Scicom-intl/clean-speech-raw-sources

Clean Speech — Raw Source Archives (mirror)

Durable public mirrors of speech corpora whose original home is not Hugging Face (external academic hosts disappear, move, or go offline). HF is used only as a faster/durable mirror — the original source links are below and remain the canonical home. Every archive is byte-for-byte unmodified from its origin and retains its original licence and attribution.

42 corpora · 297 GB · 171+ files.

Scope note. This mirror began as high-sample-rate (>=44.1 kHz) material for Scicom-intl/TTS-Clean44k. It has since broadened: it now also holds 16 kHz ASR corpora (Samromur, the Google/OpenSLR ASR sets, and others). Check the per-corpus notes before assuming sample rate.

Why mirror at all

Four of the corpora tracked alongside this repo already have dead or moved origin links, and several hosts here are slow enough that a large download is a multi-hour gamble (repository.clarin.is serves no HTTP range requests at all, so a dropped transfer restarts from zero). Mirroring makes the archives fast, resumable, and durable.

How to use

Tooling lives in malaysia-ai/dataset under multilingual-tts/prepare_raw_sources.py:

python3 prepare_raw_sources.py sources                     # registry + layouts
python3 prepare_raw_sources.py probe    --source <name>    # size/layout, no bulk download
python3 prepare_raw_sources.py mirror   --source <name>    # origin -> this repo
python3 prepare_raw_sources.py external --source <name> --from-mirror --run-prepare

--from-mirror pulls from here instead of the origin. Because HF serves ranged reads, you can also inspect an archive's layout remotely — read the zip central directory, or inflate a single metadata.tsv — without downloading gigabytes.

Ingestion status

• verified — layout confirmed by reading the real archive; safe to ingest.
• unverified — plausible layout, not yet exercised end-to-end.
• mirror-only — archive preserved, internal layout not confirmed. Do not assume an index->audio mapping; confirm it first.

Contents

OpenSLR (Google crowdsourced + Thorsten) — verified

Utterance ids are <langgender>_<speaker>_<uttnum>, so these carry real speaker labels and need no speaker clustering.

folder	lang	hours	GB	licence	origin
sundanese-asr_openslr36	su	~333	23.00	CC BY-SA 4.0	src
javanese-asr_openslr35	jv	~296	18.90	CC BY-SA 4.0	src
sinhala-asr_openslr52	si	~224	14.68	CC BY-SA 4.0	src
thorsten-german_openslr95	de	~24	3.00	CC0	src
peruvian-spanish_openslr73	es	~9	1.94	CC BY-SA 4.0	src
javanese-tts_openslr41	jv	~9	1.89	CC BY-SA 4.0	src
sundanese-tts_openslr44	su	~9	1.47	CC BY-SA 4.0	src
venezuelan-spanish_openslr75	es	~11	1.04	CC BY-SA 4.0	src
sinhala-tts_openslr30	si	~15	0.70	CC BY-SA 4.0	src
puertorico-spanish_openslr74	es	~1	0.21	CC BY-SA 4.0	src

CLARIN-IS (Icelandic)

The Samromur releases share one layout: audio at audio/<speaker_id>/<speaker_id>-<utt_id>.flac, 16 kHz mono, with a UTF-8 metadata.tsv (id, speaker_id, filename, sentence, sentence_norm, gender, age, ...). For TTS prefer sentence over sentence_norm: the latter is lowercased with punctuation stripped, discarding prosodic signal. malromur and icelandic-psc are separate corpora with their own layouts — see the caveats below.

folder	lang	hours	GB	licence	origin
malromur_clarin202	is	~136	10.22	CC BY 4.0	src
samromur-l2_clarin263	is	~150	7.94	CC BY 4.0	src
samromur-21.05_clarin189	is	~100	7.05	CC BY 4.0	src
samromur-mimic_clarin264	is	~67	3.60	CC BY 4.0	src
samromur-queries_clarin180	is	~21	1.03	CC BY 4.0	src
icelandic-psc_clarin197	is	~21	0.89	CC BY 3.0	src

Zenodo — layout verified

folder	lang	hours	GB	licence	origin
som-tts-thai_zenodo21530909	th	~30	3.63	CC BY 4.0	src
thorsten-de-2022.10_zenodo7265581	de	~34	1.39	CC BY 4.0	src
uspdatro-romanian_zenodo7898233	ro	~4	0.42	CC BY 4.0	src
thorsten-hessisch_zenodo10511260	de	~3	0.27	CC BY 4.0	src
dendi-parakou_zenodo6591186	ddn	~1	0.14	CC BY 4.0	src

Other origins

folder	lang	hours	GB	licence	origin
thai-elderly-speech_github	th	~17	5.07	CC BY-SA 4.0	src
siwis-french_edinburgh	fr	~10	2.87	CC BY 4.0	src

Zenodo — mirror-only (layout NOT confirmed)

Preserved as-is; confirm the layout before ingesting.

folder	lang	hours	GB	licence	origin
tundra-multilingual-tts_zenodo12543428	multi	?	16.00	CC BY 4.0	src
earnings25-finance-en_zenodo18762168	en	~500	12.04	CC BY 4.0	src
uvigo-gl-voices_zenodo8027725	gl	?	9.63	CC BY 4.0	src
pakistan-multilingual_zenodo19323537	ur	?	8.35	CC BY 4.0	src
lada-ukrainian-tts_zenodo7396774	uk	~10	6.78	Apache 2.0	src
chulalongkorn-spoken-thai_zenodo17366698	th	?	6.23	CC BY-NC-SA 4.0	src
cv298-japanese-tts_zenodo21119791	ja	?	5.20	CC0	src
runyankore-rukiga_zenodo20478456	nyn	?	3.94	CC BY 4.0	src
coala-dutch_zenodo8413584	nl	?	2.47	CC0	src
coala-english_zenodo8268928	en	?	1.78	CC0	src
coala-italian_zenodo8413135	it	?	1.76	CC0	src
french-tts-blizzard_zenodo13918615	fr	?	1.68	CC BY 4.0	src
gronings-nasal-besemah_zenodo7946870	gos	?	1.02	CC BY 4.0	src
dioula-words_zenodo13958016	dyu	?	0.11	CC BY 4.0	src
baule_zenodo6705861	bci	?	0.05	CC BY 4.0	src
alsatian-character-speech_zenodo10613199	gsw	?	0.01	CC BY 4.0	src

Earlier additions

• BibleTTS (OpenSLR 129) — bibletts_openslr129/*.tgz (92.8 GB). 48 kHz mono studio, up to 80 h single-speaker per language: Akuapem Twi, Asante Twi, Ewe, Hausa, Lingala, Yoruba. CC BY-SA 4.0. Source: https://www.openslr.org/129/ · Meyer et al., Interspeech 2022.
• RAVDESS — RAVDESS_Audio_Speech_Actors_01-24.zip (0.21 GB). 48 kHz, 1440 files, 24 actors (emotional). CC BY-NC-SA 4.0. Source: https://zenodo.org/records/1188976 · Livingstone & Russo, PLoS ONE 13(5).
• DAPS — DAPS_daps.tar.gz (16.1 GB). 44.1 kHz, 20 speakers. CC BY-NC-SA 4.0. Source: https://zenodo.org/records/4660670 · Mysore, IEEE SPL 22(8). Not ingestible as {audio, text} pairs — long unaligned reads with no per-utterance transcripts; needs segmentation + forced alignment first.

Per-corpus caveats

• samromur-mimic_clarin264 — ships 32,892 TTS-generated mp3 in repeat_audio/ beside 91,408 human flac in audio/, and metadata.tsv indexes both. Exclude repeat_audio/ or you will train on synthetic speech.
• icelandic-psc_clarin197 — 12 long parliamentary recordings with Transcriber .trs files, not per-utterance segments. Needs segmentation first.
• runyankore-rukiga_zenodo20478456 — 265 wav with no index file; the transcripts appear to live in the bundled .docx (mirrored for that reason).
• pakistan-multilingual_zenodo19323537 — the part inspected had no index file; transcripts may be in another part.
• french-tts-blizzard_zenodo13918615 — 39 long wav + 3 csv, not per-utterance clips.
• coala-* — distributed as .rar; needs unrar/7z.
• Split archives — samromur-21.05 and thai-elderly-speech are multi-volume. Info-ZIP volumes (.zip + .z01..) and byte-splits (.001/.002..) both need 7z; Python's zipfile reads neither.

Licensing

Licences are those of the original authors and vary per corpus (CC0, CC BY 4.0, CC BY-SA 4.0, CC BY-NC-SA 4.0, Apache 2.0) — see each table above and follow the origin link for the authoritative terms and citation. Several corpora are non-commercial. The repo-level tag is the most restrictive of the set and is not a substitute for checking the individual corpus.

Mirrored for research reproducibility. All attribution belongs to the original authors.