Understanding Non-Diegetic Music in Story-driven Games

Non-diegetic music operates outside the game world’s reality. Characters never hear it; it lives solely for the player. This invisible layer of sound shapes emotional responses, signals narrative shifts, and defines pacing in ways that diegetic sound cannot. In story-driven games, where immersion and emotional engagement are paramount, non-diegetic music becomes a direct channel for storytelling. It can foreshadow a betrayal, underscore a bittersweet victory, or create existential dread without a single line of dialogue.

However, its power comes with responsibility. When overused or mismatched, non-diegetic music breaks immersion rather than builds it. A swelling orchestra during a quiet character moment can feel manipulative; a repetitive loop during exploration can numb the player. The difference between a soundtrack that elevates a narrative and one that undermines it lies in intentional design. Every cue must earn its place.

The following sections outline actionable best practices for implementing non-diegetic music, drawing from industry standards and real-world case studies. These guidelines apply to composers, audio designers, narrative designers, and indie developers alike.

Best Practices for Implementing Non-Diegetic Music

Align Music with Narrative Beats

The most impactful non-diegetic music is synchronized with story moments. This goes beyond starting a track when a cutscene begins. True alignment means composing music that mirrors the emotional arc of a scene, changes in real time as dialogue unfolds, and hits climactic points exactly when the narrative demands it. To achieve this, create a “music map” early in development—a document that lists every major story beat and assigns a musical intention to each. For example, a character’s quiet realization of a lie might start with a subtle dissonant chord, build tension as the truth surfaces, and resolve into a haunting motif. This map becomes a shared reference for composers and narrative designers, ensuring that music evolves with the script.

Adaptive systems make this synchronization possible across interactive scenarios. Tools like FMOD and Wwise allow designers to trigger specific stems or stingers based on dialogue cues or gameplay variables. The result is music that breathes with the narrative rather than sitting atop it. Games like Journey and The Last of Us Part II exemplify this approach, where each cue feels inevitable yet organic.

Maintain Subtlety and Embrace Silence

One of the most common mistakes is over-scoring: filling every second of gameplay with music. Constant non-diegetic music loses its emotional weight. Silence and sparse scoring are equally powerful tools. In a tense conversation, a low ambient drone or a single held note can sustain atmosphere without competing with dialogue. Similarly, after a climactic scene, letting the ambient soundscape take over allows players to process the moment without musical direction.

Mixing balance is equally critical. Music should never obscure dialogue, critical sound effects, or environmental audio. Use sidechain compression or automatic volume ducking to lower music during speech. Aim for a mix where the music supports the scene without dominating it. A good reference is Hellblade: Senua’s Sacrifice, where the non-diegetic score weaves in and out of diegetic sounds, reflecting Senua’s psychosis without overwhelming the experience.

Implement Dynamic and Adaptive Music Systems

Static, looping tracks quickly become repetitive and disconnected from player actions. Adaptive music systems—also called dynamic music—allow the soundtrack to change in real time based on game state, player choices, or narrative progression. This is one of the most powerful techniques for interactive storytelling.

  • Horizontal resequencing: Different sections of a track are triggered by game events. For example, a combat track might have low-intensity and high-intensity sections that swap based on enemy proximity or health.
  • Vertical layering: Multiple stems (percussion, strings, brass, pads) are mixed dynamically. As tension rises, more layers fade in, increasing density without jarring cuts.
  • Stinger system: Short musical phrases fire on specific actions—landing a critical hit, solving a puzzle, encountering a boss—adding punctuation to key moments.
  • Procedural generation: Some experimental titles use algorithms to generate music on the fly from parameters like player speed, danger level, or narrative context.

Audio middleware like FMOD and Wwise is the industry standard for building these systems. They let designers set transition rules, crossfade durations, and real-time mixing automations without deep programming. Investing in these tools is essential for any team serious about narrative-driven audio. For a deeper dive, the official FMOD documentation offers excellent tutorials on adaptive music authoring.

Control Pacing and Tempo

Non-diegetic music is a primary driver of pacing. A fast, driving rhythm makes a stealth sequence feel urgent; a slow ambient drone turns exploration into meditation. The music’s tempo and intensity should mirror the narrative arc: sparse and neutral during exposition, more rhythmically active and harmonically complex as tension builds, full intensity at the climax, and finally a gentle fade into resolution. This arc applies not only across the whole game but within individual scenes.

Transitions are equally important. A sudden shift from a quiet exploration track to an intense battle theme can feel jarring. Use crossfades, anticipatory cues, or brief silence to prepare the player. In adaptive systems, transitional stems or “bridges” can smooth the shift. Games like Bastion and Transistor use dynamic tempo shifts tied directly to player actions, creating a seamless flow between combat and narrative pauses.

Avoid Repetitive Loops with Variation and Depth

Repetition is a common trap. Even the most beautiful track loses its magic after the fiftieth loop. To maintain engagement, build variation into the system:

  • Multiple emotional variants: For each scene or area, compose several emotional states (calm, tense, resolved) that can be triggered by context.
  • Procedural variation: Use randomization tools within middleware to alter pitch, filter, or sample selection each loop.
  • Strategic silences: Let the ambient soundscape take over during low-intensity sections. The return of music becomes more impactful.
  • Narrative triggers: Advance the music to a new phase at checkpoints or story milestones instead of looping a single track.

Ori and the Blind Forest demonstrates this elegantly. Composer Gareth Coker uses dynamic layering and thematic variations that evolve as the player progresses through different biomes and narrative arcs, so the music never feels stale despite long play sessions.

Technical Considerations for Reliable Implementation

Artistic intent must be supported by robust technical execution. The following areas are critical for ensuring non-diegetic music systems function smoothly across platforms and player contexts.

Audio Middleware and Engine Integration

FMOD and Wwise are the two dominant tools. Both offer interactive music authoring, real-time parameter control, multi-platform export, and integration with Unity and Unreal. Choosing between them often depends on team size and budget. Wwise has a steeper learning curve but advanced mixing and spatial audio features. FMOD is more accessible for small teams and offers a generous free tier. Both provide profiling tools to test music behavior in-game. Early integration with the engine allows audio designers to iterate quickly alongside programmers, preventing last-minute implementation issues.

Layered Audio Design and Frequency Management

Layered audio is a technique where multiple stems of a track are mixed in real time based on game state. For example, a single “area theme” might have stems for ambient pads, melodic lines, percussion, and bass. At rest, only pads play. As the player approaches a point of interest, melody fades in. During combat, percussion and bass join. This creates a responsive feeling without requiring large amounts of unique music.

When implementing layers, avoid frequency masking. Each layer should occupy its own sonic space: pads in low-mid, melody in mid-high, percussion in high, bass in low. Use EQ and filtering to separate layers when all are active, ensuring clarity and impact. A muddy mix will fatigue players and obscure important audio cues.

Mixing and Balancing with Dialogue and SFX

A well-mixed soundtrack is invisible—it supports the experience without drawing attention. Practical guidelines:

  • Dialogue priority: Music should always sit behind dialogue. Use volume automation or sidechain compression to lower music during speech.
  • Sound effects clarity: Critical SFX—footsteps, weapons, environmental cues—must remain audible. Avoid excessive low-frequency energy in music that masks important effects.
  • Dynamic range: Keep overall dynamic range moderate so players don’t miss cues due to sudden volume shifts. Use compression and limiting to maintain a usable range.
  • Headroom: Mix music at -18 to -12 dB RMS relative to the game’s overall output, leaving room for dialogue and SFX.

Test mixing on various output devices—headphones, TV speakers, laptop speakers, and soundbars—to ensure no critical information is lost.

Platform-Specific Optimization

Different platforms have different audio constraints. Console, PC, mobile, and streaming each present unique memory, CPU, and latency profiles. Optimize early:

  • Memory: Use compressed formats like Ogg Vorbis or Opus for playback. Stream longer tracks from disk rather than loading entirely into RAM.
  • CPU: Real-time mixing and procedural generation consume cycles. Profile your audio system on target hardware to stay within budget.
  • Latency: Audio latency varies by platform. Design transition triggers to be tolerant of small timing variations so crossfades remain smooth.

Both FMOD and Wwise provide platform-specific profiling tools. For example, Wwise includes a performance monitor to track memory and CPU usage in real time.

Case Studies in Excellence

The Last of Us Part II (Naughty Dog)

Composers Gustavo Santaolalla and Mac Quayle created a deeply adaptive score that shifts seamlessly between guitar-driven melancholy, electronic tension, and silence. The system uses horizontal resequencing and stingers aligned to story beats—such as a sudden minor key shift during a betrayal. Mixing always supports dialogue and environmental sounds, and silence is used as a powerful narrative tool. The game demonstrates how non-diegetic music can evolve with character arcs and player choices without ever feeling intrusive.

Journey (Thatgamecompany)

Austin Wintory’s score for Journey is a landmark in adaptive music. Layered stems evolve based on player proximity to goals, presence of other players, and visual landmarks. There are no loops—instead, the music flows continuously, mirroring the emotional journey from isolation to connection. This game remains a benchmark for non-diegetic music as a fully integrated narrative element. For more on Wintory’s process, visit his website and watch his GDC talks.

Conclusion

Non-diegetic music is a narrative instrument, not mere background decoration. When implemented with intention, it deepens immersion, reinforces themes, and guides emotional response. The best practices outlined here—aligning music with story beats, using subtlety and silence, employing dynamic systems, controlling pacing, avoiding repetition, and mastering technical implementation—form a framework for creating soundtracks that feel alive and responsive. Success requires early collaboration between composers, audio designers, narrative designers, and programmers. It requires iterative testing on target hardware. And it requires the courage to leave space for silence, knowing that absence can speak as powerfully as presence.

For further exploration, the GDC Vault hosts numerous talks on game audio, and blogs like Designing Sound offer deep dives into techniques and case studies. The art of non-diegetic music continues to evolve, and every new story-driven game provides an opportunity to push it forward.