* fix(loudness): target sync, -14 pre-analysis ref, queue UI, and reseed
- Front: coalesce loudness refresh by target LUFS; replay-gain IPC dedupe keys
include norm target and effective pre-attenuation so TGT changes apply.
- Rust: placeholder gain before integrated LUFS uses pivot at -14 LUFS; UI gain
from effective trim; reseed loudness after delete when waveform cache would skip.
- Pre-analysis: store attenuation relative to -14 LUFS; engine and UI use an
offset for other targets; migrate legacy absolute values on rehydrate.
- Queue/Settings: Loudness/TGT labels vs value buttons; styles; i18n for help.
* fix(i18n): simplify loudness pre-analysis helper copy
Remove reference-target wording from loudness pre-analysis helper text and keep
only the effective adjustment shown for the current LUFS target in all locales.
Serialize heavy PCM seeding through a dedicated queue with optional priority
for the current track. Emit waveform-updated once per completed seed, fix
Lucky Mix waveform refresh tokens, redact Subsonic URLs in logs, and align
hot-cache prefetch with the queued path.
When loudness normalization is on, every PARTIAL_LOUDNESS_EMIT_INTERVAL_MS
(900 ms per active stream task) the backend emits analysis:loudness-partial.
The frontend listener wrote cachedLoudnessGainByTrackId and unconditionally
called updateReplayGainForCurrentTrack, which invoked audio_update_replay_gain,
which in turn emitted audio:normalization-state back to the renderer — closing
a loop the UI had to drain on every tick. WebKitGTK absorbed it; WebView2 did
not, and after a few minutes the renderer thread stalled while Rust analysis
tasks kept progressing (visible in logs).
Five gates close the loop, smallest first:
Frontend (src/store/playerStore.ts):
- analysis:loudness-partial listener now short-circuits when the new gainDb is
within 0.05 dB of the cached value.
- audio_update_replay_gain calls go through invokeAudioUpdateReplayGainDeduped
(250 ms key-based dedupe), mirroring the audio_set_normalization helper from
#320.
- refreshLoudnessForTrack now coalesces concurrent calls per (trackId, mode)
via loudnessRefreshInflight, mirroring waveformRefreshInflight.
Backend (src-tauri/src/audio.rs):
- audio:normalization-state emits go through maybe_emit_normalization_state,
which keeps the last-emitted payload behind a Mutex and skips when engine
matches and current_gain_db (±0.05 dB) / target_lufs (±0.02) are unchanged.
Applied at all three emit sites: audio_play, audio_update_replay_gain,
audio_set_normalization.
- analysis:loudness-partial emits are gated by partial_loudness_should_emit:
per-track-identity last-emitted-gain map; an emit is suppressed when the new
gain is within PARTIAL_LOUDNESS_DELTA_THRESHOLD_DB (0.1 dB) of the previous
one. Applied to both emit_partial_loudness_from_bytes (legacy path) and the
inline ranged_download_task progress emit.
Drive-by: split estimated_output_latency_secs into two #[cfg]-gated
definitions so the sample_rate_hz parameter is no longer flagged as unused on
Windows / macOS builds. Function originally added in c6fc3ec.
Net effect: when loudness is steady (the common case), zero IPC traffic on
this path. UI stays responsive on Windows. Tested on Windows + Linux.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- Emit streaming partial loudness only when the loudness normalization engine is active
- Seed waveform/loudness from in-memory full buffers (preloaded, stream cache, gapless reuse) so offline hot cache does not rely on HTTP backfill
- Add explicit ranged/local/preload logs before full-file Symphonia + EBU analysis
- Coalesce concurrent waveform DB reads and skip duplicate audio_set_normalization within 450ms (e.g. StrictMode double init)
Resolve integrated loudness from SQLite only at decode bind; keep pre-trim
until a row exists, then allow provisional gain from live updates. Pass
DB-stable hints from the web app into play and gapless preload.
Default pre-measurement attenuation -4.5 dB with an icon reset in Settings;
drop redundant normalization copy and shorten pre-trim descriptions.
analysis_cache seed_from_bytes returns an outcome and skips redundant
waveform work on cache hits; wire callers and related frontend/backend glue.
Improve seekbar fallback behavior by using consistent bar-based rendering and smoother transitions to analyzed waveform data. Reduce hot-path analysis overhead during ranged streaming and add controls to clear cached waveform entries.
- Add Tauri command to delete loudness_cache rows for a track and helpers in AnalysisCache.
- Queue tech strip: click dB to reseed loudness; LUFS target picker via body portal; metric styling aligned with strip (no link chrome).
- Player store: reseed clears local cache and replays analysis seed; show loudness dB from SQLite cache when live state is still null; allow first numeric normalization-state update through the short duplicate filter.
- audio_update_replay_gain: resolve loudness from the requested gain when playback URL is not pinned yet.
Improve normalization consistency by separating cache refresh for non-playing tracks, hardening track-id cache lookups, and keeping UI gain state aligned with actual playback. Also reduce startup stalls by preferring non-seekable streaming when format hints are missing and by tuning ALSA/analysis paths to avoid underrun-prone churn.
On PipeWire-based distros (Debian 13, Ubuntu 22+, Gnome-on-PipeWire, and
similar), cpal's default_output_device() resolves to the raw ALSA `default`
alias. During early app-start that alias sometimes routes to a null sink:
the stream opens without error, progress ticks run, but nothing reaches the
user. Closes#234.
Prefer the pipewire-alsa / pulse-alsa aliases explicitly before falling
back to cpal's default, but only on Linux — macOS / Windows paths are
untouched. Systems that don't expose either alias (pure ALSA, no PipeWire)
fall through to the original default-device path.
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
`audio_chain_preload` runs ~30 s before the current track ends to queue the
next source for sample-accurate gapless playback. It was also calling
`Sink::set_volume(effective_volume)` with the *next* track's
volume*ReplayGain — but `Sink::set_volume` affects the whole Sink,
including the still-playing current source. Result: loudness of the
currently-playing track audibly shifted exactly 30 s before its end,
toward the next track's level.
Move `set_volume` to the gapless transition block in `spawn_progress_task`,
where `cur.replay_gain_linear` / `cur.base_volume` are already swapped.
The boundary itself is still sample-accurate; only the volume update is
deferred until the chained source is actually live.
Side effects audited:
- hi-res rate-mismatch fallback returns before the sink block — unaffected
- manual skip during chain pending: audio_play creates a fresh Sink — unaffected
- audio_set_volume / audio_update_replay_gain before transition operate on
the current track's gain — still correct
- after transition, cur.replay_gain_linear holds the chained track's gain
before the new set_volume runs, so user volume changes remain correct
- crossfade is mutually exclusive with gapless — separate code path
Closes#275
Co-authored-by: Psychotoxical <dev@psysonic.app>
Use the main window WebView user agent as the runtime source for Rust-side HTTP clients and refresh the audio engine client when the UA changes. This keeps backend and WebView requests aligned while preserving a simple startup sync flow.
Add a System setting for Off/Normal/Debug logging, apply readable local timestamps to backend logs, and enable exporting buffered runtime logs to a file when debug mode is active.
Co-authored-by: Maxim Isaev <im@friclub.ru>
* fix(player): prevent streaming seek UI freezes and progress snapbacks
Move blocking seek work off the command path with a bounded timeout, switch fullscreen/mobile seek bars to commit-on-release, and stabilize fallback progress state so rapid early seeks on streamed tracks do not freeze the UI or jump progress to zero.
* fix(player): avoid second-seek lockups and progress flicker
Add a short lock timeout for audio seek state access to prevent UI stalls when rapid seeks overlap, and keep waveform progress stable right after seek commit to eliminate brief visual snapbacks.
* fix(player): harden seek recovery and reduce stream log noise
Keep seek fallback stable during delayed backend responses to prevent occasional resets to track start, and remove per-read stream blocking logs so normal playback diagnostics stay readable.
---------
Co-authored-by: Maxim Isaev <im@friclub.ru>
Reworks the manual streaming-start path so playback is seekable from the
first frame and hot-cache hits start without a multi-hundred-MB pre-read.
Why
- The legacy AudioStreamReader was a non-seekable ring buffer. Forward
seeks during streaming let Symphonia consume the buffer through to EOF
(next song), and backward seeks were silently broken.
- fetch_data() loaded `psysonic-local://` files via tokio::fs::read,
blocking playback for seconds on hot-cache or offline files (e.g. a
100+ MB hi-res FLAC) and producing ALSA underruns when the new sink
swap arrived too late.
What changed
- New RangedHttpSource MediaSource: pre-allocates a Vec<u8> in track
size, background ranged_download_task fills linearly from offset 0
with HTTP Range reconnects. is_seekable=true; reads block on data,
seeks update the cursor only. Picked when the server response carries
Accept-Ranges: bytes + Content-Length.
- New LocalFileSource MediaSource: thin std::fs::File wrapper used for
every `psysonic-local://` URL — Symphonia probes the first ~64 KB and
starts playback before the rest is read.
- audio_play wires both paths through a generic PlayInput::SeekableMedia
variant so the build_streaming_source pipeline stays single-shape.
- current_is_seekable AtomicBool on AudioEngine: audio_seek short-circuits
with `not seekable` for the legacy fallback so the frontend restart-
fallback (playerStore seek catch) engages instead of letting a forward
seek consume the ring buffer.
- Dev-only [stream] / [seek] eprintln logs (cfg(debug_assertions)) so
source selection, codec resolution (codec name + lossless flag + bit
depth + rate + channels), download progress and seek targets are
visible in the terminal during testing.
- content_type_to_hint now covers wav/wave/opus; URL-derived format
hints strip the query string first and only accept known audio
extensions (Subsonic stream URLs have no extension — would otherwise
latch onto query-param fragments).
Compatibility
- Servers without Accept-Ranges fall back to the existing AudioStreamReader
path; seek is rejected up-front so the existing restart-fallback runs.
- Radio + audio_chain_preload paths untouched.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Add source badges for offline, cache, and stream playback in the queue tech line, including localized tooltips across shipped locales.
Track Rust preload-ready events by stream track id and latch the selected source per track, with dev logs to debug preload/source mismatches during next-track handoff.
Add resilient stream-first track playback with reconnect/range recovery, seek fallback handling for non-seekable starts, and promotion of completed streamed bytes into hot cache.
Also add crossfade/gapless backup preload windows to keep automatic transitions smooth on imperfect networks.
Downloaded and fully decoded the entire audio file on every track
change — too CPU/bandwidth intensive to keep. Removes SeekbarStyle
variant, Rust command, waveform_decode function, all frontend canvas
code, seekbarRealtime_waveform i18n keys across all 8 locales, and
unused urlencoding dependency from Cargo.toml.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Add CLI commands for stop, shuffle, repeat, mute, star, rating, reload,
server list/switch, and Subsonic search; publish extra snapshot fields and
server/search JSON for tooling. Single `play <id>` resolves track, album,
or artist; artists play a shuffled full discography.
Guard Linux tray creation with catch_unwind when Ayatana/AppIndicator .so
is missing. Harden bash completion against zsh sourcing (compopt). Narrow
AudioEngine field visibility to silence private_interfaces warnings.
Embed bash and zsh completion scripts and add a `completions` subcommand.
Extend the player CLI (library, audio device, instant mix) and surface
`music_library` in snapshots. Add a shared active-server switcher in the
header and locales; instant mix can reseed the queue from CLI and UI.
Wire a lazy CodecRegistry built with symphonia::default::register_enabled_codecs
plus symphonia_adapter_libopus::OpusDecoder for SizedDecoder and radio decoders.
Raise package rust-version to 1.89 to match the MSRV required by
symphonia-adapter-libopus.
* fix(audio): stabilize Linux output device picker and watcher
Keep pinned ALSA/cpal device ids stable when enumeration omits the active
sink or returns an equivalent name. Stop Linux device-watcher from clearing
the pin based on missing list entries; macOS and Windows still treat repeated
absence as unplugged. Settings refresh flow calls canonicalize and refetches
the list; add i18n for the out-of-list device label.
* fix(settings): sort audio output devices by label
cpal enumeration order is arbitrary; order the dropdown by readable label
and place the current OS default device first among concrete outputs.
* feat(settings): clearer audio device labels for duplicate ALSA names
Show HDMI outputs as "Card (HDMI n)" from hdmi:DEV indices; include PCM and
optional subdevice for hw/plughw/sysdefault; label other ALSA plugins with
iface and PCM. When labels still collide, append a structured hint
(iface · card · PCM) instead of only truncating the raw device string.
* feat(settings): improve audio output device picker
Parse ALSA-style ids into clearer labels (HDMI with DEV index, PCM/subdevice
for hw/plughw/sysdefault). Disambiguate colliding labels; share stderr
suppression for Linux device enumeration.
Add audio_default_output_device_name and tag the matching list entry as the
current system output (i18n). While the Audio tab is open, refresh list and
mark on audio:device-changed and audio:device-reset without toggling the
refresh spinner. Show an error toast if listing devices fails.
- Device watcher requires 3 consecutive misses (~9 s) before triggering
audio:device-reset, preventing false positives when ALSA temporarily
hides a busy device from output_devices() enumeration
- Add stderr suppression to open_stream_for_device_and_rate (last source
of ALSA terminal noise on Linux)
- Fix tooltip showing raw key 'common.refresh' — replaced with
settings.audioOutputDeviceRefresh (added to all 8 locales)
- Device switch restarts track from beginning (no seek-back)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Device watcher (3 s loop) and audio_list_devices both called
cpal output_devices(), triggering ALSA to probe unavailable backends
(JACK, OSS, dmix) and spam stderr with error messages.
Fix: redirect fd 2 to /dev/null for the duration of each enumeration
on Unix via libc dup/dup2 + RAII guard. Also add a module-level
cache in Settings.tsx so audio_list_devices is only invoked once
per app session instead of on every Audio tab activation.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Adds the ability to choose which audio output device Psysonic plays
through — useful for USB DACs, dedicated soundcards, and multi-device setups.
Rust (audio.rs):
- open_stream_for_device_and_rate(device_name, rate): opens a named
device by cpal name, falls back to system default if not found
- AudioEngine.selected_device: persists the chosen device across
stream reopens so hi-res rate switches stay on the right device
- audio_list_devices: Tauri command — returns all output device names
- audio_set_device: Tauri command — switches device immediately,
drops old sinks, emits audio:device-changed
- start_device_watcher: now handles two cases:
1. No pinned device + system default changed → reopen on new default
2. Pinned device disappeared (DAC unplugged) → clear selected_device,
fall back to system default, emit audio:device-reset
Frontend:
- authStore: audioOutputDevice (string | null), persisted
- playerStore: applies stored device on cold start
- App.tsx: listens to audio:device-reset, clears authStore device
and restarts playback on the fallback device
- Settings → Audio tab: device dropdown at top (above Hi-Res and EQ),
uses CustomSelect (portal-based, styled), all 8 locales
- CustomSelect: added disabled prop
Note: exclusive mode (WASAPI exclusive, CoreAudio exclusive) is out
of scope for now.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Lyrics Sources:
- Replace lyricsServerFirst + enableNeteaselyrics toggles with a new
Settings section — three sources (Server, LRCLIB, Netease) each
individually toggled and drag-reorderable via psy-drop DnD
- useLyrics.ts iterates sources in user-defined order, skipping disabled
ones; embedded SYLT from local files still wins unconditionally
- onRehydrateStorage migration from legacy fields on first load
ReplayGain Pre-Gain:
- New authStore fields: replayGainPreGainDb (0…+6 dB) and
replayGainFallbackDb (-6…0 dB, untagged / radio)
- audio.rs compute_gain applies pre_gain_db and fallback_db
- Settings sliders under ReplayGain mode selector
- Radio HTML5 volume scaled by fallbackDb factor on playRadio
i18n(ru): incorporate PR #148 translation improvements (kilyabin)
All 7 locales updated for new keys.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Use `let _ = msg` under `#[cfg(not(debug_assertions))]` so the binding
remains available in debug builds for the eprintln! format string.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Throttle audio:progress from 100ms to 500ms; WaveformSeek and
FsSeekbar use imperative DOM updates instead of React re-renders
- Fix all usePlayerStore() calls without selectors across pages and
components (useShallow / individual selectors throughout)
- Remove filter:blur and transform:scale from Hero, AlbumDetail and
FullscreenPlayer backgrounds — eliminated expensive software
compositing layers on WebKitGTK
- Replace translate3d with 2D translate in FS mesh and portrait
animations; remove will-change:transform from mesh blobs
- Move Hot Cache section from Storage tab to Audio tab; group Preload
and Hot Cache under a shared 'Next Track Buffering' section with a
mutual-exclusivity note and auto-disable logic
- Add 'off' toggle to Preload (replaces Off button with toggle switch);
enabling either method now automatically disables the other
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Adds a background device-watcher that polls the OS default output device
every 3 s via CPAL. When the device changes (Bluetooth headphones connecting,
USB DAC plugging in, etc.) the stream is reopened on the new device, the old
Sink is dropped (it was bound to the now-closed OutputStream), and
audio:device-changed is emitted to the frontend.
Frontend handler (TauriEventBridge): if playing, restarts the current track
from the saved position; if paused, clears the warm-pause flag so the next
resume uses the cold path (audio_play + seek) which creates a new Sink on
the new device.
Fixes#143 (audio through speakers after connecting Bluetooth headphones).
Also covers the intermittently reported single-channel output after long idle,
which can be caused by the OS reconfiguring the audio device in the background.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Revert _msg → msg in DecodeError match arm; the variable is used in the
debug_assertions eprintln! block so the underscore prefix caused a compile error
- Replace SlidersHorizontal with SlidersVertical (lucide-react) in PlayerBar,
LiveSearch, and AdvancedSearch so the EQ/filter icon shows vertical bars
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Set psysonic/<version> as User-Agent on the audio reqwest::Client to
prevent reverse proxies (nginx, Caddy, Traefik) from blocking requests
with a 403. Add #[cfg(debug_assertions)] logging in fetch_data to show
the exact URL, status, content-type and server header — useful for
diagnosing proxy-related 403s reported by users.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Replace the fixed DECODE_MAX_RETRIES (3) loop with a consecutive-error counter
that tolerates up to 100 bad frames before giving up — matching how VLC handles
files with a handful of invalid main_data offset frames. Frame-drop logs are
wrapped in #[cfg(debug_assertions)] so production builds stay silent.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Add Rust command `compute_waveform` that downloads the audio file,
seek-samples 500 evenly-spaced frames with Symphonia (fast path) and
computes peak amplitudes. Percentile-based normalisation (p5→p95)
preserves visible dynamics even for heavily compressed music.
Frontend caches results per track (module-level Map), shows the
pseudo-random fallback immediately and crossfades to the real waveform
over 400 ms (ease-in-out) once the computation finishes.
Also: Settings input-tab reset buttons now use btn-danger styling.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
M4A files store the iTunSMPB value behind a 16-byte binary 'data' atom
header. The previous parser took bytes directly after the ASCII key,
landing in the binary header and producing a corrupted parts array
(total_valid=348 instead of ~13M samples), which caused take_duration
to exhaust the source in ~8ms — silent no-playback.
Fix: search for the " 00000000 " sentinel within 128 bytes of the key
to locate the actual value string, falling back to the old offset for
ID3v2/MP3 files where no binary header exists.
Also patches symphonia-format-isomp4 (local crate under patches/) to:
- Tolerate SL descriptor predefined=0x01 (older iTunes M4A)
- Convert descriptor.unwrap() panic to a proper decode error
- Gracefully skip malformed trak atoms (e.g. MJPEG cover-art streams)
instead of failing the entire probe
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The previous optimization skipped all rate-switch logic in standard mode,
but left the stream at e.g. 88.2 kHz if Hi-Res was toggled off mid-session.
Fix: store device_default_rate (from cold-start open) in AudioEngine.
In audio_play, compute target_rate = hi_res ? output_rate : device_default_rate
and reopen only when target differs from current — no-op cost when already
at the right rate (the common case in pure standard mode).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Stream switch: skip entirely when hi_res_enabled=false. Previously the
engine compared 44100 vs. device-default 48000 Hz and re-opened the device
on every track start — unnecessary IPC + PipeWire reconfigure in safe mode.
MSS buffer: 512 KB in standard mode (was 4 MB). For in-memory MP3/AAC
the large buffer caused eager allocation competing with playback startup.
4 MB is kept only for hi-res (high-bitrate FLAC benefits from fewer reads).
ThreadPriority: no longer set Max at audio-stream-thread startup. Max is
now applied only when a hi-res reopen is requested, keeping the scheduler
budget for the actual decode threads during standard playback.
audio_preload: 8 s throttle delay + gen guard so background pre-fetch does
not compete with the decode/sink-feed cycle of the just-started track.
needs_prefill: now gated on hi_res_enabled to make the condition explicit.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The native-tls backend (macOS Security framework) was failing with
'bad protocol version' when connecting to HTTPS music servers.
Switching to rustls (statically linked, no system dependency) fixes
playback on macOS and ensures consistent TLS 1.2/1.3 support across
all platforms.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Internet Radio full release (HTML5 engine, card UI, RadioDirectoryModal,
cover upload), Backup/Restore, Albums year filter, Statistics Library
Insights (playtime/genres/formats), Playlist cover upload, resizable
tracklist columns for Playlists & Favorites, crossfade fine control,
Settings Storage tab redesign, various fixes and UI polish.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Incorporates PRs #38 (nisarg-78), #42#43#44 (JulianNymark) with fixes:
Audio / Playback
- Artist Radio starts immediately from fast getTopSongs (local library);
getSimilarSongs2 (Last.fm) enriches queue in background — no more wait
- Fix seek audio glitch: EqualPowerFadeIn only resets to zero-gain on
seeks to track start (<100 ms); all other seeks resume at unity gain
- OGG/Vorbis container support via symphonia-format-ogg (PR #42)
- Human-readable audio error messages in SizedDecoder (PR #44)
- Audio playback errors shown as themed toast notifications (PR #43)
Queue / Radio
- Infinite Queue: proactive load of 5 tracks (was 25) when ≤2 remain
- Radio: proactive reload at ≤2 remaining tracks, independent of
Infinite Queue setting — radio no longer stops at last track
- Fix: clicking Start Radio multiple times no longer stacks duplicates
- Fix: Start Radio on artist keeps current song playing
- Manual tracks always appear before Radio, Radio before Auto-added
- Queue separators: "— Radio —" and "— Auto —" dividers
- Fix: radio proactive load now works even when songs lack artistId
(uses currentRadioArtistId module var as fallback)
UI / UX
- Click synced lyrics lines to seek (PR #38)
- Volume scroll wheel on volume slider (PR #38)
- Lyrics: active / completed / upcoming visual states (PR #38)
- Shared toast utility (src/utils/toast.ts) replaces inline toast fn
- Auto-updater: relaunch_after_update Rust command exits first to
release single-instance lock before spawning new process
About / Credits
- nisarg-78 and JulianNymark added to contributors (since v1.29.0)
- netherguy4 added as Special Thanks for feature ideas and feedback
- i18n: aboutSpecialThanksLabel in all 5 languages
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
current_sample_rate was initialized to 44100, causing every track to be
resampled down to 44.1 kHz. Setting it to 0 disables resampling until
the actual track rate is known, so songs play at their native sample rate.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Replay gain was not working because track objects were created manually without
including replay gain metadata from the Subsonic API response.
Changes:
- Add songToTrack() helper function to properly map SubsonicSong to Track with
replayGainTrackDb, replayGainAlbumDb, and replayGainPeak fields
- Add audio_update_replay_gain Tauri command for dynamic volume recalculation
when replay gain settings change mid-playback
- Add updateReplayGainForCurrentTrack() to recalculate volume when toggling
replay gain setting
- Fetch fresh track data on cold resume (app relaunch) to ensure replay gain
values are current from server
- Update all files that create track objects to use songToTrack()
Fixes issue where toggling replay gain ON/OFF or changing between track/album
mode had no effect on currently playing or newly played tracks.
cucadmuh
Psychotoxical
Maxim Isaev
Frank Stellmacher
Psychotoxical
Nils Israel
trbn