The first audio example with zero SPC700 assembly. The lib's audio module ships its own resident SPC700 driver (built from source at lib build time) and exposes the whole engine as plain C: this example streams four BRR samples into APU RAM at runtime, plays them polyphonically with per-play volume/pan/pitch, shapes every voice's envelope, and switches a concert-hall echo on and off — all through snes/audio.h calls.

Input	Action
A	cello note, center
B	"AU" vowel, left
X	"SS" hiss, right
Y	"PP" pop, center
L / R	cello a fifth down / up (pitch demo)
Up / Down	master volume
START	toggle the echo hall

The backdrop tints with each action. Up to 8 simultaneous voices, round-robin allocated by the engine.

ROM mode: LoROM (project default).

SNES Concepts

Dynamic BRR loading: audioLoadSample() streams sample data over the APU I/O ports — no fixed APU memory image, no .spc700.asm
audioPlaySampleEx(): volume/pan/pitch per play; the DSP's 8 hardware voices ARE the polyphony
ADSR as note articulation: one envelope per voice set at init — looping samples ring out and die instead of sustaining forever
Echo: audioSetEcho() sizes and clears the ring, audioSetEchoFilter() sets the FIR, audioEnableEcho() routes voices in

Sample provenance

res/cello.brr is krom (Peter Lemon)'s PitchMod cello (see ATTRIBUTION.md for the caveat). The ss/pp phonemes are OpenSNES-original formant-synthesized BRRs from the speech_synth example. au.brr is an RMS-boosted regeneration of speech_synth's AU (gen_au_boost.py — the speech bank is peak-normalized, and a vowel's high crest factor made the original ~3x quieter than the recorded cello next to it).

How to Build

make

Modules Used

console, audio, input