LibriTTS

High-fidelity (24 kHz) read English audiobooks derived from LibriSpeech, with normalised transcriptions; the standard baseline for English TTS.

• Task: tts
• Languages: en
• Hours: 585
• Domain: audiobook
• License: CC BY 4.0
• Homepage: https://www.openslr.org/60
• Paper: https://arxiv.org/abs/1904.02882

Recommendation

Best first choice for English TTS training. Audio is clean and well-segmented; train-clean-360 is a solid single-subset starting point. Use train-other-500 to add acoustic diversity. Lacks spontaneous or expressive speech — supplement with VCTK or Emilia for prosody range.

Getting the data

Downloadable via VoxKitchen (libritts, source: openslr, size: 881 MB - 41.5 GB):

vkit docker download --tag slim libritts --root ./data/libritts

Subsets: dev-clean, dev-other, test-clean, test-other, train-clean-100, train-clean-360, train-other-500.

Suggested processing

A recommended VoxKitchen pipeline ships in the repository at voxkitchen/templates/pipelines/tts-data-prep.yaml — run it with vkit docker run.