manawenuz/wz-phone
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix(audio): update frame_samples on codec profile switch, fix buf sizing
Some checks failed
Mirror to GitHub / mirror (push) Failing after 27s
Details
Build Release Binaries / build-amd64 (push) Has been cancelled
Details

Browse Source 
frame_samples was immutable — when adaptive quality switched from 20ms
(Opus24k, 960 samples) to 40ms (Opus6k, 1920 samples), the send loop
kept reading 960 samples and feeding half-sized frames to the encoder.
This caused Opus6k to produce ~11 frames/s instead of 25, making audio
choppy.

Fix:
- frame_samples is now mut and updated on profile switch
- buf sized for max frame (1920) with frame_samples-bounded slices
- RMS, mute, encode, and capture reads all use &buf[..frame_samples]
- Applied to both Android and desktop send tasks

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
Siavash Sameni
2026-04-13 11:33:02 +04:00
parent f17420aa98
commit 8ff0c548a7
1 changed files with 24 additions and 17 deletions
Download Patch File Download Diff File Expand all files Collapse all files

41
desktop/src-tauri/src/engine.rs
View File

@@ -538,12 +538,14 @@ impl CallEngine {
..CallConfig::default() ..CallConfig::default()
}, },
}; };
let frame_samples = (config.profile.frame_duration_ms as usize) * 48; let mut frame_samples = (config.profile.frame_duration_ms as usize) * 48;
info!(codec = ?config.profile.codec, frame_samples, t_ms = send_t0.elapsed().as_millis(), "first-join diag: send task spawned (android/oboe)"); info!(codec = ?config.profile.codec, frame_samples, t_ms = send_t0.elapsed().as_millis(), "first-join diag: send task spawned (android/oboe)");
*send_tx_codec.lock().await = format!("{:?}", config.profile.codec); *send_tx_codec.lock().await = format!("{:?}", config.profile.codec);
let mut encoder = CallEncoder::new(&config); let mut encoder = CallEncoder::new(&config);
encoder.set_aec_enabled(false); encoder.set_aec_enabled(false);
let mut buf = vec![0i16; frame_samples]; // Sized for max frame (40ms = 1920 samples) so profile
// switches between 20ms ↔ 40ms codecs don't need realloc.
let mut buf = vec![0i16; 1920];
// Continuous DRED tuning: poll quinn path stats every 25 // Continuous DRED tuning: poll quinn path stats every 25
// frames (~500 ms at 20 ms/frame) and adjust DRED duration + // frames (~500 ms at 20 ms/frame) and adjust DRED duration +
@@ -572,7 +574,7 @@ impl CallEngine {
// to read a full frame and sleep briefly if the ring is // to read a full frame and sleep briefly if the ring is
// short. Oboe's capture callback fills at a steady rate // short. Oboe's capture callback fills at a steady rate
// so in steady state this spins once per frame. // so in steady state this spins once per frame.
let read = crate::wzp_native::audio_read_capture(&mut buf); let read = crate::wzp_native::audio_read_capture(&mut buf[..frame_samples]);
if read < frame_samples { if read < frame_samples {
short_reads += 1; short_reads += 1;
tokio::time::sleep(std::time::Duration::from_millis(5)).await; tokio::time::sleep(std::time::Duration::from_millis(5)).await;
@@ -589,8 +591,8 @@ impl CallEngine {
} }
// RMS for UI meter // RMS for UI meter
let sum_sq: f64 = buf.iter().map(|&s| (s as f64) * (s as f64)).sum(); let sum_sq: f64 = buf[..frame_samples].iter().map(|&s| (s as f64) * (s as f64)).sum();
let rms = (sum_sq / buf.len() as f64).sqrt() as u32; let rms = (sum_sq / frame_samples as f64).sqrt() as u32;
send_level.store(rms, Ordering::Relaxed); send_level.store(rms, Ordering::Relaxed);
last_rms = rms; last_rms = rms;
if !first_nonzero_rms_logged && rms > 0 { if !first_nonzero_rms_logged && rms > 0 {
@@ -603,9 +605,9 @@ impl CallEngine {
} }
if send_mic.load(Ordering::Relaxed) { if send_mic.load(Ordering::Relaxed) {
buf.fill(0); buf[..frame_samples].fill(0);
} }
match encoder.encode_frame(&buf) { match encoder.encode_frame(&buf[..frame_samples]) {
Ok(pkts) => { Ok(pkts) => {
for pkt in &pkts { for pkt in &pkts {
last_pkt_bytes = pkt.payload.len(); last_pkt_bytes = pkt.payload.len();
@@ -646,8 +648,10 @@ impl CallEngine {
let p = send_pending_profile.swap(PROFILE_NO_CHANGE, Ordering::Acquire); let p = send_pending_profile.swap(PROFILE_NO_CHANGE, Ordering::Acquire);
if p != PROFILE_NO_CHANGE { if p != PROFILE_NO_CHANGE {
if let Some(new_profile) = index_to_profile(p) { if let Some(new_profile) = index_to_profile(p) {
info!(to = ?new_profile.codec, "auto: switching encoder profile"); let new_fs = (new_profile.frame_duration_ms as usize) * 48;
info!(to = ?new_profile.codec, frame_samples = new_fs, "auto: switching encoder profile (android)");
if encoder.set_profile(new_profile).is_ok() { if encoder.set_profile(new_profile).is_ok() {
frame_samples = new_fs;
dred_tuner.set_codec(new_profile.codec); dred_tuner.set_codec(new_profile.codec);
*send_tx_codec.lock().await = format!("{:?}", new_profile.codec); *send_tx_codec.lock().await = format!("{:?}", new_profile.codec);
} }
@@ -1353,12 +1357,12 @@ impl CallEngine {
..CallConfig::default() ..CallConfig::default()
}, },
}; };
let frame_samples = (config.profile.frame_duration_ms as usize) * 48; let mut frame_samples = (config.profile.frame_duration_ms as usize) * 48;
info!(codec = ?config.profile.codec, frame_samples, "send task starting"); info!(codec = ?config.profile.codec, frame_samples, "send task starting");
*send_tx_codec.lock().await = format!("{:?}", config.profile.codec); *send_tx_codec.lock().await = format!("{:?}", config.profile.codec);
let mut encoder = CallEncoder::new(&config); let mut encoder = CallEncoder::new(&config);
encoder.set_aec_enabled(false); // OS AEC or none encoder.set_aec_enabled(false); // OS AEC or none
let mut buf = vec![0i16; frame_samples]; let mut buf = vec![0i16; 1920]; // max frame (40ms)
// Continuous DRED tuning (same as Android send task). // Continuous DRED tuning (same as Android send task).
let mut dred_tuner = wzp_proto::DredTuner::new(config.profile.codec); let mut dred_tuner = wzp_proto::DredTuner::new(config.profile.codec);
@@ -1373,19 +1377,20 @@ impl CallEngine {
tokio::time::sleep(std::time::Duration::from_millis(5)).await; tokio::time::sleep(std::time::Duration::from_millis(5)).await;
continue; continue;
} }
capture_ring.read(&mut buf); capture_ring.read(&mut buf[..frame_samples]);
// Compute RMS audio level for UI meter // Compute RMS audio level for UI meter
if !buf.is_empty() { {
let sum_sq: f64 = buf.iter().map(|&s| (s as f64) * (s as f64)).sum(); let pcm = &buf[..frame_samples];
let rms = (sum_sq / buf.len() as f64).sqrt() as u32; let sum_sq: f64 = pcm.iter().map(|&s| (s as f64) * (s as f64)).sum();
let rms = (sum_sq / pcm.len() as f64).sqrt() as u32;
send_level.store(rms, Ordering::Relaxed); send_level.store(rms, Ordering::Relaxed);
} }
if send_mic.load(Ordering::Relaxed) { if send_mic.load(Ordering::Relaxed) {
buf.fill(0); buf[..frame_samples].fill(0);
} }
match encoder.encode_frame(&buf) { match encoder.encode_frame(&buf[..frame_samples]) {
Ok(pkts) => { Ok(pkts) => {
for pkt in &pkts { for pkt in &pkts {
if let Err(e) = send_t.send_media(pkt).await { if let Err(e) = send_t.send_media(pkt).await {
@@ -1406,8 +1411,10 @@ impl CallEngine {
let p = send_pending_profile.swap(PROFILE_NO_CHANGE, Ordering::Acquire); let p = send_pending_profile.swap(PROFILE_NO_CHANGE, Ordering::Acquire);
if p != PROFILE_NO_CHANGE { if p != PROFILE_NO_CHANGE {
if let Some(new_profile) = index_to_profile(p) { if let Some(new_profile) = index_to_profile(p) {
info!(to = ?new_profile.codec, "auto: switching encoder profile"); let new_fs = (new_profile.frame_duration_ms as usize) * 48;
info!(to = ?new_profile.codec, frame_samples = new_fs, "auto: switching encoder profile (desktop)");
if encoder.set_profile(new_profile).is_ok() { if encoder.set_profile(new_profile).is_ok() {
frame_samples = new_fs;
dred_tuner.set_codec(new_profile.codec); dred_tuner.set_codec(new_profile.codec);
*send_tx_codec.lock().await = format!("{:?}", new_profile.codec); *send_tx_codec.lock().await = format!("{:?}", new_profile.codec);
} }