← Back to ninetysix Originals

Official effect

Split Atom

Currently a single high-passed fuzz branch with bias skew rather than a dual-band octave fuzz.

Despite the name, the current Split Atom patch is not a dual-band fuzz with octave partial generation. It is a single high-passed fuzz branch that raises gain with `fuzz`, adds a DC offset with `bias` before the saturation stage, and then shapes the result with a broad 980 Hz peak and a tone-controlled high shelf before `mix` blends it with the dry stereo preview source. The result then passes through the shared DC-blocking, output-gain, and soft-limiting stage.

  • ninetysix Originals
  • Library #276
  • Preview 8s

Waveform preview

Static peaks come from the compiled FAUST preview render. The live scope is drawn from the browser playback output in real time.

Timeline
0:00 / 0:08
Source
Compiled browser preview
Live oscilloscope from the browser playback output.AnalyserNode keeps the signal path unchanged while exposing the waveform.

How It Works

Technical view

Despite the name, the current Split Atom patch is not a dual-band fuzz with octave partial generation. It is a single high-passed fuzz branch that raises gain with `fuzz`, adds a DC offset with `bias` before the saturation stage, and then shapes the result with a broad 980 Hz peak and a tone-controlled high shelf before `mix` blends it with the dry stereo preview source. The result then passes through the shared DC-blocking, output-gain, and soft-limiting stage.

  • ninetysix Originals
  • waveshaping / saturation
  • delay processing
  • oscillator / LFO modulation
  • DC blocking

Signal Path

Compiled preview chain

Preview source: An internally generated picked-string phrase is widened to stereo with the shared 19-sample right-channel delay before it hits the current single-branch fuzz core.

  1. 1internal picked-string demo phrase
  2. 2shared stereo spread with a short right-channel delay
  3. 3high-pass cleanup
  4. 4gain boost into the fuzz stage
  5. 5DC bias offset before saturation
  6. 6single saturation stage
  7. 7980 Hz peak shaping
  8. 8post-fuzz high-shelf trim
  9. 9dry / wet mix
  10. 10DC blocker
  11. 11output gain trim
  12. 12final tanh limiter

Preview source

An internally generated picked-string phrase is widened to stereo with the shared 19-sample right-channel delay before it hits the current single-branch fuzz core.

Effect core

Single high-passed DC-biased fuzz branch with post-fuzz peak and shelf shaping.

Signal path

The static waveform and the live player both reflect this compiled signal chain.

  • internal picked-string demo phrase
  • shared stereo spread with a short right-channel delay
  • high-pass cleanup
  • gain boost into the fuzz stage
  • DC bias offset before saturation
  • single saturation stage
  • 980 Hz peak shaping
  • post-fuzz high-shelf trim
  • dry / wet mix
  • DC blocker
  • output gain trim
  • final tanh limiter

Controls that matter

These are the currently active controls detected in the FAUST source for this effect.

  • mix: Crossfades or scales the current wet layer against the dry stereo preview source.
  • output_gain_db: Applies final post-effect level trim after DC blocking and before the output limiter.
  • fuzz: Raises the gain feeding the fuzz stage.
  • tone: Sets the post-fuzz 980 Hz peak and high-shelf trim.
  • bias: Adds a DC offset before the saturation stage.

Code-backed analysis

The generated description is tied to the active helper blocks and routing found in the current DSP.

  • demo
  • fx
  • outputStage
  • leftDemo
  • rightDemo
  • fuzzCore
  • phraseMono
  • previewShiftSamples

Controls

Audible controls in this DSP

  • mix

    Active

    Crossfades or scales the current wet layer against the dry stereo preview source.

    Higher settings make the processed layer more dominant relative to the dry phrase.

  • output_gain_db

    Active

    Applies final post-effect level trim after DC blocking and before the output limiter.

    Matches loudness without changing the underlying effect structure.

  • fuzz

    Active

    Raises the gain feeding the fuzz stage.

    Higher settings add more edge, sustain, and fuzz density.

  • tone

    Active

    Sets the post-fuzz 980 Hz peak and high-shelf trim.

    Higher settings brighten the fuzz and push more upper-mid focus.

  • bias

    Active

    Adds a DC offset before the saturation stage.

    Higher settings skew the clipping and change the fuzz texture.

Analysis

Code-derived notes

Current implementation is one DC-biased fuzz branch with EQ, not a true split-band or octave-partial design.

Active blocks

  • demo
  • fx
  • outputStage
  • leftDemo
  • rightDemo
  • fuzzCore
  • phraseMono
  • previewShiftSamples
  • softsat
  • splitAtomOctave
  • pickedVoice
  • fund
  • pickEnv
  • ampEnv
  • triggerAt
  • maxDelaySamples

Inactive helpers

  • synthVoice
  • organVoice
  • bellVoice
  • quantize
  • combLP
  • echoOnly
  • diffuser

Warnings

No current warnings are attached to this effect.

Source

  • library/ninetysix-originals/split-atom.dsp
  • Output stage: DC blocker -> output gain trim -> final tanh limiter
  • Library position: 276

More In ninetysix Originals

Continue through the library

Each effect page uses the same generated description pipeline and compiled waveform preview assets.

Crosswind

Currently a straightforward stereo echo pair rather than a cross-feedback delay into bloom reverb.

View How It Works

Neon Melt

Currently the same straightforward stereo echo pair as Crosswind rather than a more melted or saturated delay texture.

View How It Works

Ghost Rail

Currently the same straightforward stereo echo pair as Crosswind rather than a more haunted or transformed echo design.

View How It Works