Adds internal/audiofeatures: decodes each scanned song to PCM via
ffmpeg and computes actual measured characteristics — tempo (via
autocorrelation of the frame energy envelope, left unset rather than
guessed when a clip is too short to contain a beat), an approximate
loudness figure, spectral centroid/rolloff and zero-crossing rate (via
a hand-written FFT), and a 12-bin chroma vector used both for a
Krumhansl-Kessler key estimate and as similarity-index input. No
external DSP/ML dependency.
The scanner now runs this analysis after indexing each song (as part
of the already-background scan, not on any request path) and stores
results in the existing audio_features table via the new
database.UpsertAudioFeatures/GetAudioFeatures/ListAllAudioFeatures.
internal/similarity is a small in-process nearest-neighbor index:
z-score normalizes each feature dimension across the analyzed library
and answers Nearest (ranked real distances) and Path (a feature-space
straight-line walk from one track to another, hopping to the closest
not-yet-used real song at each step — never a fabricated intermediate
track).
getSonicSimilarTracks and findSonicPath expose these through the
standard sonicMatch response, converting distance to a normalized
(0,1] similarity score. A song that hasn't been analyzed yet is simply
absent from results rather than faked.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014r98AtcBbevpscWLsfwVvv
Adds a real capability-matching decision engine: getTranscodeDecision
compares the source file's actual scanned format/channels against the
client's declared directPlayProfiles first, and only when nothing
matches does it search transcodingProfiles for a codec VibeSonic's
ffmpeg pipeline can actually encode (mp3/opus/aac), applying any
codecProfile limitations (bitrate/channel required constraints) the
client listed. canTranscode is never set true for a codec the server
has no encoder for, and canDirectPlay is never set true without a
matching client profile.
getTranscodeStream takes the opaque transcodeParams token the
decision returned (never re-derived) plus a real time offset and
streams actual transcoded audio via the existing seek-capable
transcoding.Manager.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014r98AtcBbevpscWLsfwVvv
Adds getLyricsBySongId (versions 1 and 2 of the songLyrics extension,
gated by the enhanced param): returns every stored structuredLyrics
entry for a song, including line-level start times for synced lyrics
and, only when a source actually provided word/syllable timing (an
Enhanced LRC sidecar) and enhanced=true was requested, cueLine/cue
data with the precomputed UTF-8 byte offsets. Plain and non-enhanced
synced lyrics never get fabricated cue data.
Removes the now-dead SetSongLyrics writer (superseded by
database.InsertLyrics); SongLyricsPlainText keeps serving the legacy
getLyrics endpoint unchanged.
Adds a scanner integration test proving a sidecar .lrc (with an
Enhanced LRC line) is discovered and indexed during a real scan.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014r98AtcBbevpscWLsfwVvv
owenandClaude Sonnet 5