Research Materials

This is a space where reviewers and readers can view and listen to examples of our experimental materials.
Spectral detail and intonation contour in speech intelligibility (under review)

In this study, we manipulated both the spectrotemporal complexity of spoken sentences (through vocoding) and the intonation profile of the sentences (by artificial flattening the fundamental frequency contour). Pilot testing showed that the sentences were all partially intelligible, but participants were better able to repeat back sentences with higher numbers of vocoded channels and natural intonation profiles.

 

An additional set of unintelligible, spectrally-rotated conditions allowed us to control for spectro-temporal complexity. Examples of all 10 conditions in the fMRI experiment are given below, for the BKB sentence "The clown had a funny face".

N2 - Normal intonation, 2 vocoded channels

F2 - Falling intonation, 2 vocoded channels

N4 - Normal intonation, 4 vocoded channels

F4 - Falling intonation, 4 vocoded channels

N6 - Normal intonation, 6 vocoded channels

F6 - Falling intonation, 6 vocoded channels

N2_R -Normal intonation, 2 vocoded channels 

plus spectral rotation.

F2_R - Falling intonation, 2 vocoded channels 

plus spectral rotation.

N6_R - Normal intonation, 6 vocoded channels 

plus spectral rotation.

F6_R -Falling intonation, 6 vocoded channels 

plus spectral rotation.

Evoked and Emitted ('Real' and 'Posed') laughter (under review)

 

We have carried out the first behavioural and neuroimaging studies of the perception of emotional authenticity in non-verbal vocalisations. We explored listeners' responses to genuine amusement laughter recorded from people watching funny videos (Evoked), and compared this with their perception of posed laughs that recorded in the absence of humorous stimuli (Emitted). We also 

 

Examples of the laughter we used in the behavioural study (some of which also feature in the functional MRI experiment) can be heard below. The Evoked(amb) condition refers to laughs that were generated while watching funny videos, but which could not be reliably categorized as 'Real' by our participants in the behavioural study.

Real or Posed?

Emitted

Evoked(amb)

Evoked

Emotional vocalisations and indexical speaker characteristics

 

We have run a behavioural study on how the perception of indexical speaker characteristics, such as sex or simple speaker discriminations, are affected by different kinds of vocal signals. Here are some of the stimuli we used to assess how authentic emotional content in spontaneous vocalisations versus volitionally produced vocalisations affects perception.

 

 

Spontaneous Laughter                         Sponteneous Crying                               Volitional Laughter                               Vowels