Press n or j to go to the next uncovered block, b, p or k for the previous block.
| 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 | import { useRef, useCallback } from 'react' import { useTTS } from '@/hooks/useTTS' import { useAudioManagerInstance } from '@/contexts/AudioManagerContext' // ── Adaptive timing constants ──────────────────────────────────────── /** * Max ratio of effective duration to designed duration (stretch). * 3.0 means animation won't slow below ~0.33× designed speed. * Generous to accommodate TTS clips that are significantly longer * than the designed animation (e.g. 9.5s audio for a 5s animation). */ const MAX_STRETCH_RATIO = 3.0 /** * Min ratio of effective duration to designed duration (compress). * 0.6 means animation won't speed above ~1.67× designed speed. * Only applies when `adaptiveCompress` is true on the segment. */ const MIN_COMPRESS_RATIO = 0.6 /** * When TTS finishes and the remaining animation time is within this * threshold, advance to the next segment immediately instead of waiting. * Eliminates silent gaps between segments for continuous narration flow. */ const ADVANCE_TOLERANCE_MS = 1000 /** * Compute effective animation duration based on actual TTS audio length. * * Default (stretch only): animation stretches to match longer audio but * never plays faster than the designed pace. * * With adaptiveCompress: animation can also speed up when audio is shorter, * clamped to MIN_COMPRESS_RATIO. * * Both directions are clamped to MAX_STRETCH_RATIO to prevent absurdly * slow animations. */ function computeEffectiveDuration(seg: SequencerSegment, audioDurationMs: number | null): number { if (audioDurationMs === null) return seg.animationDurationMs const designedMs = seg.animationDurationMs const maxMs = designedMs * MAX_STRETCH_RATIO if (seg.adaptiveCompress) { const minMs = designedMs * MIN_COMPRESS_RATIO return Math.max(minMs, Math.min(audioDurationMs, maxMs)) } // Stretch only: never faster than designed return Math.max(designedMs, Math.min(audioDurationMs, maxMs)) } // ── Types ─────────────────────────────────────────────────────────── /** Minimum shape a segment needs for the sequencer to drive it. */ export interface SequencerSegment { ttsText: string ttsTone?: string animationDurationMs: number /** * Allow animation to compress (speed up) when audio is shorter than * the designed duration. Default: false (stretch only — animation * slows to match longer audio but never plays faster than designed). */ adaptiveCompress?: boolean } /** Returned by `tick()` every frame while the sequencer is active. */ export interface SequencerTickResult { /** Index of the segment currently playing. */ segmentIndex: number /** 0–1 fraction of the current segment's animation duration elapsed. */ animFrac: number /** * Cumulative virtual time (ms): sum of completed segment durations * plus clamped elapsed time in the current segment. Useful for * driving time-based overlays (e.g. sieve animation). */ virtualTimeMs: number /** True when all segments have completed (both TTS and animation). */ allDone: boolean } // ── Hook ──────────────────────────────────────────────────────────── /** * Reusable TTS segment sequencer. * * Encapsulates the three-ref pattern (segmentIndex, segmentStartMs, * ttsFinished), the `speakSegment` function, and the dual-gate * advancement logic (advance only when both TTS **and** animation * duration have completed). * * Used by both `usePrimeTour` (for segmented stops) and * `useConstantDemoNarration` (for all constant demos). * * The sequencer does **not** own a RAF loop — callers invoke `tick()` * from their own animation frame. */ export function useNarrationSequencer() { const audioManager = useAudioManagerInstance() const speak = useTTS({ tone: '', say: { en: '' } }) const activeRef = useRef(false) const mutedRef = useRef(false) const segmentsRef = useRef<SequencerSegment[]>([]) const toneRef = useRef('') const segmentIndexRef = useRef(0) const segmentStartMsRef = useRef(0) const ttsFinishedRef = useRef(false) // Adaptive timing: effective durations used for completed segments, // so virtualTimeMs stays consistent after segment advancement. const completedEffDurationsRef = useRef<number[]>([]) // Timing ref: when did the current segment's TTS audio finish? const ttsFinishedAtRef = useRef(0) // Latched audio duration for the current segment — set once when // metadata becomes available, stays stable for the rest of the segment // to prevent animFrac jumps when duration changes from null to real. const segAudioDurRef = useRef<number | null>(null) // Initial fraction for mid-segment resume: when startFrom is called // with a non-zero initialFrac, tick() maps elapsed time from // initialFrac→1 instead of 0→1, so the animation picks up exactly // where the user scrubbed regardless of adaptive timing. const initialFracRef = useRef(0) /** Speak a segment's TTS text, flagging ttsFinished when done. */ const speakSegment = useCallback( (seg: SequencerSegment, tone: string, segIdx?: number) => { ttsFinishedRef.current = false // When muted, leave ttsFinished = false — the sequencer will hold at // the segment boundary (animFrac = 1) until releaseTtsGate() is called. // This lets the voice agent talk at their own pace. if (mutedRef.current) { return } speak({ say: { en: seg.ttsText }, tone: seg.ttsTone ?? tone }) .then(() => { ttsFinishedAtRef.current = performance.now() ttsFinishedRef.current = true }) .catch(() => { ttsFinishedAtRef.current = performance.now() // TTS failed or cancelled — still allow segment advancement ttsFinishedRef.current = true }) // Preload next segment's audio so the transition is instant (no network wait) if (segIdx !== undefined) { const nextSeg = segmentsRef.current[segIdx + 1] if (nextSeg) { audioManager.preloadForSpeak({ say: { en: nextSeg.ttsText }, tone: nextSeg.ttsTone ?? tone, }) } } }, [speak, audioManager] ) /** * Begin sequencing from segment 0. Speaks the first segment's TTS * immediately. */ const start = useCallback( (segments: SequencerSegment[], tone: string) => { if (segments.length === 0) return activeRef.current = true segmentsRef.current = segments toneRef.current = tone segmentIndexRef.current = 0 segmentStartMsRef.current = performance.now() ttsFinishedRef.current = false completedEffDurationsRef.current = [] segAudioDurRef.current = null initialFracRef.current = 0 speakSegment(segments[0], tone, 0) }, [speakSegment] ) /** * Begin sequencing from an arbitrary segment index. * Pre-populates completedEffDurationsRef with designed durations for * skipped segments so virtualTimeMs stays consistent. * * @param initialFrac Starting animation fraction (0–1) within the * segment. tick() maps elapsed time from * initialFrac→1 so progress resumes from the * exact scrubbed position regardless of adaptive * timing stretching the effective duration. */ const startFrom = useCallback( (segments: SequencerSegment[], tone: string, fromIndex: number, initialFrac = 0) => { if (segments.length === 0 || fromIndex >= segments.length) return activeRef.current = true segmentsRef.current = segments toneRef.current = tone segmentIndexRef.current = fromIndex segmentStartMsRef.current = performance.now() ttsFinishedRef.current = false segAudioDurRef.current = null initialFracRef.current = initialFrac // Pre-fill completed durations for skipped segments using designed durations const effDurations: number[] = [] for (let i = 0; i < fromIndex; i++) { effDurations[i] = segments[i].animationDurationMs } completedEffDurationsRef.current = effDurations speakSegment(segments[fromIndex], tone, fromIndex) }, [speakSegment] ) /** Stop sequencing and cancel current TTS playback. */ const stop = useCallback(() => { if (!activeRef.current) return activeRef.current = false audioManager.stop() }, [audioManager]) /** * Advance the sequencer by one frame. Call from your RAF loop. * * @param speedMultiplier Scale factor for elapsed time (default 1). * Values > 1 speed up; < 1 slow down. * @returns Current state, or `null` if the sequencer is not active. */ const tick = useCallback( (speedMultiplier = 1): SequencerTickResult | null => { if (!activeRef.current) return null const segments = segmentsRef.current const segIdx = segmentIndexRef.current const seg = segments[segIdx] if (!seg) { activeRef.current = false return null } const now = performance.now() const rawElapsed = (now - segmentStartMsRef.current) * speedMultiplier // Adaptive timing: latch audio duration once known, then use it for // the rest of the segment to prevent animFrac jumps. if (segAudioDurRef.current === null) { const dur = audioManager.getCurrentAudioDurationMs() if (dur !== null) segAudioDurRef.current = dur } const audioDurationMs = segAudioDurRef.current const effectiveDuration = computeEffectiveDuration(seg, audioDurationMs) // When resuming mid-segment (initialFrac > 0), only the remaining // portion of the segment's effective duration is left to play. // Map elapsed time into initialFrac→1 so the animation picks up // exactly where the user scrubbed, regardless of adaptive stretching. const startFrac = initialFracRef.current const remainingDuration = effectiveDuration * (1 - startFrac) const remainingFrac = remainingDuration > 0 ? Math.min(1, rawElapsed / remainingDuration) : 1 const animFrac = startFrac + remainingFrac * (1 - startFrac) // For advancement gating, use elapsed relative to remaining duration const segElapsed = rawElapsed // Advance when TTS is done AND animation is done or within tolerance. // The tolerance eliminates silent gaps: if the narrator finishes and // the animation has < 1s remaining, skip ahead so the next clip starts // immediately rather than waiting in silence. let allDone = false let advancedToNext = false const advanceThreshold = Math.max(0, remainingDuration - ADVANCE_TOLERANCE_MS) const shouldAdvance = ttsFinishedRef.current && segElapsed >= advanceThreshold if (shouldAdvance) { // Record effective duration for virtualTimeMs consistency completedEffDurationsRef.current[segIdx] = effectiveDuration const nextIdx = segIdx + 1 if (nextIdx < segments.length) { // Advance to next segment — speak() handles cancellation internally // and won't clear the preloaded audio cache, so the next clip // starts instantly from the already-buffered Audio element. segmentIndexRef.current = nextIdx segmentStartMsRef.current = now ttsFinishedRef.current = false segAudioDurRef.current = null initialFracRef.current = 0 // new segment starts from 0 speakSegment(segments[nextIdx], toneRef.current, nextIdx) advancedToNext = true } else { allDone = true activeRef.current = false } } // Compute cumulative virtual time using effective durations let virtualTimeMs = 0 const effDurations = completedEffDurationsRef.current for (let i = 0; i < segmentIndexRef.current; i++) { virtualTimeMs += effDurations[i] ?? segments[i].animationDurationMs } const currentSeg = segments[segmentIndexRef.current] if (currentSeg) { const currentElapsed = (now - segmentStartMsRef.current) * speedMultiplier const currentAudioMs = audioManager.getCurrentAudioDurationMs() const currentEffective = computeEffectiveDuration(currentSeg, currentAudioMs) // Account for initial frac when computing virtual time const currentStartFrac = initialFracRef.current const currentRemaining = currentEffective * (1 - currentStartFrac) virtualTimeMs += currentStartFrac * currentEffective + Math.min(currentElapsed, currentRemaining) } return { segmentIndex: allDone ? segments.length - 1 : segmentIndexRef.current, // When advancing to next segment, report animFrac=0 (new segment // just started) to avoid a one-frame discontinuity where the old // segment's partial animFrac is applied to the new segment's range. animFrac: allDone ? 1 : advancedToNext ? 0 : animFrac, virtualTimeMs, allDone, } }, [audioManager, speakSegment] ) /** Release the TTS gate so the sequencer advances past the current segment. * Used when muted: external code (e.g. voice-agent silence detection) calls * this to signal the narrator finished speaking for the current segment. */ const releaseTtsGate = useCallback(() => { if (!activeRef.current) return ttsFinishedAtRef.current = performance.now() ttsFinishedRef.current = true }, []) return { start, startFrom, tick, stop, mutedRef, releaseTtsGate, segmentIndexRef } } |