EdisonJwa 6e0bf21295 fix(audio,ios): route playback via media channel (.default + .defaultToSpeaker) (rc.8+70)
User report after the 8x boost commit (e85a6d3): playback STILL
broken, but now the diagnostic clearly shows the actual problem.
Render-callback peak_out_i16 SATURATES at 32767 on speech peaks
(cb=600, 1100, 1200, 2400) because the 8x boost amplifies an
already-loud signal into hard clipping. Quiet content reaches
audible level but loud peaks are catastrophically distorted.

The 8x boost was treating the wrong cause.

Real root cause (researched online after user prompted: 'this is
iOS a popular platform, there must be solutions'): iOS has TWO
independent audio channels:

  In-call channel  (.voiceChat / .videoChat modes)
    * Routes through the phone-call audio path.
    * Aggressively ducks non-voice content to the earpiece.
    * Volume controlled by a separate in-call hardware
      register, not the side buttons when not actively on a
      phone call.

  Media channel  (.default mode)
    * Routes through the standard media playback path.
    * No automatic ducking.
    * Volume controlled by the side volume buttons normally.

With AVAudioSession mode .voiceChat, iOS sends our output
through the in-call channel which plays at 'earpiece-level'
loudness on the speaker too. Signal is technically present but
buried under the speaker's noise floor. With mode .default +
.defaultToSpeaker option, output routes via media channel and
plays at normal loudness.

Both Twilio (video-quickstart-ios) and Daily.co (patched WebRTC
module) document the same workaround and use VPIO for AEC while
keeping the session mode at .default for loud playback:

  github.com/twilio/video-quickstart-ios/issues/522
  stackoverflow.com/questions/79834998 (Daily.co)

The user also noticed 'tx/rx almost no changes even receiving
packages' \u2014 likely a misinterpretation of the frames counter
not advancing as fast as expected during quiet voice; AudioHandler
returns silence when its jitter buffer is in buffering_samples
state which doesn't fire 'decode failed' but also doesn't
increment frames_received. The real issue is still the playback
ducking; the counter behaviour is a downstream symptom.

Changes:

1. AppDelegate.swift: AVAudioSession mode .voiceChat -> .default
   with options [.defaultToSpeaker, .allowBluetoothHFP,
   .allowBluetoothA2DP]. VPIO continues to do its job (AEC, NS,
   AGC on the mic side); only the playback routing changes.
   The earlier 'speaker selector silent under .default' bug
   does NOT apply because we no longer use cpal RemoteIO \u2014
   VPIO honours overrideOutputAudioPort under any mode.

2. ios_voice_unit.rs: revert the 8x output boost from e85a6d3.
   With media-channel routing, signal levels are correct and
   no software amplification is needed. Render callback restored
   to plain (l+r)*0.5*gain downmix.

3. ios_voice_unit.rs: revert the BypassVoiceProcessing toggle
   from c16318c. The VPIO chain stays enabled so we keep
   capture-side AEC/AGC/NS for free \u2014 the playback breakage
   it was trying to fix was the wrong layer all along.

4. ios_voice_unit.rs: drop the diagnostic render-callback log
   line. Production-clean code; can be re-enabled by reverting
   the diff in the closure if future debugging needs it.

Build counter 69 -> 70.
2026-05-17 02:11:07 +08:00

Chanora

Chanora is a cross-platform voice communication client for TeamSpeak-compatible servers.

It is built with a shared Flutter UI and a Rust core, with TeamSpeak-compatible protocol integration isolated behind tsclientlib.

Flutter UI + Rust Core + tsclientlib

Chanora is an independent project and is not affiliated with, endorsed by, sponsored by, or officially associated with TeamSpeak.


Status

Chanora is currently in early planning and baseline-candidate design.

Current documentation baseline: v0.9.2
Current status: Baseline Candidate
Implementation status: Not production-ready

The current engineering focus is:

  • defining the system and software architecture;
  • preparing the Flutter + Rust application structure;
  • validating TeamSpeak-compatible protocol integration through tsclientlib;
  • defining cross-platform audio behavior;
  • preparing release, verification, security, privacy, and legal gates.

Target Platforms

Chanora is intended to support:

  • Windows
  • macOS
  • Linux
  • Android
  • iOS / iPadOS

Current platform policy:

Platform Baseline
iOS / iPadOS runtime target iOS 13+ unless Flutter, plugin, audio, or product constraints require raising it
App Store Connect upload gate Xcode 26+ with iOS 26 / iPadOS 26 SDK+ for upload on or after 2026-04-28
Android runtime target Android API 24+ unless Flutter, plugin, audio, or product constraints require raising it
Google Play target API Target the Google Play-required API level on upload date

The App Store / Play Store upload gates are release requirements. They are separate from local development and internal testing requirements.


Architecture Overview

Chanora separates UI, protocol logic, state synchronization, audio processing, diagnostics, and platform services.

Flutter Application
  ├─ App Shell
  ├─ Material 3 / Chanora Design System
  ├─ Feature Modules
  ├─ View Models / State
  └─ Typed Flutter/Rust Bridge

Rust Core
  ├─ Connection Manager
  ├─ State Synchronization
  ├─ Protocol Adapter
  ├─ Audio Subsystem
  ├─ Storage Services
  └─ Diagnostics

Protocol Layer
  └─ tsclientlib
      └─ TeamSpeak-compatible server

Key architecture rules:

  • Flutter does not call tsclientlib directly.
  • Protocol-specific types do not leak into the Flutter UI layer.
  • Rust Core owns protocol coordination, state synchronization, audio logic, storage services, diagnostics, and bridge-facing DTOs.
  • Flutter owns presentation, navigation, Material 3 theming, accessibility, localization presentation, and platform UI behavior.
  • Product localization and server-provided content are separated.
  • UTF-8 is the internal cross-layer text representation.
  • Non-UTF-8 conversion, if needed, occurs only at explicit protocol or platform boundaries.

MVP Direction

The current recommended MVP scope is:

Area MVP decision
Active server connections One active server connection per client instance
UI baseline Material 3 + Chanora Design System
Product language English UI first, i18n-ready architecture
Server content Preserve Unicode and do not translate server-provided content
Audio processing defaults Echo Canceller, Automatic Gain Control, Noise Suppression, and High-Pass Filter enabled where supported and stable
Audio implementation path Platform-native first; fallback isolated behind the audio subsystem
Local non-secret storage SQLite or equivalent embedded database
Secret storage Platform secure storage
Flutter/Rust bridge Stable typed bridge with generated or schema-controlled DTOs
Diagnostics Local, user-initiated export only
Telemetry None in MVP
Crash reporting Disabled unless explicitly approved later

Desktop Push-to-Talk

Chanora's desktop Push-to-Talk (PTT) follows a capability-based design (see docs/architecture/desktop-ptt-architecture.md). Focused PTT — the user holds a bound key or mouse button inside the focused Chanora window — is mandatory on Windows, macOS, and Linux. Global PTT (recognised while the application is not focused) is capability-dependent: it requires the operating system, the user-granted permission set, the display server, and the available input backend to all permit it.

The application reports a PttCapabilityLevel (L0Focused, L1GlobalShortcut, L2GlobalHoldToTalk, L3GlobalWithMouseButtons) that matches actual runtime behaviour, not the platform's theoretical maximum. The UI capability badge shows the live value.

Per-platform strategy (resolved by owner rulings 2026-05-15, see docs/governance/product-decision-register.md DEC-023 through DEC-028):

  • Windows — Raw Input first, low-level keyboard hook fallback, Focused PTT terminal fallback. Mouse side buttons supported. P0 / MVP.
  • macOS — permission-aware Event Tap with Focused PTT fallback; Global PTT upgrades asynchronously when the user grants Input Monitoring / Accessibility. P0 / MVP.
  • Linux — officially tested on GNOME on Wayland using the org.freedesktop.portal.GlobalShortcuts interface; every other Linux environment falls back to Focused PTT. Release notes do not claim Global PTT support outside the tested compositor.
  • Raw key codes, scan codes, virtual-key values, keysyms, and key-press timing sequences are never logged or included in the user-initiated diagnostic export. The diagnostic export carries only capability level, backend identifier, and bound input class.

A missed-key-up watchdog (default 30 s) clears transmit_active when the OS suppresses a key-up event so a stuck-PTT bug class is ruled out by construction.

Repository Layout

The repository documentation is expected to live under docs/.

docs/
  requirements/
    sysrs.md
    srs.md

  architecture/
    sysdes.md
    sad.md
    sdd.md

  verification/
    verification-master-plan.md
    swe4-unit-verification-plan.md
    swe5-software-integration-verification-plan.md
    swe6-software-verification-plan.md
    sys4-system-integration-verification-plan.md

  release/
    release-readiness-go-nogo-record.md
    platform-release-policy.md

  security/
    security-privacy-legal-guideline.md
    threat-model.md
    secure-storage-audit-report.md
    diagnostic-redaction-audit-report.md
    dependency-and-supply-chain-report.md

  privacy/
    privacy-policy.md

  legal/
    trademark-and-attribution-review.md

  ui-ux/
    material3-guideline.md
    material3-design-tokens.md
    material3-component-catalog.md
    adaptive-layout-platform-guide.md

  i18n/
    localization-architecture.md

  governance/
    document-index.md
    document-naming-convention.md
    traceability-matrix.md
    baseline-approval-record.md
    baseline-candidate-validation-report.md
    document-review-report.md
    product-decision-register.md
    decision-impact-assessment.md
    git-commit-message-convention.md
    repo-format-validation-report.md
    path-migration-map.md

  references/
    external-references.md
    aspice-swe2-swe3-integration-note.md

Implementation source folders may be added later. A likely structure is:

apps/
  chanora_flutter/

core/
  chanora_core/

crates/
  chanora_protocol/
  chanora_audio/
  chanora_state/
  chanora_storage/
  chanora_diagnostics/
  chanora_bridge/

The exact implementation layout should be finalized when the repository scaffold is created.


Documentation Entry Points

Start here:

Topic Document
System requirements docs/requirements/sysrs.md
Software requirements docs/requirements/srs.md
System architecture docs/architecture/sysdes.md
Software architecture docs/architecture/sad.md
Software detailed design docs/architecture/sdd.md
Verification strategy docs/verification/verification-master-plan.md
Release readiness docs/release/release-readiness-go-nogo-record.md
Platform release policy docs/release/platform-release-policy.md
Product decisions docs/governance/product-decision-register.md
Traceability docs/governance/traceability-matrix.md
Security/privacy/legal gates docs/security/security-privacy-legal-guideline.md

Engineering Process

Chanora follows this documentation hierarchy:

SysRS -> SysDes -> SRS -> SAD -> SDD

Direct traceability rules:

Document Direct upstream source
SysDes SysRS
SRS SysDes only
SAD SRS only
SDD SAD only

Verification mapping:

SDD -> SWE.4 Unit Verification
SAD + SDD -> SWE.5 Software Integration Verification
SRS -> SWE.6 Software Verification
SysDes -> SYS.4 System Integration Verification

Release readiness is tracked separately through the Go/No-Go record.


Release Readiness

A release is not approved by design documents alone.

Before an external or public release, the project must complete:

docs/release/release-readiness-go-nogo-record.md

The release decision must explicitly state:

Go
Conditional Go
No-Go

Release readiness must include:

  • release scope;
  • build number;
  • commit SHA;
  • Git tag;
  • artifact hashes;
  • satisfied P0/MVP requirements;
  • deferred requirements;
  • verification results;
  • waivers;
  • security review status;
  • platform readiness;
  • legal and OSS review status;
  • privacy policy status;
  • approval decision and approvers.

Security, privacy, and legal evidence are required before public or store release.

Required documents include:

docs/security/threat-model.md
docs/security/secure-storage-audit-report.md
docs/security/diagnostic-redaction-audit-report.md
docs/security/dependency-and-supply-chain-report.md
docs/privacy/privacy-policy.md
docs/legal/trademark-and-attribution-review.md

Important gates:

  • identity secrets and server passwords must use platform secure storage;
  • logs and diagnostic exports must redact secrets;
  • diagnostic export must be user-initiated unless a later approved policy changes this;
  • dependency licenses and vulnerabilities must be reviewed;
  • OSS notices must be prepared where required;
  • public wording must not imply official TeamSpeak affiliation;
  • privacy policy must describe local storage, diagnostics, permissions, and data handling.

Git Commit Convention

Chanora uses a Conventional Commits style format:

<type>(<scope>): <summary>

Examples:

feat(voice): add push-to-talk state handling
fix(protocol): recover channel tree after reconnect snapshot
docs(sad): add interface catalog and performance view
i18n(ui): add fallback behavior for missing localization keys
sec(diagnostics): redact server password from export bundle
release(android): prepare internal alpha build metadata

See:

docs/governance/git-commit-message-convention.md

Development

Implementation commands will be added after the repository scaffold is finalized.

Expected future commands may include:

flutter pub get
flutter test
cargo test
cargo clippy
cargo fmt

Do not treat these as authoritative until the actual Flutter/Rust workspace has been created.


Contributing

Before making a change:

  1. Check the affected requirement/design document.
  2. Confirm the correct traceability layer.
  3. Use the Git commit convention.
  4. Update docs and verification plans when the change affects requirements, architecture, detailed design, release behavior, security, privacy, or legal gates.

License

Chanora is dual-licensed under either of:

at your option. This dual-license model was Accepted on 2026-05-14 as decision DEC-020 in docs/governance/product-decision-register.md.

Unless you explicitly state otherwise, any contribution intentionally submitted for inclusion in Chanora by you, as defined in the Apache-2.0 license, shall be dual-licensed as above, without any additional terms or conditions.

Third-party software bundled or linked by Chanora is listed in NOTICE with its own licenses. The complete legal review of the dependency tree (DEC-012) must complete before any public/store release. See:

docs/governance/product-decision-register.md
docs/security/dependency-and-supply-chain-report.md
docs/legal/trademark-and-attribution-review.md
S
Description
No description provided
Readme
26 MiB
Languages
Rust 55.3%
Dart 34.4%
Kotlin 3.6%
Swift 1.9%
Shell 1.4%
Other 3.3%