Evidence-Based Recommendations for Tablet Recordings From the Bridge2AI-Voice Acoustic Experiments.
Shaheen N Awan, Ruth Bahr, Stephanie Watts and 3 others
PMID 39306498WHAT IT FOUND
Tablets with cheap headset microphones produce voice data nearly identical to expensive research equipment.
Holding the tablet at arm's length fails, especially in noisy rooms. Use a headset mic at 2.5-5 cm for reliable acoustic analysis.
Key findings
01Tablets paired with low-cost headset microphones showed correlations of 0.90 or higher with research-grade equipment across all acoustic measures.
02Direct tablet recordings at typical distances failed to match the standard, particularly for fundamental frequency, which dropped by approximately 40 Hz in one exam room noise condition.
03Background noise and recording distance significantly altered spectral and cepstral measures like shimmer and HNR, making comparisons between different setups unreliable.
STILL TO COME
How it was doneWhat they foundWhat it means for SLPs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
The study used recorded samples played through a simulator, not live human speech, so real-world articulation and movement artifacts were not tested. Only two specific brands of low-cost headset microphones were evaluated; other devices may behave differently. The study focused on sustained vowels, which may not represent the variability of connected speech.
Declared interests
The text does not explicitly state funding sources or conflicts of interest, but notes the work was part of the Bridge2AI-Voice Consortium, a multi-institutional effort to build a large voice database.
The easy way to misread this
Do not assume all low-cost headset microphones are interchangeable. The study showed that the specific frequency response of different microphones significantly altered spectral measures like L/H ratio, so switching brands mid-study or mid-clinic could invalidate your comparisons.