Audio AI GUIDE
PESQ and STOI Speech Quality Metrics
PESQ and STOI are reference-based objective measures used to estimate different aspects of processed speech: perceived quality and intelligibility, respectively.
Listen to this guideAI-narrated · ~2 min
On this page2 min read
Overview
They support controlled comparisons but do not replace listening tests or guarantee a listener’s experience. PESQ’s ITU-T P.862 recommendation is superseded, so benchmark reports should state the version and use case.
Deep Dive
PESQ (Perceptual Evaluation of Speech Quality) was specified in ITU-T Recommendation P.862 as an objective method for predicting subjective quality of narrow-band telephone speech and codecs. It compares a degraded signal with a clean reference and produces a raw score on an approximately -0.5 to 4.5 scale; this is MOS-like, not a direct listener rating. The ITU lists P.862 as superseded and deleted, so current work should choose an applicable successor or standard. STOI (Short-Time Objective Intelligibility), introduced by Taal and colleagues, estimates speech intelligibility by comparing short-time temporal envelopes of clean and degraded speech across frequency bands. Its commonly reported value ranges from 0 to 1. Both are intrusive reference-based measures, and their scores summarize specific signal conditions rather than guaranteeing human judgments.
Real-World Implementation
Benchmarking speech-enhancement and noise-suppression models on standard test sets
Comparing telephone and VoIP codec quality during network engineering
Tuning hearing-aid and cochlear-implant processing for maximum intelligibility
Validating dereverberation algorithms in conferencing and voice-assistant pipelines
The Future of PESQ and STOI Speech Quality Metrics
Reference-free measures such as DNSMOS and NISQA can score a processed signal without access to a clean reference, which helps when paired recordings are unavailable. They introduce their own model and domain limitations, so objective scores are best interpreted alongside listening tests and task-specific evaluations. PESQ remains historically important, but its P.862 recommendation status and intended narrow-band use should be checked before selecting it.
Putting it into practice
Why it matters
- Access and reach. It improves accessibility through transcription, narration, and voice interfaces.
- Cost and budget. Media teams can ship polished audio faster with smaller budgets.
- Speed and scale. Customer-facing systems can process spoken interactions at larger scale.
Watch out for
- Voice misuse and impersonation risks increase when consent is missing.
- Accuracy can drop across accents, dialects, or noisy environments.
- Synthetic audio can be mistaken for authentic speech without clear labeling.
Next steps
- Obtain explicit consent for voice capture, cloning, and reuse.
- Test quality across diverse speakers and background conditions.
- Define when a human must review or approve outputs.
- Label synthetic audio and keep provenance records for accountability.
Keep Exploring
Free newsletter
Get the daily AI briefing
Three verified AI stories every weekday morning, written in plain English. Free forever, no ads.
One email each weekday. Unsubscribe in one click. We never sell or share your address.
Test yourself
Take the PESQ and STOI Speech Quality Metrics quiz
Instant feedback on every answer, and a shareable certificate with a verifiable ID once you pass a course.
Support free AI education. AI Understanding is a 501(c)(3) nonprofit — no ads, no paywall, ever. Make a donation
Frequently asked questions
What is PESQ and STOI Speech Quality Metrics?
PESQ and STOI are reference-based objective measures used to estimate different aspects of processed speech: perceived quality and intelligibility, respectively. They support controlled comparisons but do not replace listening tests or guarantee a listener’s experience. PESQ’s ITU-T P.862 recommendation is superseded, so benchmark reports should state the version and use case.
What does PESQ estimate when comparing a degraded speech signal with a reference?
PESQ was designed as an objective estimate of perceived speech quality for specified telephony and codec conditions; it does not transcribe words.
What does STOI primarily predict?
STOI estimates intelligibility, i.e. how understandable the speech is, on a 0-to-1 scale.
Both PESQ and STOI are described as 'intrusive' metrics. What does that mean?
Intrusive metrics compare the degraded signal against a clean reference to compute a score.
What is the commonly reported output range for STOI?
STOI is commonly reported on a 0-to-1 scale, with higher scores indicating greater estimated intelligibility under the evaluation setup.
PESQ models human hearing using which kind of perceptual frequency scale?
PESQ maps signals onto a psychoacoustic loudness representation using Bark-band processing.
Keep learning
Related guides
More guides picked for this topic