feat(review): vocals waveform in timeline editor — finished + fixed (takes over #895) (#927)

Vocals waveform in the timeline editor (finishes @netux's #895): backend stem-selection fix (vocals_clean), downsampled peak-envelope perf, edit-modal space reclaim, full i18n (33 locales), CodeRabbit fixes, and ReviewServer.vocals_filepath made optional. Reviewed + visually QA'd on a real job.
This commit is contained in:
Andrew Beveridge ☄️ authored and GitHub committed 2026-08-23 00:23:18 -04:00
1 parent 0497168dda
commit d7482ee85a
46 files changed
+1165 -450

No files matched your search

+27 -3
View File
@@ -19,7 +19,7 @@ import json
import os
import tempfile
from pathlib import Path
from typing import Dict, Any, List, Optional, Set, Tuple
from typing import Dict, Any, List, Literal, Optional, Set, Tuple
from fastapi import APIRouter, HTTPException, Request, Depends, Form, File, UploadFile
from fastapi.responses import RedirectResponse
@@ -372,6 +372,15 @@ async def get_correction_data(
raise HTTPException(status_code=500, detail=t("en", "review.correctionsLoadFailed", error=str(e)))
@router.get("/{job_id}/audio/vocals")
async def get_vocals_audio(
job_id: str,
auth_info: Tuple[str, str] = Depends(require_review_auth)
):
"""Stream the vocals audio file for playback."""
return await _stream_audio(job_id, stem="vocals")
@router.get("/{job_id}/audio/{audio_hash}")
async def get_audio_with_hash(
job_id: str,
@@ -392,7 +401,7 @@ async def get_audio_no_hash(
return await _stream_audio(job_id)
async def _stream_audio(job_id: str):
async def _stream_audio(job_id: str, stem: Literal["input"] | Literal["vocals"] = "input"):
"""
Redirect to a signed GCS URL for audio playback in the review interface.
@@ -407,7 +416,22 @@ async def _stream_audio(job_id: str):
if not job:
raise HTTPException(status_code=404, detail=t("en", "review.jobNotFound"))
audio_gcs_path = job.input_media_gcs_path
audio_gcs_path = None
if stem == "input":
audio_gcs_path = job.input_media_gcs_path
elif stem == "vocals":
stems = job.file_urls.get("stems", {}) if job.file_urls else {}
# Stem keys vary by separation model. We want the full vocal mix
# (lead + backing) so every word can be lined up against the waveform:
# - "vocals" : full vocal from a 2-stem split (rare)
# - "vocals_clean" : full vocal from the primary vocal/instrumental
# split (present on essentially all cloud jobs)
# - "lead_vocals" : last-resort fallback (misses backing lines)
# Use .get() so a missing key yields a clean 404 below, not a KeyError/500.
for key in ("vocals", "vocals_clean", "lead_vocals"):
if stems.get(key):
audio_gcs_path = stems[key]
break
if not audio_gcs_path:
raise HTTPException(status_code=404, detail=t("en", "review.audioNotFound"))
+68
View File
@@ -1132,3 +1132,71 @@ class TestGetPreviewVideoRedirect:
self._call(mock_storage)
assert exc_info.value.status_code == 404
mock_storage.generate_signed_url.assert_not_called()
class TestGetVocalsAudio:
"""The vocals waveform endpoint streams the full vocal stem for the timeline.
Real jobs never store a plain 'vocals' key — separation writes
'vocals_clean' / 'lead_vocals' / 'backing_vocals'. The endpoint must select
the fullest vocal mix (vocals_clean) rather than KeyError on 'vocals'.
"""
def _call(self, stems):
import asyncio
from backend.api.routes.review import _stream_audio
job = MagicMock()
job.input_media_gcs_path = "jobs/j/input.flac"
job.file_urls = {"stems": stems}
job_manager = MagicMock()
job_manager.get_job.return_value = job
transcoding = MagicMock()
transcoding.get_review_audio_url_async = AsyncMock(
return_value="https://signed.example/vocals.ogg?sig=xyz"
)
with patch("backend.api.routes.review.JobManager", return_value=job_manager), patch(
"backend.services.audio_transcoding_service.AudioTranscodingService",
return_value=transcoding,
):
resp = asyncio.run(
_stream_audio(job_id="j", stem="vocals")
)
return resp, transcoding
def test_selects_vocals_clean_when_no_plain_vocals_key(self):
from starlette.responses import RedirectResponse
# The shape produced by real cloud jobs (no plain "vocals").
resp, transcoding = self._call(
{
"instrumental_clean": "jobs/j/stems/instrumental_clean.flac",
"vocals_clean": "jobs/j/stems/vocals_clean.flac",
"lead_vocals": "jobs/j/stems/lead_vocals.flac",
"backing_vocals": "jobs/j/stems/backing_vocals.flac",
}
)
assert isinstance(resp, RedirectResponse)
assert resp.status_code == 302
# Must pick the full vocal mix (vocals_clean), not lead_vocals or KeyError.
assert transcoding.get_review_audio_url_async.call_args[0][0] == "jobs/j/stems/vocals_clean.flac"
def test_prefers_plain_vocals_then_falls_back_to_lead(self):
# 2-stem split exposes a plain "vocals" — preferred.
_, transcoding = self._call({"vocals": "jobs/j/stems/vocals.flac"})
assert transcoding.get_review_audio_url_async.call_args[0][0] == "jobs/j/stems/vocals.flac"
# Only lead_vocals available — last-resort fallback.
_, transcoding = self._call({"lead_vocals": "jobs/j/stems/lead_vocals.flac"})
assert transcoding.get_review_audio_url_async.call_args[0][0] == "jobs/j/stems/lead_vocals.flac"
def test_404_when_no_vocal_stem_present(self):
from fastapi import HTTPException
with pytest.raises(HTTPException) as exc_info:
self._call({"instrumental_clean": "jobs/j/stems/instrumental_clean.flac"})
assert exc_info.value.status_code == 404
@@ -1,8 +1,9 @@
'use client'
import { memo, useRef, useCallback } from 'react'
import { useTranslations } from 'next-intl'
import { Button } from '@/components/ui/button'
import { Play, XCircle, Hand } from 'lucide-react'
import { Hand } from 'lucide-react'
import { Word } from '@/lib/lyrics-review/types'
import TimelineEditor from './TimelineEditor'
import { cn } from '@/lib/utils'
@@ -17,6 +18,7 @@ const TapButton = memo(function TapButton({
onTapStart: () => void
onTapEnd: () => void
}) {
const t = useTranslations('lyricsReview.editTimeline')
const isPressedRef = useRef(false)
const handleTapStart = useCallback(() => {
@@ -51,7 +53,7 @@ const TapButton = memo(function TapButton({
)}
>
<Hand className="h-5 w-5 mr-2" />
{isSpacebarPressed ? 'HOLD...' : 'TAP'}
{isSpacebarPressed ? t('hold') : t('tap')}
</Button>
)
})
@@ -60,17 +62,12 @@ interface EditTimelineSectionProps {
words: Word[]
startTime: number
endTime: number
originalStartTime: number | null
originalEndTime: number | null
currentStartTime: number | null
currentEndTime: number | null
currentTime?: number
isManualSyncing: boolean
syncWordIndex: number
isSpacebarPressed: boolean
onWordUpdate: (index: number, updates: Partial<Word>) => void
onPlaySegment?: (time: number) => void
startManualSync: () => void
isGlobal?: boolean
onTapStart?: () => void
onTapEnd?: () => void
@@ -80,21 +77,17 @@ export default function EditTimelineSection({
words,
startTime,
endTime,
originalStartTime,
originalEndTime,
currentStartTime,
currentEndTime,
currentTime,
isManualSyncing,
syncWordIndex,
isSpacebarPressed,
onWordUpdate,
onPlaySegment,
startManualSync,
isGlobal = false,
onTapStart,
onTapEnd,
}: EditTimelineSectionProps) {
const t = useTranslations('lyricsReview.editTimeline')
// Simple mobile detection
const isMobile = typeof window !== 'undefined' && window.innerWidth < 640
@@ -110,8 +103,9 @@ export default function EditTimelineSection({
return (
<>
{/* Timeline Editor */}
<div className={cn('mb-0', isMobile ? 'h-20' : 'h-[120px]')}>
{/* Timeline Editor — height hugs its content (ruler + word bars + waveform)
instead of a fixed min-height, to avoid dead vertical space. */}
<div>
<TimelineEditor
words={words}
startTime={startTime}
@@ -122,70 +116,32 @@ export default function EditTimelineSection({
/>
</div>
{/* Time range info and controls */}
<div
className={cn(
'flex items-center justify-between gap-2',
isMobile ? 'flex-col items-stretch mt-2 mb-3' : 'flex-row mt-0 mb-0'
)}
>
{/* Time range info - hidden on mobile to save space */}
{!isMobile && (
<div className="text-xs text-muted-foreground">
<div>
Original Time Range: {originalStartTime?.toFixed(2) ?? 'N/A'} -{' '}
{originalEndTime?.toFixed(2) ?? 'N/A'}
</div>
<div>
Current Time Range: {currentStartTime?.toFixed(2) ?? 'N/A'} -{' '}
{currentEndTime?.toFixed(2) ?? 'N/A'}
</div>
</div>
)}
{/* Controls */}
{/* Sync helper — only visible while manually syncing. The Tap To Sync
toggle and the time-range readout now live in the modal header to
reclaim the vertical space above the word list. */}
{(currentWordInfo || (isMobile && isManualSyncing)) && (
<div
className={cn(
'flex items-center gap-2',
isMobile ? 'flex-col items-stretch' : 'flex-row'
isMobile ? 'flex-col items-stretch mt-2 mb-3' : 'flex-row mt-0 mb-0'
)}
>
{/* Tap To Sync button */}
<Button
variant={isManualSyncing ? 'outline' : 'default'}
onClick={startManualSync}
className={cn(
isManualSyncing
? 'border-destructive text-destructive hover:bg-destructive/10'
: 'bg-primary'
)}
>
{isManualSyncing ? (
<>
<XCircle className="h-4 w-4 mr-1" />
Cancel
</>
) : (
<>
<Play className="h-4 w-4 mr-1" />
Tap To Sync
</>
)}
</Button>
{/* Current word info during sync */}
{currentWordInfo && (
<div className={cn('text-center', isMobile ? '' : 'text-left')}>
<div className="text-sm">
Word {currentWordInfo.index} of {currentWordInfo.total}:{' '}
{t('wordProgress', {
index: currentWordInfo.index,
total: currentWordInfo.total,
})}{' '}
<strong>{currentWordInfo.text}</strong>
</div>
<div className="text-xs text-muted-foreground">
{isSpacebarPressed
? 'Holding... Release when word ends'
? t('holdingRelease')
: isMobile
? 'Tap the button when word starts'
: 'Press spacebar when word starts (tap for short words, hold for long words)'}
? t('tapWhenStarts')
: t('pressSpacebarHint')}
</div>
</div>
)}
@@ -199,7 +155,7 @@ export default function EditTimelineSection({
/>
)}
</div>
</div>
)}
</>
)
}
@@ -62,6 +62,7 @@ import AutoCorrectPanel from './AutoCorrectPanel'
import { useAutoCorrect } from '@/hooks/useAutoCorrect'
import { getWordsFromIds } from '@/lib/lyrics-review/utils/wordUtils'
import { applyOffsetToCorrectionData, applyOffsetToSegment } from '@/lib/lyrics-review/utils/timingUtils'
import { VocalsAudioDataLoader } from './VocalsAudioDataLoader'
// Add type for window augmentation
declare global {
@@ -81,6 +82,7 @@ interface ApiClient {
addLyrics: (source: string, lyrics: string) => Promise<CorrectionData>
searchLyrics?: (artist: string, title: string, forceSources?: string[]) => Promise<SearchLyricsResponse>
getAudioUrl: (hash: string) => string
getVocalsAudioUrl: () => string
generatePreviewVideo: (data: CorrectionData, isDuet?: boolean) => Promise<{
status: string
message?: string
@@ -154,6 +156,8 @@ export default function LyricsAnalyzer({
const [reviewMode, setReviewMode] = useState(false)
const [isSubmitting, setIsSubmitting] = useState(false)
const [vocalsAudioUrl, setVocalsAudioUrl] = useState<string | null>(null)
// Success screen state (for auto-close after submission)
const [showSuccess, setShowSuccess] = useState(false)
const [countdown, setCountdown] = useState(2)
@@ -570,6 +574,12 @@ export default function LyricsAnalyzer({
}
}, [showInstrumentalReview, isLocalMode])
useEffect(() => {
if (!apiClient) return
setVocalsAudioUrl(apiClient.getVocalsAudioUrl())
}, [apiClient])
// Countdown effect for success screen (auto-close/redirect after submission)
useEffect(() => {
if (!showSuccess) return
@@ -1352,330 +1362,332 @@ export default function LyricsAnalyzer({
}
return (
<div className="max-w-full overflow-x-hidden">
<Header
isReadOnly={isReadOnly}
onFileLoad={onFileLoad}
data={data}
onMetricClick={metricClickHandlers}
effectiveMode={effectiveMode}
onModeChange={setInteractionMode}
apiClient={apiClient}
audioHash={audioHash}
onTimeUpdate={setCurrentAudioTime}
onHandlerToggle={handleHandlerToggle}
isUpdatingHandlers={isUpdatingHandlers}
onHandlerClick={handleHandlerClick}
onFindReplace={() => setIsFindReplaceModalOpen(true)}
onEditAll={handleReplaceAllLyrics}
onTimingOffset={handleOpenTimingOffsetModal}
timingOffsetMs={timingOffsetMs}
onUndo={handleUndo}
onRedo={handleRedo}
canUndo={canUndo}
canRedo={canRedo}
onResetCorrections={handleResetCorrections}
statsVisible={statsVisible}
onStatsToggle={() => setStatsVisible((v) => !v)}
reviewMode={reviewMode}
onReviewModeToggle={setReviewMode}
onAcceptAllCorrections={handleAcceptAllCorrections}
onAcceptHighConfidenceCorrections={handleAcceptHighConfidenceCorrections}
onRevertAllCorrections={handleRevertAllCorrections}
isDuet={isDuet}
onToggleDuet={() => setIsDuet(d => !d)}
onAutoCorrect={
!isLocalMode && jobId
? () =>
autoCorrect.acceptedCount > 0
? setAutoCorrectDetailsOpen((o) => !o)
: setIsAutoCorrectModalOpen(true)
: undefined
}
autoCorrectedCount={autoCorrect.acceptedCount}
/>
<AutoCorrectPanel
controller={autoCorrect}
isReadOnly={isReadOnly}
open={autoCorrectDetailsOpen}
onClose={() => setAutoCorrectDetailsOpen(false)}
onRerun={() => {
setAutoCorrectDetailsOpen(false)
setIsAutoCorrectModalOpen(true)
}}
/>
<div className={cn('grid gap-2', isMobile ? 'grid-cols-1' : 'grid-cols-2')}>
<TranscriptionView
data={displayData}
mode={effectiveMode}
onElementClick={setModalContent}
onWordClick={handleWordClick}
flashingType={flashingType}
flashingHandler={flashingHandler}
highlightInfo={highlightInfo}
onPlaySegment={handlePlaySegment}
currentTime={isAnyModalOpen ? undefined : currentAudioTime}
anchors={data.anchor_sequences}
onDataChange={(updatedData) => {
updateDataWithHistory(updatedData, 'direct data change')
}}
<VocalsAudioDataLoader audioUrl={vocalsAudioUrl}>
<div className="max-w-full overflow-x-hidden">
<Header
isReadOnly={isReadOnly}
onFileLoad={onFileLoad}
data={data}
onMetricClick={metricClickHandlers}
effectiveMode={effectiveMode}
onModeChange={setInteractionMode}
apiClient={apiClient}
audioHash={audioHash}
onTimeUpdate={setCurrentAudioTime}
onHandlerToggle={handleHandlerToggle}
isUpdatingHandlers={isUpdatingHandlers}
onHandlerClick={handleHandlerClick}
onFindReplace={() => setIsFindReplaceModalOpen(true)}
onEditAll={handleReplaceAllLyrics}
onTimingOffset={handleOpenTimingOffsetModal}
timingOffsetMs={timingOffsetMs}
onUndo={handleUndo}
onRedo={handleRedo}
canUndo={canUndo}
canRedo={canRedo}
onResetCorrections={handleResetCorrections}
statsVisible={statsVisible}
onStatsToggle={() => setStatsVisible((v) => !v)}
reviewMode={reviewMode}
onRevertCorrection={handleRevertCorrection}
onEditCorrection={handleEditCorrection}
onAcceptCorrection={handleAcceptCorrection}
onShowCorrectionDetail={handleShowCorrectionDetail}
activeGapWordIds={activeGapWordIds}
advancedMode={advancedMode}
onAdvancedModeToggle={handleAdvancedModeToggle}
editedWordIds={editedWordIds}
aiCorrectedWordIds={aiCorrectedWordIds}
aiOriginalTextByWordId={aiOriginalTextByWordId}
aiEstimatedWordIds={aiEstimatedWordIds}
onReviewModeToggle={setReviewMode}
onAcceptAllCorrections={handleAcceptAllCorrections}
onAcceptHighConfidenceCorrections={handleAcceptHighConfidenceCorrections}
onRevertAllCorrections={handleRevertAllCorrections}
isDuet={isDuet}
onSegmentSingerChange={(segmentIdx, next) => {
const segments = [...data.corrected_segments]
segments[segmentIdx] = { ...segments[segmentIdx], singer: next }
updateDataWithHistory({ ...data, corrected_segments: segments }, 'singer change')
}}
onSegmentFocus={setFocusedSegmentIndex}
onToggleDuet={() => setIsDuet(d => !d)}
onAutoCorrect={
!isLocalMode && jobId
? () =>
autoCorrect.acceptedCount > 0
? setAutoCorrectDetailsOpen((o) => !o)
: setIsAutoCorrectModalOpen(true)
: undefined
}
autoCorrectedCount={autoCorrect.acceptedCount}
/>
<ReferenceView
referenceSources={data.reference_lyrics}
anchors={data.anchor_sequences}
gaps={data.gap_sequences}
mode={effectiveMode}
onElementClick={setModalContent}
onWordClick={handleWordClick}
flashingType={flashingType}
highlightInfo={highlightInfo}
currentSource={currentSource}
onSourceChange={setCurrentSource}
corrected_segments={data.corrected_segments}
corrections={data.corrections}
onAddLyrics={() => setIsAddLyricsModalOpen(true)}
onAddLyricsInline={handleAddLyrics}
onSearchLyrics={handleSearchLyrics}
<AutoCorrectPanel
controller={autoCorrect}
isReadOnly={isReadOnly}
open={autoCorrectDetailsOpen}
onClose={() => setAutoCorrectDetailsOpen(false)}
onRerun={() => {
setAutoCorrectDetailsOpen(false)
setIsAutoCorrectModalOpen(true)
}}
/>
<div className={cn('grid gap-2', isMobile ? 'grid-cols-1' : 'grid-cols-2')}>
<TranscriptionView
data={displayData}
mode={effectiveMode}
onElementClick={setModalContent}
onWordClick={handleWordClick}
flashingType={flashingType}
flashingHandler={flashingHandler}
highlightInfo={highlightInfo}
onPlaySegment={handlePlaySegment}
currentTime={isAnyModalOpen ? undefined : currentAudioTime}
anchors={data.anchor_sequences}
onDataChange={(updatedData) => {
updateDataWithHistory(updatedData, 'direct data change')
}}
reviewMode={reviewMode}
onRevertCorrection={handleRevertCorrection}
onEditCorrection={handleEditCorrection}
onAcceptCorrection={handleAcceptCorrection}
onShowCorrectionDetail={handleShowCorrectionDetail}
activeGapWordIds={activeGapWordIds}
advancedMode={advancedMode}
onAdvancedModeToggle={handleAdvancedModeToggle}
editedWordIds={editedWordIds}
aiCorrectedWordIds={aiCorrectedWordIds}
aiOriginalTextByWordId={aiOriginalTextByWordId}
aiEstimatedWordIds={aiEstimatedWordIds}
isDuet={isDuet}
onSegmentSingerChange={(segmentIdx, next) => {
const segments = [...data.corrected_segments]
segments[segmentIdx] = { ...segments[segmentIdx], singer: next }
updateDataWithHistory({ ...data, corrected_segments: segments }, 'singer change')
}}
onSegmentFocus={setFocusedSegmentIndex}
/>
<ReferenceView
referenceSources={data.reference_lyrics}
anchors={data.anchor_sequences}
gaps={data.gap_sequences}
mode={effectiveMode}
onElementClick={setModalContent}
onWordClick={handleWordClick}
flashingType={flashingType}
highlightInfo={highlightInfo}
currentSource={currentSource}
onSourceChange={setCurrentSource}
corrected_segments={data.corrected_segments}
corrections={data.corrections}
onAddLyrics={() => setIsAddLyricsModalOpen(true)}
onAddLyricsInline={handleAddLyrics}
onSearchLyrics={handleSearchLyrics}
defaultArtist={data.metadata?.artist || ''}
defaultTitle={data.metadata?.title || ''}
/>
</div>
{/* Spacer for sticky footer */}
{!isReadOnly && apiClient && <div className="h-16" />}
{/* Sticky footer bar */}
{!isReadOnly && apiClient && (
<div className="fixed bottom-0 left-0 right-0 bg-background border-t shadow-lg py-3 px-4 z-50 flex justify-center items-center gap-4 flex-wrap">
<GapNavigator
currentGapIndex={currentGapIndex}
totalGaps={uncorrectedGaps.length}
onPrevGap={handlePrevGap}
onNextGap={handleNextGap}
/>
{data.gap_sequences.length > 0 && (
<div className="flex items-center gap-1.5">
<div className="w-16 h-1.5 bg-muted rounded-full overflow-hidden">
<div
className="h-full bg-green-500 rounded-full transition-all"
style={{
width: `${data.gap_sequences.length > 0 ? ((data.gap_sequences.length - uncorrectedGaps.length) / data.gap_sequences.length) * 100 : 0}%`,
}}
/>
</div>
{uncorrectedGaps.length === 0 ? (
<span className="text-xs text-green-500 flex items-center gap-0.5">
<CheckCircle2 className="h-3.5 w-3.5" />
All gaps reviewed
</span>
) : (
<span className="text-xs text-muted-foreground whitespace-nowrap">
{data.gap_sequences.length - uncorrectedGaps.length}/{data.gap_sequences.length} gaps reviewed
</span>
)}
</div>
)}
{apiClient?.saveReviewSession && (
<>
<Button
variant="outline"
size="sm"
onClick={() => saveSession('manual')}
disabled={history.length <= 1}
>
Save Progress
</Button>
<Button
variant="outline"
size="sm"
onClick={handleOpenSessionBrowser}
>
Session History
</Button>
</>
)}
<span className="text-sm text-muted-foreground">Lyrics look good?</span>
<Button onClick={handleFinishReview} disabled={isReviewComplete}>
{isReviewComplete ? 'Review Complete' : 'Preview Video'}
<Video className="ml-2 h-4 w-4" />
</Button>
</div>
)}
{/* Modals */}
<EditModal
open={Boolean(editModalSegment)}
onClose={() => {
setEditModalSegment(null)
handleSetModalSpacebarHandler(undefined)
}}
segment={
editModalSegment?.segment
? timingOffsetMs !== 0
? applyOffsetToSegment(editModalSegment.segment, timingOffsetMs)
: editModalSegment.segment
: null
}
segmentIndex={editModalSegment?.index ?? null}
originalSegment={
editModalSegment?.originalSegment
? timingOffsetMs !== 0
? applyOffsetToSegment(editModalSegment.originalSegment, timingOffsetMs)
: editModalSegment.originalSegment
: null
}
onSave={handleUpdateSegment}
onDelete={handleDeleteSegment}
onAddSegment={handleAddSegment}
onSplitSegment={handleSplitSegment}
onMergeSegment={handleMergeSegment}
onPlaySegment={handlePlaySegment}
currentTime={currentAudioTime}
originalTranscribedSegment={
editModalSegment?.segment &&
editModalSegment?.index !== null &&
originalData.original_segments
? (() => {
const origSegment =
originalData.original_segments.find(
(s: LyricsSegment) => s.id === editModalSegment.segment.id
) || null
return origSegment && timingOffsetMs !== 0
? applyOffsetToSegment(origSegment, timingOffsetMs)
: origSegment
})()
: null
}
/>
<ReviewChangesModal
open={isReviewModalOpen}
onClose={() => setIsReviewModalOpen(false)}
data={data}
onSubmit={handleSubmitToServer}
isSubmitting={isSubmitting}
apiClient={apiClient}
timingOffsetMs={timingOffsetMs}
isDuet={isDuet}
/>
<AddLyricsModal
open={isAddLyricsModalOpen}
onClose={() => setIsAddLyricsModalOpen(false)}
onAdd={handleAddLyrics}
onSearch={handleSearchLyrics}
defaultArtist={data.metadata?.artist || ''}
defaultTitle={data.metadata?.title || ''}
/>
</div>
{/* Spacer for sticky footer */}
{!isReadOnly && apiClient && <div className="h-16" />}
<FindReplaceModal
open={isFindReplaceModalOpen}
onClose={() => setIsFindReplaceModalOpen(false)}
onReplace={handleFindReplace}
data={data}
/>
{/* Sticky footer bar */}
{!isReadOnly && apiClient && (
<div className="fixed bottom-0 left-0 right-0 bg-background border-t shadow-lg py-3 px-4 z-50 flex justify-center items-center gap-4 flex-wrap">
<GapNavigator
currentGapIndex={currentGapIndex}
totalGaps={uncorrectedGaps.length}
onPrevGap={handlePrevGap}
onNextGap={handleNextGap}
<TimingOffsetModal
open={isTimingOffsetModalOpen}
onClose={() => setIsTimingOffsetModalOpen(false)}
currentOffset={timingOffsetMs}
onApply={handleApplyTimingOffset}
/>
<ReplaceAllLyricsModal
open={isReplaceAllLyricsModalOpen}
onClose={() => setIsReplaceAllLyricsModalOpen(false)}
onSave={handleSaveReplaceAllLyrics}
onPlaySegment={handlePlaySegment}
currentTime={currentAudioTime}
setModalSpacebarHandler={handleSetModalSpacebarHandler}
existingSegments={data.corrected_segments}
jobId={jobId ?? ''}
artist={data.metadata?.artist}
title={data.metadata?.title}
authToken={getAccessToken() ?? undefined}
/>
<AutoCorrectModal
open={isAutoCorrectModalOpen}
onClose={() => setIsAutoCorrectModalOpen(false)}
onRun={(settings) => void autoCorrect.run(settings)}
onCancelRun={autoCorrect.cancel}
status={autoCorrect.status}
error={autoCorrect.error}
hasReferences={autoCorrect.hasReferences}
/>
<EditFeedbackBar
entry={lastEditEntry}
visible={showFeedbackBar}
onFeedback={handleEditFeedback}
onDismiss={() => setShowFeedbackBar(false)}
/>
{selectedCorrection && (
<CorrectionDetailCard
open={correctionDetailOpen}
onClose={() => {
setCorrectionDetailOpen(false)
setSelectedCorrection(null)
}}
originalWord={selectedCorrection.originalWord}
correctedWord={selectedCorrection.correctedWord}
category={selectedCorrection.category}
confidence={selectedCorrection.confidence}
reason={selectedCorrection.reason}
handler={selectedCorrection.handler}
source={selectedCorrection.source}
onRevert={() => {
handleRevertCorrection(selectedCorrection.wordId)
setCorrectionDetailOpen(false)
setSelectedCorrection(null)
}}
onEdit={() => {
handleEditCorrection(selectedCorrection.wordId)
setCorrectionDetailOpen(false)
setSelectedCorrection(null)
}}
onAccept={() => {
handleAcceptCorrection(selectedCorrection.wordId)
setCorrectionDetailOpen(false)
setSelectedCorrection(null)
}}
/>
{data.gap_sequences.length > 0 && (
<div className="flex items-center gap-1.5">
<div className="w-16 h-1.5 bg-muted rounded-full overflow-hidden">
<div
className="h-full bg-green-500 rounded-full transition-all"
style={{
width: `${data.gap_sequences.length > 0 ? ((data.gap_sequences.length - uncorrectedGaps.length) / data.gap_sequences.length) * 100 : 0}%`,
}}
/>
</div>
{uncorrectedGaps.length === 0 ? (
<span className="text-xs text-green-500 flex items-center gap-0.5">
<CheckCircle2 className="h-3.5 w-3.5" />
All gaps reviewed
</span>
) : (
<span className="text-xs text-muted-foreground whitespace-nowrap">
{data.gap_sequences.length - uncorrectedGaps.length}/{data.gap_sequences.length} gaps reviewed
</span>
)}
</div>
)}
{apiClient?.saveReviewSession && (
<>
<Button
variant="outline"
size="sm"
onClick={() => saveSession('manual')}
disabled={history.length <= 1}
>
Save Progress
</Button>
<Button
variant="outline"
size="sm"
onClick={handleOpenSessionBrowser}
>
Session History
</Button>
</>
)}
<span className="text-sm text-muted-foreground">Lyrics look good?</span>
<Button onClick={handleFinishReview} disabled={isReviewComplete}>
{isReviewComplete ? 'Review Complete' : 'Preview Video'}
<Video className="ml-2 h-4 w-4" />
</Button>
</div>
)}
)}
{/* Modals */}
<EditModal
open={Boolean(editModalSegment)}
onClose={() => {
setEditModalSegment(null)
handleSetModalSpacebarHandler(undefined)
}}
segment={
editModalSegment?.segment
? timingOffsetMs !== 0
? applyOffsetToSegment(editModalSegment.segment, timingOffsetMs)
: editModalSegment.segment
: null
}
segmentIndex={editModalSegment?.index ?? null}
originalSegment={
editModalSegment?.originalSegment
? timingOffsetMs !== 0
? applyOffsetToSegment(editModalSegment.originalSegment, timingOffsetMs)
: editModalSegment.originalSegment
: null
}
onSave={handleUpdateSegment}
onDelete={handleDeleteSegment}
onAddSegment={handleAddSegment}
onSplitSegment={handleSplitSegment}
onMergeSegment={handleMergeSegment}
onPlaySegment={handlePlaySegment}
currentTime={currentAudioTime}
originalTranscribedSegment={
editModalSegment?.segment &&
editModalSegment?.index !== null &&
originalData.original_segments
? (() => {
const origSegment =
originalData.original_segments.find(
(s: LyricsSegment) => s.id === editModalSegment.segment.id
) || null
return origSegment && timingOffsetMs !== 0
? applyOffsetToSegment(origSegment, timingOffsetMs)
: origSegment
})()
: null
}
/>
<ReviewChangesModal
open={isReviewModalOpen}
onClose={() => setIsReviewModalOpen(false)}
data={data}
onSubmit={handleSubmitToServer}
isSubmitting={isSubmitting}
apiClient={apiClient}
timingOffsetMs={timingOffsetMs}
isDuet={isDuet}
/>
<AddLyricsModal
open={isAddLyricsModalOpen}
onClose={() => setIsAddLyricsModalOpen(false)}
onAdd={handleAddLyrics}
onSearch={handleSearchLyrics}
defaultArtist={data.metadata?.artist || ''}
defaultTitle={data.metadata?.title || ''}
/>
<FindReplaceModal
open={isFindReplaceModalOpen}
onClose={() => setIsFindReplaceModalOpen(false)}
onReplace={handleFindReplace}
data={data}
/>
<TimingOffsetModal
open={isTimingOffsetModalOpen}
onClose={() => setIsTimingOffsetModalOpen(false)}
currentOffset={timingOffsetMs}
onApply={handleApplyTimingOffset}
/>
<ReplaceAllLyricsModal
open={isReplaceAllLyricsModalOpen}
onClose={() => setIsReplaceAllLyricsModalOpen(false)}
onSave={handleSaveReplaceAllLyrics}
onPlaySegment={handlePlaySegment}
currentTime={currentAudioTime}
setModalSpacebarHandler={handleSetModalSpacebarHandler}
existingSegments={data.corrected_segments}
jobId={jobId ?? ''}
artist={data.metadata?.artist}
title={data.metadata?.title}
authToken={getAccessToken() ?? undefined}
/>
<AutoCorrectModal
open={isAutoCorrectModalOpen}
onClose={() => setIsAutoCorrectModalOpen(false)}
onRun={(settings) => void autoCorrect.run(settings)}
onCancelRun={autoCorrect.cancel}
status={autoCorrect.status}
error={autoCorrect.error}
hasReferences={autoCorrect.hasReferences}
/>
<EditFeedbackBar
entry={lastEditEntry}
visible={showFeedbackBar}
onFeedback={handleEditFeedback}
onDismiss={() => setShowFeedbackBar(false)}
/>
{selectedCorrection && (
<CorrectionDetailCard
open={correctionDetailOpen}
onClose={() => {
setCorrectionDetailOpen(false)
setSelectedCorrection(null)
}}
originalWord={selectedCorrection.originalWord}
correctedWord={selectedCorrection.correctedWord}
category={selectedCorrection.category}
confidence={selectedCorrection.confidence}
reason={selectedCorrection.reason}
handler={selectedCorrection.handler}
source={selectedCorrection.source}
onRevert={() => {
handleRevertCorrection(selectedCorrection.wordId)
setCorrectionDetailOpen(false)
setSelectedCorrection(null)
}}
onEdit={() => {
handleEditCorrection(selectedCorrection.wordId)
setCorrectionDetailOpen(false)
setSelectedCorrection(null)
}}
onAccept={() => {
handleAcceptCorrection(selectedCorrection.wordId)
setCorrectionDetailOpen(false)
setSelectedCorrection(null)
}}
/>
)}
{!isReadOnly && (
<SessionRestoreDialog
open={sessionRestoreOpen}
onClose={() => setSessionRestoreOpen(false)}
onRestore={handleSessionRestore}
sessions={savedSessions}
apiClient={sessionClient as LyricsReviewApiClient}
isLoading={sessionsLoading}
/>
)}
</div>
{!isReadOnly && (
<SessionRestoreDialog
open={sessionRestoreOpen}
onClose={() => setSessionRestoreOpen(false)}
onRestore={handleSessionRestore}
sessions={savedSessions}
apiClient={sessionClient as LyricsReviewApiClient}
isLoading={sessionsLoading}
/>
)}
</div>
</VocalsAudioDataLoader>
)
}
@@ -3,6 +3,8 @@
import { useRef, useState } from 'react'
import { Word } from '@/lib/lyrics-review/types'
import { cn } from '@/lib/utils'
import { WaveformVisualizer } from './WaveformVisualizer'
import { VocalsAudioDataLoaderContext } from './VocalsAudioDataLoader'
interface TimelineEditorProps {
words: Word[]
@@ -241,14 +243,14 @@ export default function TimelineEditor({
return (
<div
ref={containerRef}
className="relative h-[75px] bg-card rounded my-2 px-2 border border-border"
className="relative bg-card rounded border border-border"
onMouseMove={handleMouseMove}
onMouseUp={handleMouseUp}
onMouseLeave={handleMouseUp}
>
{/* Timeline ruler */}
<div
className="absolute top-0 left-0 right-0 h-10 border-b border-border cursor-pointer"
className="h-10 border-b border-border cursor-pointer"
onClick={handleTimelineClick}
>
{generateTimelineMarks()}
@@ -263,52 +265,65 @@ export default function TimelineEditor({
)}
{/* Word blocks */}
{words.map((word, index) => {
if (word.start_time === null || word.end_time === null) return null
<div className="relative h-[30px]">
{words.map((word, index) => {
if (word.start_time === null || word.end_time === null) return null
const leftPosition = timeToPosition(word.start_time)
const rightPosition = timeToPosition(word.end_time)
const width = rightPosition - leftPosition
const leftPosition = timeToPosition(word.start_time)
const rightPosition = timeToPosition(word.end_time)
const width = rightPosition - leftPosition
return (
<div
key={index}
className={cn(
'absolute h-[30px] top-10 bg-primary rounded text-primary-foreground px-2 py-1',
'cursor-move select-none flex items-center text-sm font-sans transition-colors',
isWordHighlighted(word) && 'bg-purple-500 dark:bg-purple-600'
)}
style={{
left: `${leftPosition}%`,
width: `${width}%`,
maxWidth: `calc(${100 - leftPosition}%)`,
}}
onMouseDown={(e) => {
e.stopPropagation()
handleMouseDown(e, index, 'move')
}}
onContextMenu={(e) => handleContextMenu(e, index)}
>
{/* Left resize handle */}
return (
<div
className="absolute top-0 left-0 w-2.5 h-full cursor-col-resize hover:bg-primary-foreground/20 rounded-l"
key={index}
className={cn(
'absolute bg-primary rounded text-primary-foreground px-2 py-1',
'cursor-move select-none flex items-center text-sm font-sans transition-colors',
isWordHighlighted(word) && 'bg-purple-500 dark:bg-purple-600'
)}
style={{
left: `${leftPosition}%`,
width: `${width}%`,
maxWidth: `calc(${100 - leftPosition}%)`,
}}
onMouseDown={(e) => {
e.stopPropagation()
handleMouseDown(e, index, 'resize-left')
handleMouseDown(e, index, 'move')
}}
/>
{word.text}
{/* Right resize handle */}
<div
className="absolute top-0 right-0 w-2.5 h-full cursor-col-resize hover:bg-primary-foreground/20 rounded-r"
onMouseDown={(e) => {
e.stopPropagation()
handleMouseDown(e, index, 'resize-right')
}}
/>
</div>
)
})}
onContextMenu={(e) => handleContextMenu(e, index)}
>
{/* Left resize handle */}
<div
className="absolute top-0 left-0 w-2.5 h-full cursor-col-resize hover:bg-primary-foreground/20 rounded-l"
onMouseDown={(e) => {
e.stopPropagation()
handleMouseDown(e, index, 'resize-left')
}}
/>
{word.text}
{/* Right resize handle */}
<div
className="absolute top-0 right-0 w-2.5 h-full cursor-col-resize hover:bg-primary-foreground/20 rounded-r"
onMouseDown={(e) => {
e.stopPropagation()
handleMouseDown(e, index, 'resize-right')
}}
/>
</div>
)
})}
</div>
<VocalsAudioDataLoaderContext.Consumer>
{({ audioData: vocalsAudioData }) => (
vocalsAudioData && <WaveformVisualizer
startTime={startTime}
endTime={endTime}
audioData={vocalsAudioData}
className="w-[100%] h-[35px]"
/>
)}
</VocalsAudioDataLoaderContext.Consumer>
</div>
)
}
@@ -0,0 +1,43 @@
import { AudioData, fetchAudioData } from '@/lib/audio-data'
import { createContext, PropsWithChildren, useEffect, useState } from 'react'
export const VocalsAudioDataLoaderContext = createContext<{ audioData: AudioData | null }>({ audioData: null })
export interface AudioDataLoaderProps extends PropsWithChildren {
audioUrl: string | null
}
export const VocalsAudioDataLoader = ({ audioUrl, children }: AudioDataLoaderProps) => {
const [audioData, setAudioData] = useState<AudioData | null>(null)
useEffect(() => {
if (!audioUrl) return
// Guard against (a) an unhandled rejection when the endpoint 404s (no vocal
// stem for this job) and (b) a stale in-flight fetch resolving after a newer
// audioUrl has already been set.
let cancelled = false
fetchAudioData(audioUrl)
.then((audioData) => {
if (!cancelled) setAudioData(audioData)
})
.catch((error) => {
if (!cancelled) {
console.error('Failed to load vocals audio data', error)
setAudioData(null)
}
})
return () => {
cancelled = true
setAudioData(null)
}
}, [audioUrl])
return (
<VocalsAudioDataLoaderContext.Provider value={{ audioData }}>
{children}
</VocalsAudioDataLoaderContext.Provider>
)
}
@@ -0,0 +1,113 @@
import { HTMLProps, useCallback, useEffect, useRef, useState } from 'react'
import colors from 'tailwindcss/colors'
import { AudioData } from '@/lib/audio-data'
import { useAudioReady } from '@/lib/lyrics-review/hooks/useAudioReady'
export interface WaveformVisualizerProps extends Omit<HTMLProps<HTMLCanvasElement>, 'width' | 'height'> {
startTime: number
endTime: number
audioData: AudioData | null
barColor?: string
}
export const WaveformVisualizer = ({
startTime,
endTime,
audioData,
className,
barColor = colors.yellow['500'],
... canvasProps
}: WaveformVisualizerProps) => {
const containerRef = useRef<HTMLDivElement | null>(null)
const canvasCtxRef = useRef<CanvasRenderingContext2D | null>(null)
const audioReady = useAudioReady()
const [canvasSize, setCanvasSize] = useState<{ width: number; height: number; }>({ width: 0, height: 0 })
useEffect(() => {
if (!containerRef.current) return
const resizeObserver = new ResizeObserver((entries) => {
const size = entries.at(-1)!.contentBoxSize[0]
setCanvasSize({
width: size.inlineSize,
height: size.blockSize
})
})
resizeObserver.observe(containerRef.current, {
box: 'content-box'
})
return () => {
resizeObserver.disconnect()
}
}, [containerRef.current])
const readAudioData = useCallback((startTime: number, endTime: number) => {
if (audioData == null) return null
// Index directly into the time-based peak envelope (peaksPerSecond buckets
// per second) so we only ever touch the window being drawn, never the whole
// track.
const startIndex = Math.max(0, Math.floor(startTime * audioData.peaksPerSecond))
const endIndex = Math.min(audioData.peaks.length, Math.ceil(endTime * audioData.peaksPerSecond))
return audioData.peaks.slice(startIndex, endIndex)
}, [audioData])
const render = useCallback((startTime: number, endTime: number) => {
const ctx = canvasCtxRef.current
if (!ctx) return
ctx.clearRect(0, 0, ctx.canvas.width, ctx.canvas.height)
ctx.fillStyle = barColor
const audioData = readAudioData(startTime, endTime)
if (!audioData) return
for (let x = 0; x < ctx.canvas.width; x++) {
// Clamp the first pixel: (x - 0.5) is negative at x=0, and a negative
// slice start would sample peaks from the end of the window instead.
const dataPointIndex = Math.max(0, Math.floor((x - 0.5) / ctx.canvas.width * audioData.length))
// When fewer peaks than pixels are visible (zoomed in), adjacent columns
// floor to the same index; force at least one peak per column so the
// waveform doesn't render zero-height gaps.
const nextDataPointIndex = Math.max(
dataPointIndex + 1,
Math.floor((x + 0.5) / ctx.canvas.width * audioData.length)
)
const amplitude = audioData
.slice(dataPointIndex, nextDataPointIndex)
.reduce((maxAmplitude, point) => Math.max(maxAmplitude, Math.abs(point)), 0)
const barHeight = amplitude * ctx.canvas.height
ctx.fillRect(x, (ctx.canvas.height - barHeight) / 2, 1, barHeight)
}
}, [readAudioData, barColor])
// Re-render when the data (readAudioData is memoized on audioData), the window,
// the canvas size, or the colour changes. audioData arrives asynchronously, so
// omitting `render` here would leave the canvas blank until an unrelated prop
// changed.
useEffect(() => {
render(startTime, endTime)
}, [render, startTime, endTime, canvasSize, audioReady.ready])
return (
<div
ref={containerRef}
className={className}
>
<canvas
ref={(el) => {
canvasCtxRef.current = el?.getContext('2d') ?? null
}}
className={"w-[100%] h-[100%]"}
width={canvasSize.width}
height={canvasSize.height}
{... canvasProps}
></canvas>
</div>
)
}
@@ -5,7 +5,8 @@ import { useState, useEffect, useCallback, useRef } from 'react'
import { toast } from 'sonner'
import { Dialog, DialogContent, DialogHeader, DialogTitle, DialogFooter } from '@/components/ui/dialog'
import { Button } from '@/components/ui/button'
import { Loader2, Play, Square, RotateCcw, Trash2, History } from 'lucide-react'
import { Loader2, Play, Square, RotateCcw, Trash2, History, XCircle } from 'lucide-react'
import { cn } from '@/lib/utils'
import { LyricsSegment, Word } from '@/lib/lyrics-review/types'
import { nanoid } from 'nanoid'
import EditWordList from '../EditWordList'
@@ -56,6 +57,7 @@ export default function EditModal({
const t = useTranslations('lyricsReview.modals.editAll')
const tActions = useTranslations('lyricsReview.editActionBar')
const tHeader = useTranslations('lyricsReview.header')
const tTimeline = useTranslations('lyricsReview.editTimeline')
const { ready: audioReady } = useAudioReady()
const [editedSegment, setEditedSegment] = useState<LyricsSegment | null>(segment)
const [timingFixCount, setTimingFixCount] = useState(0)
@@ -383,23 +385,65 @@ export default function EditModal({
return (
<Dialog open={open} onOpenChange={(isOpen) => !isOpen && handleClose()}>
<DialogContent className="max-w-[960px] max-h-[90vh] overflow-hidden flex flex-col" onKeyDown={handleKeyDown}>
<DialogContent className="max-w-[960px] max-h-[90vh] overflow-hidden flex flex-col gap-2" onKeyDown={handleKeyDown}>
<DialogHeader>
<DialogTitle className="flex items-center gap-2">
{isGlobal ? t('title') : t('editSegment', { index: segmentIndex ?? 0 })}
{segment && segment.start_time !== null && onPlaySegment && (
{/* Left: title (with the segment's time range folded into the heading)
+ play. Right: Tap To Sync, moved up here to reclaim vertical space
above the word list. pr-10 keeps it clear of the dialog close button. */}
<div className="flex items-center justify-between gap-4 pr-10">
<DialogTitle className="flex items-center gap-2">
<span>
{isGlobal ? t('title') : t('editSegment', { index: segmentIndex ?? 0 })}
{!isGlobal &&
editedSegment &&
editedSegment.start_time !== null &&
editedSegment.end_time !== null && (
<span className="font-normal text-muted-foreground">
{' '}
({editedSegment.start_time.toFixed(2)} - {editedSegment.end_time.toFixed(2)})
</span>
)}
</span>
{segment && segment.start_time !== null && onPlaySegment && (
<Button
variant="ghost"
size="icon"
className="h-8 w-8"
onClick={handlePlayToggle}
disabled={!audioReady}
title={audioReady ? undefined : tHeader('audioStillLoading')}
>
{isPlaying ? <Square className="h-4 w-4" /> : <Play className="h-4 w-4" />}
</Button>
)}
</DialogTitle>
{/* Tap To Sync toggle (relocated from EditTimelineSection) */}
{editedSegment && editedSegment.words.some((w) => w.start_time !== null) && (
<Button
variant="ghost"
size="icon"
className="h-8 w-8"
onClick={handlePlayToggle}
disabled={!audioReady}
title={audioReady ? undefined : tHeader('audioStillLoading')}
variant={isManualSyncing ? 'outline' : 'default'}
onClick={startManualSync}
className={cn(
'shrink-0',
isManualSyncing
? 'border-destructive text-destructive hover:bg-destructive/10'
: 'bg-primary'
)}
>
{isPlaying ? <Square className="h-4 w-4" /> : <Play className="h-4 w-4" />}
{isManualSyncing ? (
<>
<XCircle className="h-4 w-4 mr-1" />
{tTimeline('cancelSync')}
</>
) : (
<>
<Play className="h-4 w-4 mr-1" />
{tTimeline('tapToSync')}
</>
)}
</Button>
)}
</DialogTitle>
</div>
</DialogHeader>
{isLoading ? (
@@ -410,7 +454,7 @@ export default function EditModal({
</p>
</div>
) : editedSegment ? (
<div className="flex flex-col flex-1 space-y-4 overflow-hidden">
<div className="flex flex-col flex-1 space-y-2 overflow-hidden">
{timingFixCount > 0 && (
<div
role="status"
@@ -429,17 +473,12 @@ export default function EditModal({
words={editedSegment.words}
startTime={startTime}
endTime={endTime}
originalStartTime={originalSegment?.start_time ?? null}
originalEndTime={originalSegment?.end_time ?? null}
currentStartTime={editedSegment.start_time}
currentEndTime={editedSegment.end_time}
currentTime={currentTime}
isManualSyncing={isManualSyncing}
syncWordIndex={syncWordIndex}
isSpacebarPressed={isSpacebarPressed}
onWordUpdate={handleWordChange}
onPlaySegment={onPlaySegment}
startManualSync={startManualSync}
isGlobal={isGlobal}
onTapStart={handleTapStart}
onTapEnd={handleTapEnd}
+10
View File
@@ -3488,6 +3488,7 @@ export interface LyricsReviewApiClient {
addLyrics: (source: string, lyrics: string) => Promise<CorrectionData>
searchLyrics: (artist: string, title: string, forceSources?: string[]) => Promise<SearchLyricsResponse>
getAudioUrl: (hash: string) => string
getVocalsAudioUrl: () => string
generatePreviewVideo: (data: CorrectionData, isDuet?: boolean) => Promise<{
status: string
message?: string
@@ -3627,6 +3628,15 @@ export function createLyricsReviewApiClient(jobId: string): LyricsReviewApiClien
return token ? `${base}?token=${encodeURIComponent(token)}` : base
},
/**
* Get vocals audio URL for playback
*/
getVocalsAudioUrl(): string {
const token = getAccessToken()
const base = `${API_BASE_URL}/api/review/${jobId}/audio/vocals`
return token ? `${base}?token=${encodeURIComponent(token)}` : base
},
/**
* Generate a preview video from the current correction data
*/
+67
View File
@@ -0,0 +1,67 @@
export interface AudioData {
duration: number
// Downsampled absolute-amplitude envelope: one peak per bucket. Storing the
// envelope instead of the raw PCM keeps a whole vocal track in ~hundreds of KB
// (peaksPerSecond * duration * 4 bytes) rather than the tens of MB a full
// Float32Array of decoded samples would occupy, which noticeably slows the app.
peaks: Float32Array
peaksPerSecond: number
}
// Resolution of the precomputed peak envelope. High enough to stay crisp when a
// short segment is zoomed across the full timeline width, low enough to keep the
// whole track tiny in memory. 400/s over a 4-minute song ≈ 96k floats (~384 KB).
const PEAKS_PER_SECOND = 400
export async function fetchAudioData(url: string): Promise<AudioData> {
// fetch() resolves even for 4xx/5xx; the vocals endpoint 404s when a job has
// no vocal stem. Fail explicitly so the caller sees "no vocals" rather than an
// opaque decodeAudioData error on the JSON/HTML error body.
const response = await fetch(url)
if (!response.ok) {
throw new Error(`Failed to fetch vocals audio: ${response.status}`)
}
const arrayBuffer = await response.arrayBuffer()
const context = new AudioContext()
try {
const decoded = await context.decodeAudioData(arrayBuffer)
const { duration, numberOfChannels, length } = decoded
const bucketCount = Math.max(1, Math.ceil(duration * PEAKS_PER_SECOND))
const samplesPerBucket = Math.max(1, Math.floor(length / bucketCount))
const peaks = new Float32Array(bucketCount)
// Collapse all channels into a single envelope by taking the maximum
// absolute sample value in each bucket. Iterate per bucket (not per sample)
// so we avoid a division + Math.floor on every one of the ~millions of
// samples, which would otherwise freeze the main thread while loading.
// The raw decoded buffer is released when this function returns; only the
// compact `peaks` array is retained.
for (let channelIdx = 0; channelIdx < numberOfChannels; channelIdx++) {
const channelData = decoded.getChannelData(channelIdx)
for (let bucket = 0; bucket < bucketCount; bucket++) {
const start = bucket * samplesPerBucket
if (start >= length) break
const end = bucket === bucketCount - 1 ? length : Math.min(length, start + samplesPerBucket)
let peak = peaks[bucket]
for (let i = start; i < end; i++) {
const amplitude = Math.abs(channelData[i])
if (amplitude > peak) {
peak = amplitude
}
}
peaks[bucket] = peak
}
}
return {
duration,
peaks,
peaksPerSecond: PEAKS_PER_SECOND
}
} finally {
context.close()
}
}
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "إعادة التنفيذ",
"revertAllAuto": "التراجع عن الكل",
"showDetails": "عرض التفاصيل"
},
"editTimeline": {
"holdingRelease": "جاري الضغط... أفلت عندما تنتهي الكلمة",
"pressSpacebarHint": "اضغط على زر المسافة عندما تبدأ الكلمة (انقر للكلمات القصيرة، واستمر بالضغط للكلمات الطويلة)",
"tapToSync": "انقر للمزامنة",
"tapWhenStarts": "انقر على الزر عندما تبدأ الكلمة",
"wordProgress": "الكلمة {index} من {total}:",
"cancelSync": "إلغاء",
"hold": "اضغط مطولاً...",
"tap": "انقر"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Torna a executar",
"revertAllAuto": "Reverteix-ho tot",
"showDetails": "Mostra els detalls"
},
"editTimeline": {
"holdingRelease": "Mantenint premut... Deixa anar quan acabi la paraula",
"pressSpacebarHint": "Prem la barra espaiadora quan comenci la paraula (toca per a paraules curtes, mantén premut per a paraules llargues)",
"tapToSync": "Toca per sincronitzar",
"tapWhenStarts": "Toca el botó quan comenci la paraula",
"wordProgress": "Paraula {index} de {total}:",
"cancelSync": "Cancel·la",
"hold": "MANTÉN...",
"tap": "TOCA"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Spustit znovu",
"revertAllAuto": "Vrátit vše",
"showDetails": "Zobrazit podrobnosti"
},
"editTimeline": {
"holdingRelease": "Držíš... Pusť, když slovo skončí",
"pressSpacebarHint": "Stiskni mezerník, když slovo začne (ťukni pro krátká slova, podrž pro dlouhá slova)",
"tapToSync": "Ťukni pro synchronizaci",
"tapWhenStarts": "Ťukni na tlačítko, když slovo začne",
"wordProgress": "Slovo {index} z {total}:",
"cancelSync": "Zrušit",
"hold": "DRŽ...",
"tap": "KLEPNI"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Kør igen",
"revertAllAuto": "Fortryd alle",
"showDetails": "Vis detaljer"
},
"editTimeline": {
"holdingRelease": "Holder nede... Slip, når ordet slutter",
"pressSpacebarHint": "Tryk på mellemrumstasten, når ordet starter (tryk for korte ord, hold nede for lange ord)",
"tapToSync": "Tryk for at synkronisere",
"tapWhenStarts": "Tryk på knappen, når ordet starter",
"wordProgress": "Ord {index} af {total}:",
"cancelSync": "Annuller",
"hold": "HOLD...",
"tap": "TRYK"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Erneut ausführen",
"revertAllAuto": "Alle rückgängig machen",
"showDetails": "Details anzeigen"
},
"editTimeline": {
"holdingRelease": "Gedrückt halten... Loslassen, wenn das Wort endet",
"pressSpacebarHint": "Drücke die Leertaste, wenn das Wort beginnt (tippen für kurze Wörter, gedrückt halten für lange Wörter)",
"tapToSync": "Tippen zum Synchronisieren",
"tapWhenStarts": "Tippe auf den Button, wenn das Wort beginnt",
"wordProgress": "Wort {index} von {total}:",
"cancelSync": "Abbrechen",
"hold": "HALTEN...",
"tap": "TIPPEN"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Επανεκτέλεση",
"revertAllAuto": "Αναίρεση όλων",
"showDetails": "Εμφάνιση λεπτομερειών"
},
"editTimeline": {
"holdingRelease": "Κρατάς πατημένο... Άφησέ το όταν τελειώσει η λέξη",
"pressSpacebarHint": "Πάτησε το spacebar όταν ξεκινάει η λέξη (πάτησε στιγμιαία για μικρές λέξεις, κράτα πατημένο για μεγάλες λέξεις)",
"tapToSync": "Πάτησε για συγχρονισμό",
"tapWhenStarts": "Πάτησε το κουμπί όταν ξεκινάει η λέξη",
"wordProgress": "Λέξη {index} από {total}:",
"cancelSync": "Ακύρωση",
"hold": "ΚΡΑΤΑ...",
"tap": "ΠΑΤΑ"
}
},
"instrumentalReview": {
+10
View File
@@ -1337,6 +1337,16 @@
},
"upcomingMore": "+{count} more"
},
"editTimeline": {
"tapToSync": "Tap To Sync",
"cancelSync": "Cancel",
"wordProgress": "Word {index} of {total}:",
"holdingRelease": "Holding... Release when word ends",
"tapWhenStarts": "Tap the button when word starts",
"pressSpacebarHint": "Press spacebar when word starts (tap for short words, hold for long words)",
"tap": "TAP",
"hold": "HOLD..."
},
"duet": {
"markAsDuet": "Duet Mode",
"duetOn": "Duet: ON",
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Volver a ejecutar",
"revertAllAuto": "Revertir todo",
"showDetails": "Mostrar detalles"
},
"editTimeline": {
"holdingRelease": "Manteniendo presionado... Suelta cuando termine la palabra",
"pressSpacebarHint": "Presiona la barra espaciadora cuando empiece la palabra (toca para palabras cortas, mantén presionado para palabras largas)",
"tapToSync": "Toca para sincronizar",
"tapWhenStarts": "Toca el botón cuando empiece la palabra",
"wordProgress": "Palabra {index} de {total}:",
"cancelSync": "Cancelar",
"hold": "MANTÉN...",
"tap": "TOCA"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Suorita uudelleen",
"revertAllAuto": "Kumoa kaikki",
"showDetails": "Näytä tiedot"
},
"editTimeline": {
"holdingRelease": "Pidä painettuna... Vapauta, kun sana päättyy",
"pressSpacebarHint": "Paina välilyöntiä, kun sana alkaa (napauta lyhyiden sanojen kohdalla, pidä painettuna pitkien kohdalla)",
"tapToSync": "Napauta synkronoidaksesi",
"tapWhenStarts": "Napauta painiketta, kun sana alkaa",
"wordProgress": "Sana {index}/{total}:",
"cancelSync": "Peruuta",
"hold": "PIDÄ...",
"tap": "NAPAUTA"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Relancer",
"revertAllAuto": "Tout rétablir",
"showDetails": "Afficher les détails"
},
"editTimeline": {
"holdingRelease": "Maintiens... Relâche quand le mot se termine",
"pressSpacebarHint": "Appuie sur la barre d'espace quand le mot commence (appuie brièvement pour les mots courts, maintiens pour les mots longs)",
"tapToSync": "Appuie pour synchroniser",
"tapWhenStarts": "Appuie sur le bouton quand le mot commence",
"wordProgress": "Mot {index} sur {total} :",
"cancelSync": "Annuler",
"hold": "MAINTIENS...",
"tap": "APPUIE"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "הפעל מחדש",
"revertAllAuto": "בטל הכל",
"showDetails": "הצג פרטים"
},
"editTimeline": {
"holdingRelease": "לחוץ... שחרר כשהמילה מסתיימת",
"pressSpacebarHint": "לחץ על מקש הרווח כשהמילה מתחילה (הקש למילים קצרות, החזק למילים ארוכות)",
"tapToSync": "הקש לסנכרון",
"tapWhenStarts": "הקש על הכפתור כשהמילה מתחילה",
"wordProgress": "מילה {index} מתוך {total}:",
"cancelSync": "ביטול",
"hold": "החזק...",
"tap": "הקש"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "फिर से चलाओ",
"revertAllAuto": "सब पहले जैसा करो",
"showDetails": "विवरण दिखाओ"
},
"editTimeline": {
"holdingRelease": "दबाए हुए... शब्द खत्म होने पर छोड़ दो",
"pressSpacebarHint": "शब्द शुरू होने पर स्पेसबार दबाओ (छोटे शब्दों के लिए टैप करो, लंबे शब्दों के लिए दबाए रखो)",
"tapToSync": "सिंक करने के लिए टैप करो",
"tapWhenStarts": "शब्द शुरू होने पर बटन टैप करो",
"wordProgress": "{total} में से {index} शब्द:",
"cancelSync": "कैंसल",
"hold": "दबाए रखो...",
"tap": "टैप करो"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Pokreni ponovno",
"revertAllAuto": "Poništi sve",
"showDetails": "Prikaži detalje"
},
"editTimeline": {
"holdingRelease": "Drži... Otpusti kada riječ završi",
"pressSpacebarHint": "Pritisni razmaknicu kada riječ počne (dodirni za kratke riječi, drži za duge riječi)",
"tapToSync": "Dodirni za sinkronizaciju",
"tapWhenStarts": "Dodirni gumb kada riječ počne",
"wordProgress": "Riječ {index} od {total}:",
"cancelSync": "Odustani",
"hold": "DRŽI...",
"tap": "TAPNI"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Újrafuttatás",
"revertAllAuto": "Összes visszavonása",
"showDetails": "Részletek megjelenítése"
},
"editTimeline": {
"holdingRelease": "Nyomva tartva... Engedd fel, amikor a szó véget ér",
"pressSpacebarHint": "Nyomd meg a szóköz billentyűt, amikor a szó elkezdődik (rövid szavaknál röviden nyomd meg, hosszú szavaknál tartsd nyomva)",
"tapToSync": "Koppints a szinkronizáláshoz",
"tapWhenStarts": "Koppints a gombra, amikor a szó elkezdődik",
"wordProgress": "{index}. szó a(z) {total} közül:",
"cancelSync": "Mégse",
"hold": "TARTSD...",
"tap": "KOPPINTS"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Jalankan ulang",
"revertAllAuto": "Kembalikan semua",
"showDetails": "Tampilkan detail"
},
"editTimeline": {
"holdingRelease": "Menahan... Lepaskan saat kata selesai",
"pressSpacebarHint": "Tekan spasi saat kata dimulai (ketuk untuk kata pendek, tahan untuk kata panjang)",
"tapToSync": "Ketuk Untuk Sinkronisasi",
"tapWhenStarts": "Ketuk tombol saat kata dimulai",
"wordProgress": "Kata {index} dari {total}:",
"cancelSync": "Batal",
"hold": "TAHAN...",
"tap": "KETUK"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Riesegui",
"revertAllAuto": "Ripristina tutto",
"showDetails": "Mostra dettagli"
},
"editTimeline": {
"holdingRelease": "Stai premendo... Rilascia quando la parola finisce",
"pressSpacebarHint": "Premi la barra spaziatrice quando inizia la parola (premi brevemente per le parole brevi, tieni premuto per quelle lunghe)",
"tapToSync": "Tocca per sincronizzare",
"tapWhenStarts": "Tocca il pulsante quando inizia la parola",
"wordProgress": "Parola {index} di {total}:",
"cancelSync": "Annulla",
"hold": "TIENI PREMUTO...",
"tap": "TOCCA"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "再実行",
"revertAllAuto": "すべて元に戻す",
"showDetails": "詳細を表示"
},
"editTimeline": {
"holdingRelease": "長押し中... 単語が終わったら離してください",
"pressSpacebarHint": "単語が始まったらスペースキーを押してください(短い単語はタップ、長い単語は長押し)",
"tapToSync": "タップして同期",
"tapWhenStarts": "単語が始まったらボタンをタップしてください",
"wordProgress": "単語 {index} / {total}:",
"cancelSync": "キャンセル",
"hold": "長押し...",
"tap": "タップ"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "다시 실행",
"revertAllAuto": "모두 되돌리기",
"showDetails": "세부 정보 보기"
},
"editTimeline": {
"holdingRelease": "누르는 중... 단어가 끝나면 떼세요",
"pressSpacebarHint": "단어가 시작될 때 스페이스바를 누르세요 (짧은 단어는 짧게, 긴 단어는 길게 누르세요)",
"tapToSync": "탭하여 동기화",
"tapWhenStarts": "단어가 시작될 때 버튼을 탭하세요",
"wordProgress": "단어 {index} / {total}:",
"cancelSync": "취소",
"hold": "길게 누르기...",
"tap": "탭"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Jalankan semula",
"revertAllAuto": "Kembalikan semua",
"showDetails": "Papar butiran"
},
"editTimeline": {
"holdingRelease": "Sedang ditahan... Lepaskan apabila perkataan tamat",
"pressSpacebarHint": "Tekan bar ruang apabila perkataan bermula (ketik untuk perkataan pendek, tahan untuk perkataan panjang)",
"tapToSync": "Ketik Untuk Segerakkan",
"tapWhenStarts": "Ketik butang apabila perkataan bermula",
"wordProgress": "Perkataan {index} daripada {total}:",
"cancelSync": "Batal",
"hold": "TAHAN...",
"tap": "TEKAN"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Kjør på nytt",
"revertAllAuto": "Tilbakestill alle",
"showDetails": "Vis detaljer"
},
"editTimeline": {
"holdingRelease": "Holder inne... Slipp når ordet slutter",
"pressSpacebarHint": "Trykk på mellomromstasten når ordet starter (trykk kort for korte ord, hold inne for lange ord)",
"tapToSync": "Trykk for å synkronisere",
"tapWhenStarts": "Trykk på knappen når ordet starter",
"wordProgress": "Ord {index} av {total}:",
"cancelSync": "Avbryt",
"hold": "HOLD...",
"tap": "TRYKK"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Opnieuw uitvoeren",
"revertAllAuto": "Alles ongedaan maken",
"showDetails": "Details tonen"
},
"editTimeline": {
"holdingRelease": "Vasthouden... Laat los wanneer het woord eindigt",
"pressSpacebarHint": "Druk op de spatiebalk wanneer het woord begint (tik voor korte woorden, houd ingedrukt voor lange woorden)",
"tapToSync": "Tik om te synchroniseren",
"tapWhenStarts": "Tik op de knop wanneer het woord begint",
"wordProgress": "Woord {index} van {total}:",
"cancelSync": "Annuleren",
"hold": "HOUD VAST...",
"tap": "TIK"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Uruchom ponownie",
"revertAllAuto": "Cofnij wszystko",
"showDetails": "Pokaż szczegóły"
},
"editTimeline": {
"holdingRelease": "Przytrzymujesz... Puść, gdy słowo się skończy",
"pressSpacebarHint": "Naciśnij spację, gdy zaczyna się słowo (naciśnij krótko przy krótkich słowach, przytrzymaj przy długich)",
"tapToSync": "Stuknij, aby zsynchronizować",
"tapWhenStarts": "Stuknij przycisk, gdy zaczyna się słowo",
"wordProgress": "Słowo {index} z {total}:",
"cancelSync": "Anuluj",
"hold": "PRZYTRZYMAJ...",
"tap": "STUKNIJ"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Executar novamente",
"revertAllAuto": "Reverter tudo",
"showDetails": "Mostrar detalhes"
},
"editTimeline": {
"holdingRelease": "Segurando... Solte quando a palavra terminar",
"pressSpacebarHint": "Pressione a barra de espaço quando a palavra começar (toque para palavras curtas, segure para palavras longas)",
"tapToSync": "Toque para sincronizar",
"tapWhenStarts": "Toque no botão quando a palavra começar",
"wordProgress": "Palavra {index} de {total}:",
"cancelSync": "Cancelar",
"hold": "SEGURE...",
"tap": "TOQUE"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Rulează din nou",
"revertAllAuto": "Anulează tot",
"showDetails": "Arată detaliile"
},
"editTimeline": {
"holdingRelease": "Ții apăsat... Eliberează când se termină cuvântul",
"pressSpacebarHint": "Apasă tasta Space când începe cuvântul (apasă scurt pentru cuvinte scurte, ține apăsat pentru cuvinte lungi)",
"tapToSync": "Apasă pentru a sincroniza",
"tapWhenStarts": "Apasă butonul când începe cuvântul",
"wordProgress": "Cuvântul {index} din {total}:",
"cancelSync": "Anulează",
"hold": "ȚINE...",
"tap": "APASĂ"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Перезапустить",
"revertAllAuto": "Отменить все",
"showDetails": "Показать подробности"
},
"editTimeline": {
"holdingRelease": "Удерживай... Отпусти, когда слово закончится",
"pressSpacebarHint": "Нажми пробел, когда начнется слово (нажми коротко для коротких слов, удерживай для длинных)",
"tapToSync": "Нажми для синхронизации",
"tapWhenStarts": "Нажми кнопку, когда начнется слово",
"wordProgress": "Слово {index} из {total}:",
"cancelSync": "Отмена",
"hold": "УДЕРЖИВАЙ...",
"tap": "НАЖМИ"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Spustiť znova",
"revertAllAuto": "Vrátiť všetko",
"showDetails": "Zobraziť podrobnosti"
},
"editTimeline": {
"holdingRelease": "Držíš... Uvoľni, keď sa slovo skončí",
"pressSpacebarHint": "Stlač medzerník, keď sa slovo začne (ťukni pre krátke slová, podrž pre dlhé slová)",
"tapToSync": "Ťukni pre synchronizáciu",
"tapWhenStarts": "Ťukni na tlačidlo, keď sa slovo začne",
"wordProgress": "Slovo {index} z {total}:",
"cancelSync": "Zrušiť",
"hold": "DRŽ...",
"tap": "ŤUKNI"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Kör igen",
"revertAllAuto": "Återställ alla",
"showDetails": "Visa detaljer"
},
"editTimeline": {
"holdingRelease": "Håller in... Släpp när ordet slutar",
"pressSpacebarHint": "Tryck på mellanslag när ordet börjar (tryck snabbt för korta ord, håll in för långa ord)",
"tapToSync": "Tryck för att synka",
"tapWhenStarts": "Tryck på knappen när ordet börjar",
"wordProgress": "Ord {index} av {total}:",
"cancelSync": "Avbryt",
"hold": "HÅLL IN...",
"tap": "TRYCK"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "รันอีกครั้ง",
"revertAllAuto": "คืนค่าทั้งหมด",
"showDetails": "แสดงรายละเอียด"
},
"editTimeline": {
"holdingRelease": "กำลังกดค้าง... ปล่อยเมื่อจบคำ",
"pressSpacebarHint": "กดสเปซบาร์เมื่อเริ่มคำ (แตะสำหรับคำสั้นๆ กดค้างสำหรับคำยาวๆ)",
"tapToSync": "แตะเพื่อซิงค์",
"tapWhenStarts": "แตะปุ่มเมื่อเริ่มคำ",
"wordProgress": "คำที่ {index} จาก {total}:",
"cancelSync": "ยกเลิก",
"hold": "กดค้าง...",
"tap": "แตะ"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Patakbuhin muli",
"revertAllAuto": "Ibalik lahat",
"showDetails": "Ipakita ang mga detalye"
},
"editTimeline": {
"holdingRelease": "Naka-hold... Bitawan kapag tapos na ang salita",
"pressSpacebarHint": "Pindutin ang spacebar kapag nagsimula ang salita (i-tap para sa maiikling salita, i-hold para sa mahahabang salita)",
"tapToSync": "I-tap Para I-sync",
"tapWhenStarts": "I-tap ang button kapag nagsimula ang salita",
"wordProgress": "Salita {index} ng {total}:",
"cancelSync": "Kanselahin",
"hold": "I-HOLD...",
"tap": "I-TAP"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Yeniden çalıştır",
"revertAllAuto": "Tümünü geri al",
"showDetails": "Detayları göster"
},
"editTimeline": {
"holdingRelease": "Basılı tutuluyor... Kelime bittiğinde bırak",
"pressSpacebarHint": "Kelime başladığında boşluk tuşuna bas (kısa kelimeler için dokun, uzun kelimeler için basılı tut)",
"tapToSync": "Senkronize Etmek İçin Dokun",
"tapWhenStarts": "Kelime başladığında düğmeye dokun",
"wordProgress": "Kelime {index} / {total}:",
"cancelSync": "İptal",
"hold": "BASILI TUT...",
"tap": "DOKUN"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Перезапустити",
"revertAllAuto": "Скасувати все",
"showDetails": "Показати деталі"
},
"editTimeline": {
"holdingRelease": "Тримай... Відпусти, коли слово закінчиться",
"pressSpacebarHint": "Натисни пробіл, коли починається слово (натисни для коротких слів, утримуй для довгих)",
"tapToSync": "Натисни для синхронізації",
"tapWhenStarts": "Натисни кнопку, коли починається слово",
"wordProgress": "Слово {index} з {total}:",
"cancelSync": "Скасувати",
"hold": "УТРИМУЙ...",
"tap": "НАТИСНИ"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "Chạy lại",
"revertAllAuto": "Hoàn tác tất cả",
"showDetails": "Hiện chi tiết"
},
"editTimeline": {
"holdingRelease": "Đang giữ... Thả ra khi từ kết thúc",
"pressSpacebarHint": "Nhấn phím cách khi từ bắt đầu (nhấn thả cho từ ngắn, nhấn giữ cho từ dài)",
"tapToSync": "Nhấn để đồng bộ",
"tapWhenStarts": "Nhấn nút khi từ bắt đầu",
"wordProgress": "Từ {index} trên {total}:",
"cancelSync": "Hủy",
"hold": "GIỮ...",
"tap": "CHẠM"
}
},
"instrumentalReview": {
+10
View File
@@ -1389,6 +1389,16 @@
"rerun": "重新运行",
"revertAllAuto": "全部还原",
"showDetails": "显示详情"
},
"editTimeline": {
"holdingRelease": "按住... 在字词结束时松开",
"pressSpacebarHint": "在字词开始时按空格键(短词轻按,长词按住)",
"tapToSync": "点击同步",
"tapWhenStarts": "在字词开始时点击按钮",
"wordProgress": "第 {index} 个词,共 {total} 个:",
"cancelSync": "取消",
"hold": "按住...",
"tap": "点击"
}
},
"instrumentalReview": {
@@ -73,6 +73,11 @@ class ReviewServer:
output_config: OutputConfig,
audio_filepath: str,
logger: logging.Logger,
# Optional vocals stem powering the timeline waveform. May be absent when a
# separation produced no vocals stem; the vocals endpoint 404s and the
# frontend hides the waveform. Keyword-only in practice (callers pass it by
# name), so relocating it here from before `logger` is safe.
vocals_filepath: Optional[str] = None,
# Instrumental review data (optional - for combined review flow)
instrumental_options: Optional[List[Dict[str, Any]]] = None,
backing_vocals_analysis: Optional[Dict[str, Any]] = None,
@@ -85,7 +90,8 @@ class ReviewServer:
Args:
correction_result: The lyrics correction result to review
output_config: Output configuration
audio_filepath: Path to the main audio file (vocals)
audio_filepath: Path to the main audio file
vocals_filepath: Path the vocals (lead & backing) audio file
logger: Logger instance
instrumental_options: List of instrumental options for selection
Each option: {"id": str, "label": str, "audio_path": str}
@@ -97,6 +103,7 @@ class ReviewServer:
self.correction_result = correction_result
self.output_config = output_config
self.audio_filepath = audio_filepath
self.vocals_filepath = vocals_filepath
self.logger = logger or logging.getLogger(__name__)
self.review_completed = False
self.corrections_saved = False # Flag for intermediate save (before instrumental review)
@@ -416,6 +423,8 @@ class ReviewServer:
async def get_audio_with_job_id(job_id: str, audio_hash: str):
if audio_hash in _stem_names:
return await self.get_instrumental_audio(audio_hash)
elif audio_hash == "vocals":
return await self.get_vocals_audio()
return await self.get_audio(audio_hash)
self.app.add_api_route("/api/review/{job_id}/audio/{audio_hash}", get_audio_with_job_id, methods=["GET"])
@@ -1072,6 +1081,25 @@ class ReviewServer:
return FileResponse(audio_path, media_type=content_type, filename=os.path.basename(audio_path))
async def get_vocals_audio(self):
"""Stream the vocals audio file."""
audio_path = self.vocals_filepath
if not audio_path or not os.path.exists(audio_path):
raise HTTPException(status_code=404, detail="Vocals audio not found")
# Determine content type based on file extension
ext = os.path.splitext(audio_path)[1].lower()
content_types = {
".flac": "audio/flac",
".mp3": "audio/mpeg",
".wav": "audio/wav",
".m4a": "audio/mp4",
}
content_type = content_types.get(ext, "application/octet-stream")
return FileResponse(audio_path, media_type=content_type, filename=os.path.basename(audio_path))
async def generate_preview_video(self, updated_data: Dict[str, Any] = Body(...)):
"""Generate a preview video with the current corrections."""
# Check if preview video generation is allowed (disabled with --no-video)
+14 -4
View File
@@ -213,11 +213,20 @@ def run_combined_review(
backing_vocals_path = None
clean_instrumental_path = None
vocals_path = None
clean_result = separated.get("clean_instrumental", {})
if isinstance(clean_result, dict) and clean_result.get("instrumental"):
clean_instrumental_path = _resolve_path_for_cwd(clean_result["instrumental"], track_dir)
if not os.path.exists(clean_instrumental_path):
clean_instrumental_path = None
if isinstance(clean_result, dict):
if clean_result.get("instrumental"):
clean_instrumental_path = _resolve_path_for_cwd(clean_result["instrumental"], track_dir)
if not os.path.exists(clean_instrumental_path):
clean_instrumental_path = None
# Vocals stem drives the (optional) timeline waveform. A partial separation
# may have an instrumental but no "vocals" key — use .get() and never crash
# the review over a missing waveform; the frontend tolerates its absence.
vocals_source = clean_result.get("vocals")
if vocals_source:
candidate = _resolve_path_for_cwd(vocals_source, track_dir)
vocals_path = candidate if os.path.exists(candidate) else None
with_backing_path = None
combined_result = separated.get("combined_instrumentals", {})
@@ -303,6 +312,7 @@ def run_combined_review(
correction_result=correction_result,
output_config=output_config,
audio_filepath=resolved_audio or "",
vocals_filepath=vocals_path,
logger=logger,
# Instrumental review data
instrumental_options=instrumental_options,
+1 -1
View File
@@ -1,6 +1,6 @@
[tool.poetry]
name = "karaoke-gen"
version = "0.197.0"
version = "0.197.1"
description = "Generate karaoke videos with synchronized lyrics. Handles the entire process from downloading audio and lyrics to creating the final video with title screens."
authors = ["Andrew Beveridge <andrew@beveridge.uk>"]
license = "MIT"