Module wasm 

WASM-compatible synthesis API

Provides an API that works without filesystem access, suitable for WebAssembly (wasm32) targets. Model and config data are passed as byte slices instead of file paths.

This module is available on all platforms but is designed primarily for WASM. On native platforms, prefer PiperVoice::load() for convenience.

Structs

WasmModelCapabilities

Model capabilities detected from ONNX input/output node names.

WasmSynthesisResult

WASM-friendly synthesis result (no file I/O)

WasmVoice

WASM-compatible voice synthesizer. Loads model from bytes rather than file paths.

Functions

parse_config

Parse config JSON string into VoiceConfig

samples_i16_to_f32

Convert i16 PCM samples to f32 normalized audio (-1.0 to 1.0)

samples_to_wav_bytes

Convert f32 audio to WAV bytes (in-memory, no filesystem)