NONLINEAR LABS C15 - 5. The Synthesis Engine

Synthesis Engine
Synthesis Engine
80 81
On the other hand, the engine is also capable of producing classical subtractive and FM
sounds, so a wide spectrum and variety of sounds can be produced.
The combination of the components can also be understood in a physical modeling
sense, as the branches can function as exciters for the Comb Filter as a resonator, while
the State Variable Filter and the eect chain work as an acoustic body, amplifier and
spatial setting at once.
5.2 Structure
Signal Flow
The overall signal flow of the synthesis engine is illustrated in the following diagram.
Two Oscillators and Shapers produce signals which are passed to the Comb Filter, to the
State Variable Filter, and to the Output Mixer. The filter output signals are also connected
to this mixer. The resulting mix is passed to the eect section and finally to the output. A
second mix from the filters and eects is defined for the Feedback bus.
There is a symmetric approach for the Oscillators, Shapers and Envelopes. Envelope
A, Oscillator A and Shaper A (also called “Branch A”) are quasi-identical to Envelope B,
Oscillator B and Shaper B (“Branch B”). Envelope C can be mapped to the filters, pitches,
and feedback amounts.
5. The Synthesis Engine
The C15’s first synthesis engine is named „Phase 22“. It can be seen as the latest stage of
an ongoing, long-time development, combining and advancing experiences gained from
precursor projects like Spark, Prism, Skanner and Kontour (available for Native Instru-
ments’ acclaimed Reaktor series).
There has been a steady evolution and optimization of the overall architecture and signal
flow, as well as the modulation mechanism and feedback structures. This process will
continue, further refining and advancing the synthesis engine.
5.1 Concept
The main design goal for the „Phase 22“ project is to exploit digital synthesis algorithms in
order to build a highly responsive and expressively playable instrument with a strong indi-
vidual character. At the same time, the set of parameters should be small enough to be
easily accessible. It is intended to harness complexity by a reduced, minimalistic engine
architecture, avoiding multi-mode switches for multiple waveforms or selectable filter
and eect types. Furthermore, (almost) all parameters of the synth engine are designed to
be continuous and time-variant in order to produce smooth transitions between sounds
and increasing the overall sensitivity of the instrument.
The core structure is a phase modulation synth that is based on two sine oscillators and
two sine shapers only (phase modulation is closely related to frequency modulation aka
„FM“). The combination of an oscillator and its envelope can be understood as a classical
FM operator. Together with a Shaper and a Feedback injection point it forms one of the
two „branches“. The signals from Branch A and B can be passed through the Comb Filter,
the State Variable Filter or directly to the Output Mixer.
The Comb Filter is a tool for complex spectral shaping and also works as a resonator, the
State Variable Filter can be applied for subtractive filtering. Signal routing is determined
by both the Output Mixer (defining the mix being passed to the eect section) and the
Feedback Mixer (defining the mix for flexibly applyable feedback loops).
Both filters and the eects have a profound impact on the feedback behavior, their
amplitude and phase responses determine the frequencies where the feedback can result
in self oscillation. In fact, the feedback bus is a unique and versatile feature of the synth
engine, raising the perceived acoustic complexity of a sound (if taken use of).