Neural correlates of voice perception
Though speech is recent to evolution, vocalisations prominent in audition well before
- Extracting paralinguistic information important for survival e.g cough=ill
- Little known of voice perception compared to speech perception
- Neuroimaging suggests processing happens in dissociated functional pathways
Ontogenesis and phylogenesis
- Voice perception abilities in other species
o Northern fur seal – mother and pup can recognise each other’s vocalisations in
breeding season in large colony and retain this for 4 years
Variability in voice quality
- Sourced in larynx, filtered by supralaryngeal vocal tract
- Vocal folds periodically open and close in larynx producing buzzing sound with
waveform, determining the f0
- Fry mode – lowest f0, growling sound
- Falsetto mode – thinner, higher sounds
- Formants – enhanced frequencies, depend on size and shape (configuration of
articulators) of vocal tract
- Men – lower f0 and formants
- Women – can use vocal folds in more open configuration leading to higher ratio low
to high frequencies
‘auditory face’
- Voices have small inter and intra individual variations helping carry information of:
speech, identity and affective
- Identity
o Static features e.g timbre is influenced by physical factors (age)
o Dynamic information e.g patterns of pronunciation due to region (accent)
- Affective
o Modification of acoustic limits due to autonomic influence and specific patterns of
muscular contraction
o Information on emotional and motivational state
o emotion prosody = amplitude, pause duration, f0 and f0 variation influence affect
o non-speech interjections = laughs, cries, screams (equivalent to facial expressions)
- formants directly linked to body of speaker, providing information
Bruce & young’s model of face perception
- framework for perceptual and cognitive processes in voice perception
- low level analysis in subcortical nuclei and primary auditory cortex, then structural
encoding involving bilateral regions of STS
- Dissociated in 3 systems:
o speech information analysis in anterior and posterior STS and inferior prefrontal
regions mainly in left
Though speech is recent to evolution, vocalisations prominent in audition well before
- Extracting paralinguistic information important for survival e.g cough=ill
- Little known of voice perception compared to speech perception
- Neuroimaging suggests processing happens in dissociated functional pathways
Ontogenesis and phylogenesis
- Voice perception abilities in other species
o Northern fur seal – mother and pup can recognise each other’s vocalisations in
breeding season in large colony and retain this for 4 years
Variability in voice quality
- Sourced in larynx, filtered by supralaryngeal vocal tract
- Vocal folds periodically open and close in larynx producing buzzing sound with
waveform, determining the f0
- Fry mode – lowest f0, growling sound
- Falsetto mode – thinner, higher sounds
- Formants – enhanced frequencies, depend on size and shape (configuration of
articulators) of vocal tract
- Men – lower f0 and formants
- Women – can use vocal folds in more open configuration leading to higher ratio low
to high frequencies
‘auditory face’
- Voices have small inter and intra individual variations helping carry information of:
speech, identity and affective
- Identity
o Static features e.g timbre is influenced by physical factors (age)
o Dynamic information e.g patterns of pronunciation due to region (accent)
- Affective
o Modification of acoustic limits due to autonomic influence and specific patterns of
muscular contraction
o Information on emotional and motivational state
o emotion prosody = amplitude, pause duration, f0 and f0 variation influence affect
o non-speech interjections = laughs, cries, screams (equivalent to facial expressions)
- formants directly linked to body of speaker, providing information
Bruce & young’s model of face perception
- framework for perceptual and cognitive processes in voice perception
- low level analysis in subcortical nuclei and primary auditory cortex, then structural
encoding involving bilateral regions of STS
- Dissociated in 3 systems:
o speech information analysis in anterior and posterior STS and inferior prefrontal
regions mainly in left
