The Nonlinear Labs C15 is that rarest of things — a synth that dares to be different.
Stephan Schmitt was the founder of Native Instruments and its Chief Technical Officer for many years until he realised that he had lost his passion for what NI was doing and moved on to establish Nonlinear Labs. The company's first keyboard, the C15, is a digital polysynth delivered in two sections that you connect using the supplied brackets and cable. The Base Unit contains its five-octave keyboard, a pitch-bender, two long ribbons that you can use as controllers and as data entry devices, a small control panel with an OLED display, plus volume controls for the main and headphone outputs. The Panel Unit contains four panels of 24 parameter selectors (96 buttons to access a rather larger number of parameters) plus a central editing section comprising a further 18 buttons, a rotary encoder and a larger OLED screen.
In a novel twist, it's possible to play the C15 without the Panel Unit attached, using the Base Unit as a preset synth that you can also program using its browser-based editor/librarian (see 'The Editor/Librarian' box later). But if you attempt to edit the C15 before doing any homework, you'll probably end up with little more than a selection of weird noises that will do nothing to enhance your reputation as a sound designer. This is because its 'Phase 22' phase modulation synthesis engine (which is related to, but a significant advance upon, the Kontour engine that Nonlinear Labs developed for NI) eschews common digital synthesis techniques such as PCM playback and virtual analogue modelling and instead incorporates elements of 2-op FM and its sibling Phase Distortion synthesis, while not being the same as either of them.
There are two initial sound generators for each of the C15's 20 voices. These are called Branch A and Branch B, and each comprises an Oscillator and a Shaper. The Oscillator generates a sine wave whose pitch is controlled by a tuning parameter, an Envelope and key tracking. The initial phase is also programmable, ranging from -180º to +180º. However, there are three ways — all controlled by Envelopes — in which you can modulate this wave to obtain complex spectra. The first is by feeding the output from the Oscillator back to modulate itself or by applying the output from its Shaper as the modulator, or a mix of both. The second is cross-modulation derived from the output of the Oscillator or the Shaper of the other Branch, or a mix of both. The third is the modulation generated by a feedback loop tapped from further down the signal path. If this wasn't complex enough, each oscillator offers a parameter called Fluctuation that applies a random frequency offset to each cycle of the oscillation, creating anything ranging from an unusual smearing of the sound at low values to wide-band noise at higher ones. The amount of Fluctuation can also be controlled by an Envelope, which makes all manner of strange effects possible. A low-pass filter called Chirp lies in the summed modulation path and removes some of the more aggressive consequences of high modulation amounts.
Like the Oscillators, the Shapers generate sine waves that can be modified in various ways. Most obviously, you can overdrive a Shaper to create clipped waveforms with increasing numbers and amplitudes of harmonics, while Fold and Asymetry (that's not my spelling, I promise) further modify its output in interesting ways. There are then three output level controls within each Shaper panel, and these determine what the Branch sends to the next stage in the audio path. The first mixes the proportions of unshaped and shaped signals within the Branch, while the second (controlled by an Envelope) mixes the output from the first mixer with another tap of the feedback signal before passing the audio to the inputs of the Ring Modulators that lie in each of the Branches. The third control determines the mix between the input and output of the local Ring Modulator, and this is the signal that is passed down the audio path.
If there were nothing more to the C15, it would still be a powerful synthesizer, but Nonlinear Labs have added two filter sections to shape the audio yet further. The first is called Comb Filter, but offers three functions. Its first parameter determines the mix of Branch A and Branch B accepted at the input of the comb filter itself, while the next sets the initial pitch of the first notch in its spectrum. This is followed by a 2-pole all-pass filter that affects the phases of the signal components passing through it, with Tune and Resonance parameters determining the shape of the phase-shifting curve. The third is a high-cut filter that tames the high frequencies in the resulting signals. This section also contains an internal feedback path that can turn it into a resonator excited by the signals generated by the Branches. Its Decay parameter helps to determine the nature of the resulting signal and how long it takes to decay when the keyboard gate is removed, while the Gate parameter determines whether its oscillation decays naturally or is curtailed by a shorter release. All three filters can be modified by key tracking and Envelope C, while the Comb Filter can also be modulated by the outputs of Branches A and B.
The second filter section is a State Variable Filter that you can place in parallel with the Comb Filter section, or in series, or in a mix of both. It comprises two 12dB/octave filters whose configurations range from LP>LP, though BP>LP, HP>LP, HP>BP and finally HP>HP. The cutoff/centre frequencies of the two filters can be separated by a Spread parameter that provides all manner of additional profiles, while the Parallel parameter allows you to configure the two filters in series or in parallel, or anywhere between the two. You can determine the initial cut-off/centre frequency and resonance of the pair, and control these using key tracking and Envelope C. Filter FM is again possible using the outputs from the Branches.
The audio now reaches the Output Mixer, which accepts signals from Branches A and B plus the two filter sections, with individual pans for each. There's a master Level control, and Key Panning allows you to spread the output with low-pitched notes to the left and high ones to the right. If this seems straightforward, it isn't, because the Output Mixer contains another Shaper whose Drive, Fold and Asymetry parameters perform the same actions here as they do in the Branches.
Next come five effects sections in a predetermined series. The first is a stereo Flanger. The delay that creates the effect can be modulated by contributions from its internal oscillator and Decay contour, with parameters to determine the initial delay time, the modulation depth and (if desired) an offset in the delay times in the left and right channels. The signal then passes to another all-pass filter that creates phasing effects and the output from this can then be fed back to the flanger's input with either positive or negative polarity, and (again if desired) crossed so that the output from the right channel is fed back into the input of the left, and vice-versa. A high-cut filter then attenuates some of the higher frequencies in the effected signal, and a wet/dry mix parameter does what you might expect, but with the option of inverting the delayed signal for yet another flavour of effect.
The audio now passes to the Cabinet effect, which is yet another Shaper with an additional parameter that allows you to bias the shaping of the audio toward high or low frequencies. It also includes a band-pass filter with independent high-cut and low-cut controls, plus a saturation stage that adds yet more distortion.
It's been a long time since I encountered a synthesizer that made me think 'wow, that's different!', but there are many aspects of the C15 that have surprised me.
Next, the Gap Filter comprises independent low-pass/high-pass filter pairs for the left and right channels. You can control the centre frequencies for these, the offset between the channels, the gap between the high-pass and low-pass cutoff frequencies, and the resonances of all four filters. There's also a Balance control that emphasises the low or the high frequencies of the effected signal. The Mix control then determines the wet/dry mix, with positive values configuring the filters in parallel, and negative values configuring them in series. Nothing in the C15 is as simple as it seems!
The signal now reaches the Echo and Reverb sections. You can set the delay time for the Echo and determine an offset for the left and right channels for yet more stereo effects. There are controls for feedback, for cross-feedback from right to left and vice-versa, for a damping filter, and for the wet/dry mix. The maximum gain of the feedback loops is a tad under unity which means that, while the echoes can last for a long time, wacky sci-fi effects are beyond it. The output from this is then passed to the surprisingly good Reverb which offers control over pre-delay and reverb time, low-frequency damping and a Chorus effect that modulates the delay times of the reverberated signals. The audio then passes through a stereo master volume and tuning stage that includes a soft clipper to tame things if the signal gets a bit too wild.
At this point, you may want to breathe a huge sigh of relief at having reached the end of the C15's complexities, but you can't because you haven't. This is because, as already mentioned, you can feed the mixed outputs from the two filter sections and the effects (with a programmable contribution from the Reverb, or not, as you choose) back into the modulation and mix inputs of both Branches. Of course, this wouldn't be the C15 without yet another Shaper integrated into the Feedback Mixer, offering the now usual Drive, Fold and Asymetry controls. You can also set the base level of the signal in the feedback loop and use key tracking to bias this toward low or high frequencies. Inevitably, the choice of positions from which you derive the feedback signal and the relative amplitudes of the components can dramatically alter the sound that you hear.
Next, we come to a Unison section that can stack up to 12 voices per note with detuning of up to 120 semitones. You can also spread the start phases of the stacked voices to create phase cancellations and you can spread their positions in the stereo field for an expansive sound. Finally, we reach the Scale section, which acts on a per-patch basis and allows you to determine an offset of up to eight semitones for each note in the octave.
Throughout this description, I've referred to the C15's Envelopes. There are three of these, called A, B and C, with the first two acting upon parameters within Branches A and B, while the third controls the filter sections as well as the Feedback levels. They each offer five stages — Attack, Decay 1, [Breakpoint], Decay 2, Sustain, and Release — and allow you to control the Attack time and level as well as the Release time using velocity, while a Curve parameter allows you to shape the Attack from logarithmic to exponential responses. The durations of the Attack and Decay stages of all three Envelopes can range from near instantaneous to 16 seconds, while their Releases can range from near instantaneous to infinite, so sounds can be sustained indefinitely. You can control each contour's amplitude using velocity and key tracking, and shorten its duration as you play higher up the keyboard, which is vital for emulating many natural sounds. Envelopes A and B offer an overall gain control that Envelope C lacks, but C's Breakpoint and Sustain levels can be negative, so it can generate a much wider range of shapes.
The C15 also offers a small modulation matrix that allows you to direct your choice of eight hardware controllers — its four pedal inputs, the bender, the dual ribbons and channel aftertouch — to four programmable macros that can be directed to any combination of 86 possible destinations. I had some problems with the pedals during the review, but the manual admits that this section is "still in development and proper functionality is not guaranteed for now", so I'm not going to make a fuss.