Game Development Game Audio Syllabus

What is the difference between background music and a dynamic (adaptive) music system?

Background music is a fixed track/loop that plays regardless of state; a dynamic music system changes the music in real time in response to gameplay (intensity, location, events) via layers or branches.

Contrast the two main adaptive music techniques: horizontal re-sequencing and vertical layering.

Horizontal re-sequencing rearranges/branches between musical segments over time (transitions to new sections). Vertical layering (re-orchestration) adds or removes simultaneous stems/instrument layers to change intensity without changing the underlying progression.

What is a musical 'stem' in vertical layering?

An individual instrument or group track (e.g., drums, strings, bass) mixed to play simultaneously with other stems; the system fades stems in/out to raise or lower musical intensity.

What is a musical transition marker (sync point) and why is it used?

A defined point in the music (beat, bar, or grid boundary) where the system is allowed to switch segments; quantizing changes to markers keeps transitions musically smooth instead of jarring mid-phrase.

What is a stinger in game music?

A short, one-shot musical phrase layered over existing music to punctuate an event (e.g., an enemy spotting the player), without replacing the underlying track.

What is interactive vs. adaptive music (as commonly distinguished)?

Interactive music responds to direct player actions/input; adaptive music responds to the changing game state/environment. Both are forms of dynamic music, often used together.

In music theory, how many semitones are in an octave and what is the frequency ratio of an octave?

$12$ equal semitones per octave in equal temperament, with an octave corresponding to a frequency ratio of $2:1$.

Give the equal-temperament formula for the frequency of a note $n$ semitones from a reference pitch $f_0$.

$f = f_{0}\cdot 2^{\,n/12}$; each semitone multiplies frequency by $2^{1/12}\approx 1.0595$.

What reference frequency is standard concert pitch A4, and why does it matter for game audio?

$A_{4} = 440\,\text{Hz}$; it anchors tuning so recorded/synth music and interactive layers stay in tune with one another.

What is a diatonic major scale's interval pattern in semitones?

Whole–Whole–Half–Whole–Whole–Whole–Half, i.e. $2,2,1,2,2,2,1$ semitones over the octave.

How are major and minor triads built and how do they differ emotionally in game scoring?

Major triad = root + major third (4 semitones) + perfect fifth (7): $\{0,4,7\}$, typically bright/heroic. Minor triad = root + minor third (3) + fifth (7): $\{0,3,7\}$, typically darker/tense. Composers pick mode to match emotional tone.

What are BPM and tempo, and why must a game composer expose them to the audio engine?

BPM (beats per minute) is the tempo, the number of beats per minute. Exposing it lets the engine quantize transitions, stingers, and gameplay events to musical beats/bars for tight sync.

Convert a tempo of $120$ BPM into the duration of one beat and one 4/4 bar.

One beat $= \frac{60}{120} = 0.5\,\text{s}$; one $4/4$ bar $= 4\times 0.5 = 2\,\text{s}$.

What is a leitmotif and how is it used in games?

A short recurring musical theme associated with a character, place, or idea; games recall and vary it (tempo, orchestration, mode) to reinforce narrative and player recognition.

What is audio middleware, and why use it instead of coding sound directly in the engine?

A specialized authoring tool + runtime (e.g., FMOD, Wwise) that lets sound designers build interactive audio behavior (mixing, adaptive music, DSP, randomization) without programmer intervention, decoupling audio design from code and iterating faster.

Name two industry-standard audio middleware tools and their vendors.

FMOD (by Firelight Technologies) and Wwise (Audiokinetic); both integrate with Unity and Unreal and expose designer-driven event systems.

In Wwise, what is an 'Event' and what is an 'RTPC'?

An Event is a named trigger the game calls to start/stop/modify audio actions. An RTPC (Real-Time Parameter Control) is a game parameter (speed, health, etc.) mapped to audio properties like volume or pitch for continuous real-time control.

In FMOD Studio, what is an 'Event' and what is a 'Parameter'?

An Event is the basic container of sound content and behavior the game plays; a Parameter is a continuous or labeled/discrete value the game sets to drive changes within the event (e.g., intensity, surface type).

What is a middleware 'snapshot' (mixer state)?

A saved set of mixer/bus settings (volumes, effects, ducking) that can be activated to override the mix for a situation (e.g., 'underwater' or 'menu paused'), often blended in/out over time.

What is audio ducking (sidechaining) and give a common game use.

Automatically lowering one bus's level when another is active, e.g., dipping music and ambience when dialogue plays so speech stays intelligible; often driven by a sidechain compressor keyed on the priority signal.

What is voice limiting / voice stealing (virtualization) and why is it needed?

Capping the number of simultaneously playing voices for CPU/memory limits; when the cap is exceeded the engine 'steals' (stops or virtualizes) the least important voice, chosen by priority, distance, or age.

What is an audio bus (mixer group / submix)?

A routing channel that groups multiple sounds so they can be processed and level-controlled together (e.g., an SFX bus, Music bus, Dialogue bus feeding a Master bus).

What is randomization/variation in a sound engine and why use it?

Selecting among multiple samples and randomizing pitch/volume per playback to avoid the 'machine-gun' repetition of identical triggers (e.g., footsteps), increasing perceived realism.

What is a Blend/Random Container in middleware?

A container object holding multiple child sounds: a Random Container picks one (or a shuffled sequence) each trigger, while a Blend Container crossfades between children based on a parameter/RTPC.

What is streaming vs. in-memory (decompressed) audio, and the trade-off?

Streaming reads audio from disk on demand (low RAM, some latency/disk I/O), used for long music/ambience; in-memory keeps decoded/short assets resident for instant, low-latency playback (higher RAM), used for reactive SFX.

Why are short, latency-critical SFX often stored uncompressed (e.g., PCM) while music is compressed (e.g., Vorbis/ADPCM)?

Uncompressed PCM avoids per-play decode cost so gameplay SFX fire with minimal latency; long music tolerates decode overhead and benefits from compression's large storage savings.

State the Nyquist sampling theorem and the standard game/CD sample rate it justifies.

To capture a signal without aliasing, the sample rate must exceed twice the highest frequency: $f_{s} > 2 f_{max}$. Human hearing tops near $20\,\text{kHz}$, so $44.1\,\text{kHz}$ (and $48\,\text{kHz}$) is standard.

What bit depth is standard for game audio production and what dynamic range does it provide?

$16$-bit (delivery) or $24$-bit (production); dynamic range $\approx 6.02\times\text{bits}$, so about $96\,\text{dB}$ for 16-bit and $\approx 144\,\text{dB}$ for 24-bit.

What is an audio 'listener' in an in-engine audio tool, and where is it usually attached?

The virtual microphone/receiver whose position and orientation define how 3D sounds are panned and attenuated; typically attached to the camera or player character.

In Unity's built-in audio, what components handle emitting and receiving sound?

An AudioSource component emits/plays an AudioClip (with 2D/3D, spatial blend, roll-off settings), and a single AudioListener component (usually on the camera) receives and mixes all sources.

What is Loudness Units Full Scale (LUFS) and why do console/platform audio guidelines specify a target (e.g., around $-23$ to $-24$ LUFS)?

LUFS is a perceptual, integrated loudness measurement; platforms mandate a target integrated loudness so games have consistent volume relative to other content and to each other, preventing one title from being far louder than another.