narrator-ts

TypeScript reimplementation of the Amiga's speech pair: translator.library (English to phonemes) and narrator.device 33.2 (phoneme formant synthesis).

Output is Paula-native — 8-bit signed samples and the period they were written with. Verified against the real binaries running under a 68000 emulator: byte-exact translation on 15,405 phrases, sample-exact synthesis end to end.

npm install narrator-ts

import { speak } from 'narrator-ts'
import { translate } from 'narrator-ts/translator'
import rules from 'narrator-ts/reference/nrl-table.json' with { type: 'json' }
import voice from 'narrator-ts/reference/voice-free.json' with { type: 'json' }

const { phonemes } = translate('hello world', rules)   // '/HEHLOW WERLD '
const { pcm, sampleRate } = speak(Buffer.from(phonemes, 'latin1'), voice)
// Int8Array(26112) at 22030 Hz

Both tables are the free ones, so that runs with nothing else installed. Swap either for an extracted Amiga table to get the authentic voice — see below.

pcm is 8-bit signed. speak takes narrator_rb's own parameters — pitch, rate, sex, mode, sampfreq, mouths — and returns one entry per sentence alongside the joined samples, because that is how the device produces them.

Voices

A voice is ~12 KB of tables: formant frequencies and amplitudes per phoneme, durations, allophonic rewrite rules, the glottal waveform, fricative noise. They are a parameter, not built in.

The authentic Amiga voice is Commodore and SoftVoice's, so it is not in this repository. Extract it from Workbench disk images you already own:

python3 tools/extract-devices.py /path/to/workbench-disks -o fixtures/amiga
python3 tools/gen-voice.py   fixtures/amiga/narrator_device-33.2-*.bin     -o data
python3 tools/gen-tables.py  fixtures/amiga/translator_library-33.2-1*.bin -o data

The free voice is reference/voice-free.json, built from published phonetics by tools/gen-free-voice.py and owing nothing to anyone. It is the default, so a fresh clone speaks. It is not the Amiga's voice and does not claim to be — reference/README.md says what is measured and what is still provisional.

The letter-to-sound half has its free equivalent too, and is the default when no Amiga table has been extracted:

import nrl from 'narrator-ts/reference/nrl-table.json' with { type: 'json' }
translate('hello world', nrl)   // '/HEHLOW WERLD '

That is the rule set of NRL Report 7948 (Elovitz et al., 1976 — a US Government work), rebuilt from the report alone. It is not byte-compatible with the Amiga and is not meant to be: 64.6% of distinct words match. research/03-nrl-provenance.md has the measurements.

Command line

npm run say -- 'hello world' -o hello.wav          # the free voice
npm run say -- 'hello world' -o hello.wav -V 33.2  # an extracted Amiga one
npm run say -- -p '/HEH4LOW WER4LD' -o hello.wav
npm run say -- 'is this a question' --pitch 200 --rate 100

--pitch --rate --sex --mode --sampfreq --mouths are narrator_rb's own fields. npm run say with no arguments prints them.

Every tool under tools/ takes --help.

AROS resource export

Create the versioned IFF resource consumed by AROS translator.library, narrator.device, and speech.device:

npm run export:aros -- -o speech.iff
npm run export:aros -- -o speech.iff \
  --translator data/translator-33.2.json \
  --voice data/narrator-33.2.json \
  --translator-license "User supplied; redistribution not granted" \
  --voice-license "User supplied; redistribution not granted"
npm run export:aros -- -o speech-resource.c --format c

The first command uses the redistributable reference tables. The second uses locally extracted authentic tables; the resulting file has the same legal status as those inputs and should not be redistributed. The C form contains the identical IFF payload as a byte array for ROM builds.

FORM NARR uses ordinary length-delimited Latin-1 IFF chunks rather than an embedded serialization format. VERS is the binary resource schema version; FVER records the narrator-ts package version. TVER/TSRC/TLIC describe the translator and VVER/VSRC/VLIC describe the voice, followed by their LTRS and NVOI table chunks. The built-in free inputs are labelled public domain. Explicit table paths conservatively default to User supplied; redistribution not granted; the two --*-license options override that text.

Status

| | | |---|---| | translator | byte-exact against all 6 shipped builds | | narrator 33.2 | byte-exact front half, sample-exact renderer, end to end | | free letter-to-sound table | built, divergence measured | | free voice | speaks; texture within a few percent of 33.2's, no allophonic rules yet |

narrator.device 37.7 is a rewrite and is out of scope — one voice done properly. 1.6, 31.13, 33.2 and 36.9 are sample-identical over 4,865 phrases, so 33.2 covers four of the five shipped builds anyway.

Licence

Code is MIT.

The Amiga binaries and everything derived by running them are not redistributable and are gitignored: fixtures/amiga/*.bin, fixtures/golden/, data/. None has ever been committed.

reference/ is the opposite — material that can be redistributed, with its provenance recorded in reference/README.md.

narrator.device's copyright line reads Mark Barton / Joseph Katz. It was licensed in from SoftVoice, Inc.; Commodore never owned it.

Documentation

| | | |---|---| | docs/development.md | building the oracle, regenerating fixtures, coverage | | research/00-findings.md | what the binaries turned out to be | | research/01-translator.md | the letter-to-sound engine | | research/02-narrator.md | the synthesizer, stage by stage | | research/03-nrl-provenance.md | which rules came from NRL and which SoftVoice added |