Related Experiment Videos
Acoustic analysis of pathological voices compressed with MPEG system
Julio Gonzalez1, Teresa Cervera, M José Llau
1Department of Basic and Clinical Psychology and Psychobiology, Universitat Jaume I, Castellon, Spain. gonzalez@psb.uji.es
Summary
MP3 audio compression maintains high voice quality for dysphonic individuals at high bitrates (96-128 kbps). Lower bitrates significantly alter acoustic properties, impacting voice analysis.
Area of Science:
- Acoustics
- Speech Science
- Digital Signal Processing
Background:
- MPEG-1 Layer 3 (MP3) compression offers high audio quality at reduced file sizes.
- Previous research indicated MP3 fidelity for normal voices.
- The acoustic impact of MP3 compression on dysphonic voices remained largely unexplored.
Purpose of the Study:
- To investigate the acoustic effects of MP3 compression on dysphonic voice signals.
- To determine the fidelity of compressed dysphonic voices across different bitrates.
Main Methods:
- Comparative analysis of acoustic parameters (spectrograms, LTAS, FFT spectra) for running speech.
- Evaluation of voice parameters (intensity, formants, multidimensional measures) in sustained phonations.
- Testing MP3 compression at various bitrates (96, 128 kbps, and lower).
Main Results:
- High-bitrate MP3 compression (96 and 128 kbps) preserved key acoustic properties of dysphonic voices.
- Lower compression bitrates led to decreased fidelity and significant acoustic alterations.
- Consistent findings with prior research on normal voices.
Conclusions:
- High-bitrate MP3 compression is suitable for storing and transmitting dysphonic voice recordings without significant acoustic distortion.
- Careful consideration of bitrate is crucial to avoid altering pathological voice characteristics.