. All
of these modulators can be routed to pretty much any effect parameter
and/or a complex X/Y vector matrix for morphing your patches.
If your head isn’t spinning yet, then throw in a bunch of
dedicated pattern sequencers and the ability to quantize the pitch and
timing of any relevant parameter. If Molekular was a synth, it would be
absurdly powerful. Instead, it’s a MIDI-controlled, pitch-savvy effects
processor that you can route anything into. Pianos, audio tracks,
other synths, vocals, drum loops . . . everything is fair game, and the
results are almost always mind-blowing.
Now that we have an aerial view of Molekular, let’s zoom
in at ground level, starting with its DSP modules. For starters, all
four DSP units include five common effects that can be selected,
regardless of the module: dual delay, single-band EQ, multimode resonant
filter, Metaverb (the same reverb as found in NI’s Maschine), and
“level.” The first four effects are pretty self-explanatory, but the
Level effect deserves a closer look. At first glance, it’s quite similar
to Ableton Live’s Utility device, which includes control over volume,
panning, stereo width, and phase. Dig a little deeper and you’ll find a
nifty integrated 16-step sequencer that offers discrete control over
volume and panning, with an additional layer that lets you sequence the
gate time of each step. It’s worth noting that Molekular sprinkles this
type of amenity throughout its effects, so while there are four global
assignable step-sequencers, there are also a number of dedicated ones,
depending on the effect.
The first block is dedicated to what NI calls spectral
effects. These include FFT tools, as well as more synth-like items like a
vocoder and a couple of resonators. The Vokoder effect includes three
tunable sawtooth oscillators that serve as the carrier, while the FFT
applies the original audio input as the modulator.
Another standout in DSP 1’s array of tools is the oddly
named Plagiarism effect. This option applies an envelope follower to the
audio input, which then triggers 16 envelopes that correlate with
tunable tone generators that include sine and pulse options. These
tunings are also variable. In sine-wave mode, they’re tuned to
harmonics. In pulse mode, the tunings align with musical intervals.
There are still five more effects in this module alone!
ReSonitarium and Dual Comb are generally similar in sound,
with Dual Comb providing two tunable comb filters, while ReSonitarium
includes four discrete resonators, each with tuning, delay time, and
panning, along with overall brightness and harmonic parameters. These
tools are great for juxtaposing dynamic ringing chordal artifacts on
rhythmic material like drum hits and percussion loops.
Spektral Hold and Spektral Shift play Fourier games with
the input that lean toward the glitchy, experimental side of things, but
Spektral Smear is another standout. Without getting super-technical
here, it splits the FFT analysis into two audio streams that have
individual envelope controls. The end result sounds like a cross between
a reverb, a flanger, and a physically modeled resonant cavity. Although
it’s tempting to say these sounds are tailor-made for electronic music,
they’re equally well suited to modern progressive rock as well as
cinematic sound design.
This module enters more familiar territory with its delay-based
tools—or “temporal structure modifiers,” if you prefer NI’s
terminology—including several multi-tap modes (up to 16 in the Angel
Delay), a granular Cloud delay, and a wonderfully straightforward Dub
delay, which does a fantastic job at both analog and tape-style
On the exotic side, there are Freezer, Iteratron,
Reversoid, and Trails effects that cover looping, stuttering and buffer
shuffling approaches in a manner that’s similar to iZotope’s StutterEdit
and Ableton’s Beat Repeat, but with a more experimental flair, thanks
to Molekular’s modulation amenities.
The final DSP 2 model, Ryuchi, was another favorite, with
its combination of tempo-synched time and pitch-shifting algorithms. The
results here were a lot more direct and intuitive than the other
buffer-shuffling tools in this module, with more obvious and musical
The third DSP module focuses on modulation effects like
chorus, flanging, and frequency shifting (which I like as an alternative
to phasers for many applications). The chorus is especially lovely, as
it includes a triple-tap delay with dual LFOs and phase offsets. I know
that sounds technical, but hey, that’s the basic topology of the classic
ARP Solina string synth, so if you’ve been looking for that sound, here
The Phlanger effect crossbreeds the tonalities of both
phasers and flangers. While the parameters are quite complex, it’s
almost impossible to get this module to sound bad.
The stereo frequency shifter is also implemented
beautifully, capable of ’60s-style phasing, Doppler weirdness, or ring
Naturally, there are bunch of cutting-edge processors in
here, too, notably Dark Forces with its Demon, Witch, and Zombie modes
(essentially four bizarre delay/pitch shifters with modulation), and
Half Wave, which applies discrete band-pass filters to each polarity of
the audio’s wave cycle.
Back in the real world, there are also excellent
pitch-shifters, dual filters, and a Filterbank mode that’s reminiscent
of the Moogerfooger MIDI MuRF pedal.
The final DSP module is devoted to distortion, saturation, bit-crushing,
and compression, with modes that are optimized for straight-up digital
mangling, as opposed to more warm, organic textures.
The two saturation effects—Modulo Fry and Wave Fold—bear a
strong resemblance to the results you’d get with Ableton’s Saturator
effect: filthy, shredded highs that sound like someone kicked in a
speaker and filled it with broken glass.
Slam Dunk is a compressor that adds new dimensions to the term “brick wall.” If you’re making hard music of any kind, be it rock or industrial, this could be your ticket to redline heaven.
Track OSC tracks both the pitch and dynamics of the input
signal, kind of like the Moogerfooger FreqBox, but a lot more
My favorite of the bunch is Track Pulses, which does some
crazy math on the input waveform. (I’ll spare you the details, but it
involves flip-flop division and single sample pulses.). It adds a
resonant filter and outputs a signal that sounds like a distorted
TB-303, no matter what you feed it.
With the certifiably insane DSP blocks covered, let’s dig into how
you can manipulate almost every parameter in real time using decidedly
more familiar tools.
Molekular’s four LFOs can be approached with a minimum of
head-scratching. All of the traditional waveforms are present, along
with both Hertz-based and tempo-synced rates. If you dig deeper, there
are some simple parameters for warping the waveforms in useful ways, as
well as bipolar and unipolar options for each LFO. Assigning LFOs is a
painless click-and-drag process, and you can modulate multiple
destinations simultaneously from a single LFO’s output.
The three envelopes feature a really cool implementation.
Remember, this is an audio processor, so the envelopes here are
optimized for effects use, not key triggering. Essentially, each
envelope is a simple attack-release affair, with hold time that’s
triggered via a TR-808-style step sequencer synced to tempo. This means
you can use each one for either triggering effects or LFO-type burbles
if you set the attack and release parameters to longer times.
Polyrhythms and unusual time signatures are a breeze, since each
sequencer can have from one to 16 steps. Want to put one envelope
pattern in 7/8 while the rest are in 5/4? Have at it! There’s an
envelope follower in here, too, if you want to control specific
parameters via the dynamics of your input signal.
The step sequencers are equally polyrhythmic, but with
individual values for each step, as expected. You can set it to run
forward, backward, or back and forth for added complexity.
Finally, there’s a “Logic” section (not to be confused
with Apple’s DAW) for combining the properties of two modulation
elements. While this sounds a bit complex, it’s easiest to think of it
as the section where you can assign the mod wheel to control the depth
of an LFO. Of course, you can apply math to these combined parameters as
well, so if you want the values of a step sequencer to serve as, say,
an exclusive-or gate for a sine-wave LFO, you can go there instead.
These modulation tools, combined with the vast processing
resources of Molekular would be more than ample for an entire career’s
worth of sound design, but there’s even more to consider. For one thing,
you can set up four completely different values and a base state for
almost every parameter in this system, then morph between the states
using an X/Y vector, which can be controlled by MIDI or by any of the
What’s more, you can pitch-quantize any relevant parameter
(like pitch-shifters, the vocoder, FFT processes, and resonators) to up
to eight different user-defined keys and scales, then switch between
them in real time or with one of the modulation tools (see Figure 1
Finally, you can quantize the motion of any
knob/parameter, whether it’s modulated or controlled in real time via
MIDI, to a tempo-synced note value. This means if you turn a knob, or
have a slow tempo-synced triangle LFO modulating a parameter, it will be
stepped in time with your DAW. Unreal.
The level of detail and control in Molekular is absolutely
mind-boggling. Just experimenting took more than a week before I got the
hang of it. If you’re an instant-gratification sound designer, this
software will drive you insane. But if you’re the kind of keyboardist
who thinks that Keith Emerson’s modular rig could use a few more cables
and another wall of synth modules, you’ll be in hog heaven.
In a lot of ways, Molekular reminds me of Symbolic Sound’s
Kyma hardware DSP system, but at a much more approachable price. If
you’re familiar with Kyma, you’ll be smiling right now. If not, then let
me explain it this way: Molekular is the single most flexible effects
processor I have ever used.
That said, it has a sound that’s deep, cinematic,
expansive, and decidedly digital. It doesn’t “ooze warmth” in any way,
but that doesn’t mean it’s not absolutely brilliant, because it is.
Whether you’re a die-hard progressive rock fan, a budding film composer,
or an EDM freak, you can’t afford to ignore Molekular. With a little
investment of time and patience, it could very well change the way you
approach recorded sound and music. I don’t say that very often, but in
this case, I can’t say it loud enough. All this processing power for
$149 also makes it a clear winner of our Key Buy award
PROS: Completely modular approach to effects processing. Tons of
exotic DSP effects. Patch morphing. Pitch and timing quantization on
relevant parameters. Totally unique sound.
CONS: FFT processes are CPU-intensive, even on a quad-core computer. Learning curve can be daunting for newcomers.
Bottom Line: The most sophisticated multi-effects plug-in we’ve seen yet.