Authoring Sound
Audio authoring is the process of setting up and tailoring audio content inside the chosen tool. Terms like authoring, implementation, integration, or technical audio design are sometimes used interchangeably, but they all refer to crafting audio behavior using middleware, game engines, or other systems. Authoring is usually done by technical sound designers, though sound designers often take on this role too.
In middleware, assets are imported, organized into a hierarchy (tree structure), and customized. Parent, child, and sibling relationships allow higher-level nodes to pass down properties such as volume or pitch, making adjustments more efficient. Middleware also allows auditioning sounds directly, or connecting live to the game to test in runtime. Once configured, assets and settings are exported as soundbanks, which the game engine then uses as audio data.
Without middleware, authoring must be done directly in the game engine. This is usually more code-heavy, as engine tools may lack certain built-in features. Typical workflows include:
- Importing and organizing audio clips
- Adjusting format and compression settings
- Triggering events via code, prefabs, or audio components
- Mixing with the engine’s built-in tools
Mixing is closely tied to authoring, but we will examine it separately. As seen here, authoring is a deep and technical process that blends artistry with system design.
With the basics in place, let’s look at key technical audio design concepts that shape how audio behaves in-game.
Here are some of the most essential sound properties in any audio implementation system:
- Volume, pitch, filtering
- Playback (looping, transitions, fades, delay, voice limiting)
- Attenuation (sound over distance)
- Positioning (location in space)
- Panning (distribution across speakers)
- Spatialization (immersive 3D propagation)
Randomizing properties like volume or pitch helps reduce repetition and make frequently used sounds (e.g. footsteps, impacts) feel natural. Controlled randomization also increases variety without requiring more assets. For example, we can set a sound’s volume to be randomized +/- 2 decibels each time it’s played.
Parameters (or as commonly referred to as RTPCs coming from Wwise) are game parameters mapped to sound properties in real time. For example, mapping player health (0–100+) to a “player_health” parameter could drive a high-pass filter that muffles gameplay sounds as health decreases. RTPCs allow highly responsive audio.
Containers (or audio groups) are structures that hold audio assets, set their properties, and define how they play back. You can think of them as building blocks for interactive sound behavior.
- Common container types include:
- Random: plays a random variation from its set
- Sequence: plays a series of sounds in order
- Blend: layers multiple sounds together
- Interval: delays and plays sounds at random or fixed intervals
- Switch: selects sounds based on game data (e.g. character state or environment)
Some middleware, such as Wwise, provides a range of individual container types, while others offer modular containers that can be configured to behave in different ways. Both approaches allow designers to create flexible and reusable audio structures.
Containers can also be nested in tree hierarchies, where parent containers control or pass down properties to children. This makes it possible to design complex systems like adaptive footsteps. For example, the following hierarchy plays the correct variation depending on the player’s armor type and the surface they walk on:
PlayerFootseps (switch container - light/heavy armor)
---LightArmor (blend container - mixes base and material sounds)
------BaseSoundLight (random container - picks a random light base variation)
---------Player_footstep_base_light_01 (audio clip)
---------…other variations
------Material (switch container - surface material type)
---------Dirt (random container - picks a random dirt variation)
------------player_footstep_dirt_01 (audio clip)
------------…other variations
---------Concrete(random container - picks a random concrete variation)
------------player_footstep_concrete_01 (audio clip)
------------…other variations
---HeavyArmor (blend container)
------BaseSoundHeavy (random container)
---------player_footstep_base_heavy_01 (audio clip)
---------…other variations
------Material (duplicated from the light armor hierarchy)
---------…
States represent game conditions like “combat,” “stealth,” or “menu.” They can switch sound sets or apply different mixes at runtime, making them essential for dynamic mixing.
Unity: https://unity.com/
FMOD: https://fmod.com/
Wwise: https://www.audiokinetic.com/en/wwise/overview/
Home
