Skip to content

Sound Design Direction

Goal

The audio system should support generated ambient music and environmental soundscapes for a wilderness-focused terrain prototype. The target is not a general-purpose synthesizer or rhythm workstation. It should become a specialized landscape music generator for a world that feels ancient, lonely, mythical, and physically tied to terrain.

The tool should still preserve low-level experimentation paths. Raw synthesis is useful for learning, debugging, and discovering textures. Higher-level modes should add constraints that make good results easier and more coherent.

Creative Identity

The desired music should lean toward:

  • drones instead of songs
  • modal and pentatonic pitch material instead of chromatic movement
  • evolving layers instead of loops or beats
  • imperfect detuning instead of clean digital precision
  • sparse events instead of dense rhythm
  • natural noise textures such as wind, insects, water, and distant rumble
  • ancient or folk-adjacent intervals without becoming a melody generator first

Avoid early piano-roll, MIDI-grid, or step-sequencer metaphors. The better mental model is a soundscape that is constrained by a musical landscape.

Audio Studio Modes

The Audio workspace should remain split into modes:

  • Spectrogram: source-audio analysis, scribble painting, and resynthesis.
  • Synthesis: raw additive oscillator experimentation.
  • Soundscape: constrained modal/layer-based generation for the game style.

Synthesis should stay raw. It is the lab bench. Soundscape should be the opinionated music-design tool.

Soundscape Model

The long-term structure should be:

Soundscape
  -> root note
  -> scale or mode
  -> global register / tonal range
  -> drone layers
  -> resonance layers
  -> shimmer layers
  -> noise and weather layers
  -> rare motif events
  -> modulation and drift rules

The first implementation should focus on tonal constraints:

  • root note
  • scale or mode
  • scale degree
  • octave/register
  • detune in cents
  • waveform type
  • amplitude

Raw frequency should be secondary in this mode. The user should mostly choose musical relationships rather than exact Hz values.

Initial Scales

Start with a small set of strong choices:

  • Minor Pentatonic: lonely, ancient, and safe.
  • Major Pentatonic: open, natural, and pastoral.
  • Dorian: old-world, earthy, and not too sad.
  • Aeolian: darker natural minor.
  • Phrygian: mythical, ritual, and harsher.
  • Suspended Pentatonic: ancient, neutral, and spacious.

More scales can be added later, but too many options early will make the tool feel generic.

Layer Roles

Soundscape layers use roles to apply useful defaults and keep the generator opinionated:

  • Drone: long sustained root, fifth, or modal anchor.
  • Resonance: quiet harmonic or overtone support.
  • Shimmer: high, soft, detuned, and slowly moving.
  • Pulse: very slow swelling, not beat-driven.
  • Call: sparse higher modal movement.
  • Wind: filtered broad noise for air movement.
  • Rumble: low filtered noise for distant earth/weather weight.
  • Air: high filtered noise for hiss, insects, and thin atmosphere.

Roles now drive defaults, modulation behavior, register ranges, source type, and randomization rules.

Feature Roadmap

Recommended order:

  1. Scale/harmonic lock: root, scale, degree, octave, detune.
  2. Detune ranges: cents offsets and slight randomization for rich drones.
  3. Envelope: especially slow attack and release.
  4. LFO modulation: slow amplitude and pitch drift.
  5. Noise sources: filtered wind, rumble, and texture layers.
  6. Randomize within constraints: generate valid layers without chaos.
  7. Motif events: rare scale-locked notes, not rhythmic sequencing.

This order keeps the system musically coherent before adding motion and generative behavior.

Constraints

Important constraints:

  • Preserve raw Synthesis mode for freeform experiments.
  • Keep Soundscape opinionated and game-specific.
  • Prefer a few meaningful controls over a full workstation.
  • Keep generated results slow and ambient by default.
  • Use constraints to avoid bad or overly busy output.

Current Implementation

The implemented Soundscape mode provides:

  • a third Audio mode
  • root note selection
  • scale selection
  • seeded constrained randomization
  • modal tone layers
  • filtered noise layers
  • role preset layer add buttons
  • per-layer role, source, waveform, degree, octave, detune, amplitude, and enabled state
  • per-layer attack/release envelopes
  • per-layer amplitude drift, pitch drift, and drift cycle controls
  • per-layer modal motion: static, wander, call
  • motion controls: change interval, change chance, glide
  • live playback through the existing synthesis engine
  • waveform visualization based on computed layer frequencies
  • targeted tests for modal frequency, noise normalization, role presets, and constrained randomization

This gives the project a practical modal soundscape tool while keeping raw Synthesis available for experiments.