Journal article · AI & technology← The news desk

✦ The Dispatch

The voice clone intelligibility benefit in noise in middle-aged listeners

A dispatch from PubMed — filed

Research with younger adults showed that cloned voices are more intelligible than human voices in noise, with a benefit of 13.4%. This study tested whether this benefit extends to 40 middle-aged listeners (45-65 years), as this population may show emerging difficulties with speech-in-noise. Participants recognised sentences by ten human voices and ten voice clones in four noise levels....

Continue reading at PubMed

✦ The floor

Discussion

Signed responses from readers of the wire.

Clinical Takeaway

Findings are preliminary and study-specific; no actionable change to fitting or counseling practice is warranted until replication with larger, clinically representative samples.

Why It Matters

If voice-clone intelligibility benefits extend across age groups, personalized AI voices could become a practical tool for improving communication in noise for hearing aid and assistive listening device users.

Key Points
  1. 01AI voice clones previously showed a ~13.4% intelligibility-in-noise benefit in younger adults.
  2. 02This JASA Express Letters study extends the investigation to middle-aged listeners.
  3. 03Personalized voice familiarity may reduce the cognitive effort needed to understand speech in noise.
  4. 04Results will inform whether voice-clone benefits are age-dependent or broadly applicable.
  5. 05Published in JASA Express Letters, a peer-reviewed acoustics journal.
Claims & Evidence

AI-generated voice clones produced a 13.4% intelligibility benefit in noise in younger adults in prior research.

studypartially supported

Voice clones may offer an intelligibility benefit in noise for middle-aged listeners.

studyunclear
Research metadata
PMID
42678253
DOI
10.1121/10.0046178.
Journal
JASA Express Letters
Publication type
research_article
Evidence level
2b
Population
Middle-aged listeners with typical or near-typical hearing
Intervention
AI-generated voice clones presented in noise
Comparator
Non-cloned or unfamiliar voices in noise

Primary outcomes

Speech intelligibility in noise for voice-cloned vs. non-cloned stimuli

Related stories