Ecosystem animation — voice-over

Narration assets for ../ecosystem-animation-standalone.html (40 s loop, seven sections).

• VO-SCRIPT.md — the timecoded script (93 words / 40 s) with delivery cues. Source of truth for the copy.
• cues.json — the same table, machine-readable: [{ "section", "start", "end", "line", "cue" }], start/end in seconds. Regenerate from VO-SCRIPT.md; never edit the copy here first.
• ecosystem-vo-v{N}.wav — recorded takes, added here as they land (see naming below).

Recording spec

• Format: mono WAV, 48 kHz, 24-bit.
• Loudness: −16 LUFS integrated, −1 dBTP true-peak ceiling.
• Delivery: measured, declarative read, ~140 wpm, no sell.
• SRPVDAL: read as its seven stage names — sense, reason, plan, validate, decide, act, learn — never as a word.
• Player state while recording: captions must be off in the player (Tweaks → captions); the on-screen captions duplicate the track and will throw the read.
• Hit the timing cues in VO-SCRIPT.md; finish the CLOSE line by 0:39 so the fade plays in silence.

File naming

ecosystem-vo-v{N}.wav — N is an integer take/version number starting at 1 (ecosystem-vo-v1.wav, ecosystem-vo-v2.wav, …). Never overwrite a numbered take; bump N. Any mixed or trimmed derivative carries the same N plus a suffix (ecosystem-vo-v2-trimmed.wav).

Status

Take 1 mounted. ecosystem-vo-v1.wav (48 kHz mono 16-bit, −16 LUFS, 40.0s) and ecosystem-vo-v1.mp3 are generated and synchronized with cues.json.