feat(desktop): upgrade Pocket TTS to April model

Cargo.toml

@@ -26,9 +26,10 @@ members = [
     "crates/buzz-pair-relay",
     "crates/buzz-relay-mesh",
     "crates/buzz-dev-mcp",
+    "crates/buzz-voice",
     "examples/countdown-bot",
 ]
-exclude = ["desktop/src-tauri"]
+exclude = ["desktop/src-tauri", "crates/sherpa-onnx-sys"]
 resolver = "2"
 
 [workspace.package]
@@ -170,3 +171,4 @@ strip = true
 # allowlist for the auth token). Revert to crates.io once #449 lands upstream.
 [patch.crates-io]
 aws-creds = { git = "https://github.com/tlongwell-block/rust-s3", rev = "c9fce3620dd434c1f810101d672cf384268dbb0f" }
+sherpa-onnx-sys = { path = "crates/sherpa-onnx-sys" }

Dockerfile

@@ -63,6 +63,10 @@ RUN apt-get update \
 # locations. The normal runtime strips it below; runtime-debug retains it.
 ENV CARGO_PROFILE_RELEASE_DEBUG=line-tables-only
 COPY --from=planner /build/recipe.json recipe.json
+# The sherpa sys patch keeps its published 1.13.4 version so Cargo can replace
+# the registry crate. It is excluded from the workspace recipe because
+# cargo-chef masks workspace package versions to 0.0.1.
+COPY --from=planner /build/crates/sherpa-onnx-sys crates/sherpa-onnx-sys
 # Cook the full workspace recipe — relay deps include workspace siblings, so
 # scoping to -p buzz-relay misses transitive deps and re-builds them later.
 RUN cargo chef cook --release --recipe-path recipe.json

Dockerfile.push-gateway

@@ -15,6 +15,9 @@ RUN apt-get update \
     && apt-get install -y --no-install-recommends build-essential pkg-config libssl-dev ca-certificates \
     && rm -rf /var/lib/apt/lists/*
 COPY --from=planner /build/recipe.json recipe.json
+# Keep the patched sherpa sys crate at its published version; cargo-chef masks
+# workspace package versions, so the patch remains outside the workspace recipe.
+COPY --from=planner /build/crates/sherpa-onnx-sys crates/sherpa-onnx-sys
 RUN cargo chef cook --release --recipe-path recipe.json
 COPY . .
 RUN cargo build --release --locked -p buzz-push-gateway --bin buzz-push-gateway \

Justfile

@@ -276,6 +276,7 @@ test-unit:
     #!/usr/bin/env bash
     if command -v cargo-nextest &>/dev/null; then
         cargo nextest run -p buzz-core -p buzz-auth --lib
+        cargo nextest run -p buzz-voice --lib
         # buzz-db migrator/lint tests: pure SQL-parsing unit tests (no infra).
         # They guard the embedded-migrator invariant (exactly the consolidated
         # 0001; cutover/backfill stays an operator script, not startup state)

crates/buzz-voice/Cargo.toml

@@ -0,0 +1,23 @@
+[package]
+name = "buzz-voice"
+version.workspace = true
+edition.workspace = true
+rust-version.workspace = true
+license.workspace = true
+repository.workspace = true
+description = "Reusable local voice primitives for Buzz"
+
+[features]
+default = ["static"]
+static = ["sherpa-onnx/static"]
+shared = ["sherpa-onnx/shared"]
+
+[dependencies]
+ort = { version = "=2.0.0-rc.12", default-features = false, features = ["api-24", "ndarray", "std"] }
+ort-sys = { version = "=2.0.0-rc.12", features = ["disable-linking"] }
+rand = { workspace = true }
+sentencepiece-model = "0.1"
+serde = { workspace = true }
+serde_json = { workspace = true }
+sherpa-onnx = { version = "1.12", default-features = false }
+tokenizers = { version = "0.22", default-features = false, features = ["fancy-regex"] }

crates/buzz-voice/examples/pocket_april_onset.rs

@@ -0,0 +1,75 @@
+//! Generate one raw, untrimmed April Pocket TTS onset-stress clip.
+//!
+//! Each sentence is a separate synthesis call, making every sentence start an
+//! explicit model-generation boundary. Outputs are concatenated without fades,
+//! trimming, normalization, or gain; only 250 ms of digital silence is inserted
+//! between calls so boundaries are easy to identify while listening.
+//!
+//! Usage:
+//!   cargo run --release -p buzz-voice --example pocket_april_onset -- \
+//!     <model-dir> <output.wav>
+
+use std::path::PathBuf;
+
+use buzz_voice::pocket::{
+    load_text_to_speech, load_voice_style, DEFAULT_VOICE, SAMPLE_RATE, VOICE_FILE_EXT,
+};
+
+const PASSAGE: &[&str] = &[
+    "Yep.",
+    "Hello there.",
+    "What happened?",
+    "Try this now.",
+    "Please check again.",
+    "Start with the smallest change.",
+    "I'm listening.",
+    "I've got it.",
+    "Sounds good to me.",
+];
+
+fn main() -> Result<(), String> {
+    let mut args = std::env::args().skip(1);
+    let model_dir = PathBuf::from(args.next().ok_or("missing model directory argument")?);
+    let output = PathBuf::from(args.next().ok_or("missing output WAV argument")?);
+    if args.next().is_some() {
+        return Err("unexpected extra argument".to_string());
+    }
+
+    let engine = load_text_to_speech(
+        model_dir
+            .to_str()
+            .ok_or_else(|| format!("model path is not UTF-8: {}", model_dir.display()))?,
+    )?;
+    let voice_path = model_dir.join(format!("{DEFAULT_VOICE}.{VOICE_FILE_EXT}"));
+    let voice = load_voice_style(&voice_path)?;
+    let boundary_silence = vec![0.0_f32; SAMPLE_RATE as usize / 4];
+    let mut combined = Vec::new();
+
+    for (index, text) in PASSAGE.iter().enumerate() {
+        eprintln!("boundary {}: {text}", index + 1);
+        let samples = engine.synth_chunk(text, "en", &voice, 1)?;
+        eprintln!(
+            "  raw samples: {} ({:.3} s)",
+            samples.len(),
+            samples.len() as f32 / SAMPLE_RATE as f32
+        );
+        combined.extend_from_slice(&samples);
+        if index + 1 != PASSAGE.len() {
+            combined.extend_from_slice(&boundary_silence);
+        }
+    }
+
+    let output_str = output
+        .to_str()
+        .ok_or_else(|| format!("output path is not UTF-8: {}", output.display()))?;
+    if !sherpa_onnx::write(output_str, &combined, SAMPLE_RATE as i32) {
+        return Err(format!("could not write {}", output.display()));
+    }
+    eprintln!(
+        "wrote {} raw samples ({:.3} s) to {}",
+        combined.len(),
+        combined.len() as f32 / SAMPLE_RATE as f32,
+        output.display()
+    );
+    Ok(())
+}

crates/buzz-voice/src/lib.rs

@@ -0,0 +1,3 @@
+//! Reusable local voice primitives for Buzz.
+
+pub mod pocket;