Directional Audio Navigation Cues

1. Overview

Pattern Level: Sub-pattern of Audio-First Dual-Mode Navigation
Primary Phase: Guiding
One-line Summary: A navigation mechanism that plays periodic spatially positioned musical cues derived from each character’s theme, giving players directional guidance and character awareness in the absence of any visual interface.
Source: Voices of the Boyne (Battle of Boyne, Hill of Tara, National Botanic Garden)

Name

• Directional Audio Navigation Cues

Intent

To replace the directional feedback that a visual interface would normally provide — a map marker or waypoint arrow — with periodic audio cues that tell players where a character is and roughly how far away. The cues also carry musical identity, so players begin to recognise a character by sound before they arrive.

2. Target

Problem

In a screen-free location-based experience, players have no persistent directional awareness. There is no minimap, no waypoint arrow, no compass overlay. When a character is distant and outside the proximity zone, there is nothing to guide the player toward it. Without repeated directional feedback, the experience risks feeling aimless between encounters. Audio is the natural solution, but standard real-time spatial rendering can produce front-back ambiguity for short bursts, where the listener does not have enough time to resolve the ambiguity through natural head movement. A player who mistakes a rear cue for a frontal one will walk in exactly the wrong direction. Generic tones such as beeps would also clutter the soundscape and feel disconnected from the characters players are trying to find.

Context

• Players are in Wander mode with no character in proximity.
• No visual map or directional overlay is available.
• Players are wearing headphones capable of spatial audio rendering.
• Each character has its own musical theme.
• GPS position and IMU heading data are available in real time.

Use When

• A player is in Wander mode and one or more characters are within navigation range but outside the proximity zone.
• No target is locked and the system is cycling cues across eligible characters.
• A target is locked and the system is playing sequential cues for that character.
• A new layer has loaded and players need to orient to a fresh set of characters.

Forces

• Navigation function vs. audio identity: cues must communicate direction clearly, but generic tones feel out of place in a narrative experience. Musical fragments serve both purposes but require more production effort.
• Rendering accuracy vs. production cost: real-time binaural rendering is flexible but produces front-back ambiguity for short bursts. Pre-recording from fixed directions is more accurate but requires recording each cue multiple times. Tests with synthetically processed mono audio showed it did not match the localisation accuracy of pre-recorded binaural takes, particularly for directions behind the listener.
• Angular resolution vs. practicality: more recording directions improve precision but increase production effort. Too few and cues feel coarse.
• Cue frequency vs. audio fatigue: more frequent cues give better guidance but make the soundscape feel relentless.

Consequences

Weaknesses:

• Pre-recording from eight directions requires significant audio production effort per character.
• If the IMU sensor disconnects, the system falls back to phone sensors requiring the phone to be held upright. A phone left in a pocket will produce incorrect heading data.
• Binaural recordings are optimised for a generalised ear geometry and accuracy varies by listener.
• Wind and outdoor noise can mask spatial characteristics of the recording.

Strengths:

• Players can navigate accurately without looking at a screen.
• Front-back ambiguity is significantly reduced compared to real-time rendering for short bursts.
• Cues reinforce character identity and build anticipation before arrival.
• The sequential preview during lock gives players meaningful confirmation that their orientation is correct.

3. Application

Solution

1. For each character, navigation cues are pre-recorded from eight compass directions using a binaural microphone. Cues are short musical fragments derived from the character’s theme.
2. At runtime, the system calculates the bearing from the player’s GPS position to the character, then calculates the player’s heading from the IMU. The relative angle is quantised to the nearest of the eight recorded directions and the corresponding cue plays.
3. This repeats at a set interval. Bearing and heading are recalculated on each trigger. If the player has moved or turned, the cue updates automatically.
4. Standard wander cycling — when no target is locked, eligible characters take turns in rotation with a gap delay between each cue.
5. Sequential preview during lock — when a player locks onto a target, the system plays the character’s cues in sequence rather than repeating a single fragment. This gives a preview of the full theme and provides continuous confirmation that the lock is maintained.

Rationale

• Pre-recorded binaural captures authentic spatial characteristics that real-time rendering approximates. Short bursts do not give the listener enough time to resolve ambiguity through head movement. Pre-recording removes that dependency.
• Eight directions gives 45-degree resolution. At walking pace, a maximum error of 22.5 degrees self-corrects within a few steps. More directions add production cost without proportionate benefit.
• Musical fragments serve two purposes: direction and character identity. When the full theme begins on arrival, it feels like a continuation rather than something new.
• The sequential preview during lock rewards commitment to a direction and signals that the system has responded to the player’s intent.

Design Parameters

Example

A character named Sarsfield has a theme built around a slow fiddle phrase. His four cues are short extracts from that phrase, each recorded from one of eight directions. A player is 60 metres from Sarsfield, facing north, while Sarsfield is to the southeast. The system quantises to southeast and plays that recording. The player hears the fiddle from their right ear and slightly behind. They turn right. Six seconds later the cue fires again. The player is now facing southeast. The cue plays from directly ahead. They walk forward.

Implementation Notes

• Direction quantisation: divide the relative angle by 45, round to the nearest integer, and take modulo 8.
• IMU heading should be stabilised before direction calculation. A low-pass filter or higher-priority external sensor reduces noise.
• Do not interrupt a cue mid-play if the player rotates. Wait for completion then recalculate. Interrupting produces a jarring effect.

Evidence / Source

• Deployed at Battle of Boyne heritage site across multiple user testing sessions.
• 25-participant testing. Players reported cues were reliably localisable outdoors.
• Synthetic mono processing was tested as an alternative and found less effective for front-back differentiation, particularly for directions behind the listener.

References

[1] Wightman, F. L., & Kistler, D. J. (1999). Resolution of front-back ambiguity in spatial hearing by listener and source movement. Hearing Research, 113, 61-74.

Composition Recipes

Recipe	Scenario	Guiding	Indicating	Presenting + Notes
R1	Single character navigation	Directional Audio Navigation Cues (one active target)	Audio-First Dual-Mode triggers Interact on proximity	Progressive Proximity Audio Zones — cues guide approach, zones handle content on arrival
R2	Multi-character navigation with lock	Directional Audio Navigation Cues cycling across targets	Head-Directed Target Locking activates sequential preview for chosen character	Progressive Proximity Audio Zones — on arrival, cues stop and zones begin
R3	New layer orientation	Directional Audio Navigation Cues restart with new character set after portal	Audio-First Dual-Mode resets to Wander with new character set	Character-Triggered Layer Transition delivers the new character set that cues navigate toward