stefantaubert/mel-cepstral-distance

A Python library for computing the Mel-Cepstral Distance (Mel-Cepstral Distortion, MCD) between two inputs. This implementation is based on the method proposed by Robert F. Kubichek in "Mel-Cepstral Distance Measure for Objective Speech Quality Assessment".

43
/ 100
Emerging

This tool helps researchers and practitioners in speech technology objectively compare the quality of two speech audio inputs, like a reference recording and a synthesized voice. It calculates the Mel-Cepstral Distance (MCD), a metric indicating how different two sounds are, and can also provide a penalty for misaligned audio segments. Speech synthesis developers, voice conversion researchers, and anyone evaluating speech generation models would use this.

No commits in the last 6 months.

Use this if you need to quantify the perceptual difference between a generated speech audio and a reference speech audio, or compare different speech synthesis models.

Not ideal if you are looking for subjective human-perception based speech quality assessment or a tool for general audio comparison outside of speech.

speech-synthesis voice-conversion audio-quality-assessment speech-evaluation voice-AI
Stale 6m No Package No Dependents
Maintenance 2 / 25
Adoption 8 / 25
Maturity 16 / 25
Community 17 / 25

How are scores calculated?

Stars

65

Forks

12

Language

Python

License

MIT

Last pushed

Aug 24, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/voice-ai/stefantaubert/mel-cepstral-distance"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.