OPUS to SPH Converter
Produce SPHERE speech research audio from OPUS
Speech Corpus Format
SPH is the standard behind major speech datasets — convert OPUS recordings into research-ready audio.
Dataset Preparation
Process entire OPUS collections to SPH at once — prepare corpora in one operation.
Online Conversion
No speech toolkit needed — produce SPH from OPUS directly in your browser.
How to convert OPUS to SPH
Select files from Computer, Google Drive, Dropbox, URL or by dragging it on the page.
Choose sph or any other format you need as a result (more than 200 formats supported)
Let the file convert and you can download your sph file right afterwards
About formats
Frequently Asked Questions
SPH (SPHERE) is the NIST-defined standard for speech research corpora. ASR pipelines and linguistic tools expect SPHERE input.
Kaldi, HTK, NIST evaluation tools, and academic speech datasets like TIMIT use SPHERE as their audio format.
Yes — SPH and NIST both refer to SPHERE (SPeech HEader Resources) defined by the National Institute of Standards and Technology.
Speech corpora typically use 8 or 16 kHz — the converter resamples from OPUS automatically.
Upload an entire OPUS speech collection and produce SPH for every file — ready for research.