Kid mode 🧒

This page is Matilda Voice Inventory. It is one of the projects Brandon built and put online. The grown-up details are below, tap Grown-up mode to read them the normal way.

Matilda Voice Inventory

Complete manifest of every file produced with the ElevenLabs Matilda voice (ID: XrExE9yKIg1WjnnlVkGX) across the Chiron Intuition tarot pipeline. All runs occurred June 5, 2026 in a single batch session.

TTS Runs
7
Characters
60,177
Words Voiced
11,339
API Chunks
21
Processing
49m
Audio Files
1,715
Total Audio
3.2GB
Credits Left
0

Final Video Deliverables (Matilda-Voiced)

These MP4s contain ElevenLabs Matilda narration mixed over a chillwave ambient bed with beat-synced chalkboard card visuals.

VideoSourceSizeDateStatus
JpWA2IRTFNw_chalkboard_MATILDA.mp4 chiron-chalkboard-FULL/JpWA2IRTFNw/ 32 MB Jun 5 02:33 SHIPPED
JpWA2IRTFNw_first2min_chalkboard_MATILDA.mp4 chiron-chalkboard-2min-v10-caps/ 4.7 MB Jun 5 02:30 SHIPPED
chiron_longform_MATILDA.mp4 scribe-video/out/ 9.4 MB Jun 5 02:30 SHIPPED
mihAR69tVtc_chalkboard.mp4 chiron-chalkboard-FULL/mihAR69tVtc/ 38 MB Jun 5 02:37 MATILDA RUN
wpUCkqhE8v4_chalkboard.mp4 chiron-chalkboard-FULL/wpUCkqhE8v4/ 43 MB Jun 4 22:40 MATILDA RUN
XCZXHP0fY4o_chalkboard.mp4 chiron-chalkboard-FULL/XCZXHP0fY4o/ 45 MB Jun 5 00:23 MATILDA RUN

Total video output: ~172 MB across 6 Matilda-voiced videos

ElevenLabs API Runs (matilda_run.json)

Each row is a confirmed paid API call with its matilda_run.json receipt on disk.

#ProjectCharsWordsChunksRuntimeTimestamp
1 chiron-chalkboard-2min-v8 1,609311132s Jun 5 02:08
2 chiron-chalkboard-2min-v10-caps 1,609311132s Jun 5 02:07
3 chiron-chalkboard-FULL / JpWA2IRTFNw 12,2522,338421m 14s Jun 5 02:31
4 chiron-chalkboard-FULL / wpUCkqhE8v4 12,7682,436413m 53s Jun 5 02:35
5 chiron-chalkboard-FULL / mihAR69tVtc 12,8542,42544m 10s Jun 5 02:39
6 chiron-chalkboard-FULL / XCZXHP0fY4o 13,9082,58156m 54s Jun 5 02:46
7 scribe-video (longform concept map) 5,17793721m 33s Jun 5 02:10
TOTAL 60,17711,3392148m 48s
Credit depletion (60,177 / 60,177 used)

Voice Settings (All Runs)

Voice ID
XrExE9yKIg1WjnnlVkGX
Model
eleven_multilingual_v2
Stability
0.50
Similarity
0.75
Style
0.35
Speaker Boost
ON

Full Narration Audio (WAV per Project)

JpWA2IRTFNw (Full)
12,252 chars / 2,338 words / 4 chunks
  • narration_raw.wav (65 MB)
  • narration.wav (65 MB)
  • mixed.wav (65 MB, + bed)
  • ~130 scene WAVs
wpUCkqhE8v4 (Full)
12,768 chars / 2,436 words / 4 chunks
  • narration_raw.wav (71 MB)
  • narration.wav (71 MB)
  • mixed.wav (71 MB, + bed)
  • matilda_edited/ subfolder
mihAR69tVtc (Full)
12,854 chars / 2,425 words / 4 chunks
  • narration_raw.wav (70 MB)
  • narration.wav (70 MB)
  • mixed.wav (70 MB, + bed)
XCZXHP0fY4o (Full)
13,908 chars / 2,581 words / 5 chunks
  • narration_raw.wav (78 MB)
  • narration.wav (78 MB)
  • mixed.wav (78 MB, + bed)
2min v8 + v10 (Short)
1,609 chars / 311 words / 1 chunk (x2 runs)
  • v8: narration.wav + mixed.wav (9 MB each)
  • v10: narration_raw.wav + narration.wav + mixed.wav (9 MB each)
  • 27 scene WAVs per version
Scribe Longform
5,177 chars / 937 words / 2 chunks
  • narration.wav (25 MB)
  • narration_raw.wav (25 MB)
  • mixed.wav (25 MB, + bed)

Approval Markers (.matilda_approved)

The TTS spend gate requires user approval before any paid synthesis. These markers prove each run was authorized:

Longform Batch (8 Readings Queued)

8 Chiron reading transcripts were cleaned, ranked, and queued for Matilda chalkboard rendering. Only 2 received approval markers before the batch driver crashed on a missing stdbuf command. None completed rendering.

RankTitleTranscript SizeApprovedRendered
1You're the talk of the town75 KBNOFAILED
2They've got to go through god to get to you72 KBYESFAILED
3You've gone GODMODE82 KBYESFAILED
4MULTIPLE INJUSTICES BEING CORRECTED71 KBNOFAILED
5The destruction of dumb and dumber83 KBNOFAILED
6The lone wolf, we're up against these glu...81 KBNOFAILED
7YOU ARE the biggest glitch in the game57 KBNOFAILED
8IT'S ALL COMING BACK72 KBNOFAILED

All 8 failed with: chiron-master-driver.sh: line 125: stdbuf: command not found. No ElevenLabs credits were spent on these — they died before reaching the TTS stage.

Archived Iterations (Pre-Matilda)

These earlier versions in .chiron-archive-20260604/ used XTTS (free, local) voice, not ElevenLabs Matilda. Listed for completeness:

VersionVideoSizeDate
prototypeJpWA2IRTFNw_chalkboard.mp413 MBMay 31 12:15
2min-v1JpWA2IRTFNw_first2min_chalkboard.mp43.3 MBMay 31 12:57
2min-v3JpWA2IRTFNw_first2min_chalkboard.mp43.8 MBMay 31 13:42
2min-v4JpWA2IRTFNw_first2min_chalkboard.mp43.7 MBMay 31 14:15
2min-v5JpWA2IRTFNw_first2min_chalkboard.mp44.0 MBMay 31 14:41
2min-v6JpWA2IRTFNw_first2min_chalkboard.mp44.1 MBMay 31 14:53
2min-v7JpWA2IRTFNw_first2min_chalkboard.mp44.0 MBMay 31 16:05

iCloud Copies

Post-Depletion Fallback (Jun 8)

WgWy35I643I — Lester XTTS (free, local) XTTS
After ElevenLabs hit 0 credits, the pipeline fell back to local XTTS Lester voice. This is the most recent render.

Skill Infrastructure

chiron-matilda-rws-video
~/.claude/skills/chiron-matilda-rws-video/ (64 MB)
  • _lib/synth_matilda_bed.py — synthesis + bed mixing
  • _lib/approve_matilda.py — TTS spend gate
  • _lib/edit_narration.py — narration editor
  • tools/transcript_chalkboard.py — card renderer
  • beds/ — chillwave (81.6s) + shiner ambient
  • 78 Rider-Waite-Smith card images
tarot-chapter-voice-video
~/.claude/skills/tarot-chapter-voice-video/
  • _lib/synth_matilda_bed.py (shared copy)
  • _lib/approve_matilda.py (shared copy)

Voice Bakeoff (Jun 4)

Before committing to Matilda, a 4-voice comparison was run at chiron-voice-bakeoff/:

VoiceEngineReferenceTake Size
BrandonXTTS local_ref16k.wav (750 KB)5.9 MB
AshXTTS local_ref16k.wav (750 KB)7.6 MB
JohnXTTS local_ref16k.wav (750 KB)6.7 MB
Lester v3XTTS local_ref16k.wav (690 KB)6.0 MB

A/B comparison file: A_compare.wav (29 MB). Matilda (ElevenLabs) was chosen over all four XTTS options.

Kate Engine Work (Jun 6)

Kate tarot voice engine at Projects/_outputs/kate-engine-v3/. 17 text variations of 5-card Empress-reversed readings, no audio synthesis (text-only voice calibration):

Kate voice at 70% authenticity (vs Chiron 95-96%). Awaits the Chiron v2 patch before audio synthesis.

Session Lineage