Sex-related communicative functions of voice spectral energy in human chorusing
Data files
Oct 17, 2023 version files 191.14 KB
-
Preference_data.csv
87.49 KB
-
README.md
7.04 KB
-
Sensitivity_data.csv
94.28 KB
-
Stimulus_Audio_Stats.csv
2.32 KB
Abstract
Music is a human communicative art whose evolutionary origins may lie in capacities that support cooperation and/or competition. A mixed account favoring simultaneous cooperation and competition draws on analogous interactive displays produced by collectively signalling non-human animals (e.g., crickets and frogs). In these displays, rhythmically coordinated calls serve as a beacon whereby groups of males “cooperatively” attract potential female mates, while the likelihood of each male competitively attracting an actual mate depends on the precedence of his signal. Human behaviour consistent with the mixed account was previously observed in a renowned boys choir, where the basses—the oldest boys with the deepest voices—boosted their acoustic prominence by increasing energy in a high-frequency band of the vocal spectrum when girls were in an otherwise male audience. The current study tested female and male sensitivity and preferences for this subtle vocal modulation in online listening tasks. Results indicate that while female and male listeners are similarly sensitive to enhanced high-spectral energy elicited by the presence of female audience members, only female listeners exhibit a reliable preference for it. Findings suggest that human chorusing is a flexible form of social communicative behavior that allows simultaneous group cohesion and sexually motivated competition.
https://doi.org/10.5061/dryad.jh9w0vthn
This document describes the code and data files used for analyses reported in the main manuscript and the Supplementary Material.
The manuscript presents data on human listener sensitivity and preference for sex-related modulations of voice quality in choir singing.
Description of the data and file structure
There are two sets of analyses.
Analysis Set 1 addresses the stimulus audio analyses reported in the Supplementary Material and shown Figure 1 in the main manuscript.
Analysis Set 2 addresses the analyses of online listener ratings for the sensitivity study and the preference study.
Sharing/Access information
Links to other publicly accessible locations of the data:
Code/Software
————————————————————
SET 1: STIMULUS AUDIO ANALYSES
————————————————————
Step 1: Extraction of audio features from stimulus files in Matlab
Step 2: Statistical analysis of audio features in R
————
Step 1: Extraction of audio features from stimulus files in Matlab
Matlab script:
Stimulus_Audio_Analysis_Basic.m
Input:
Audio files in .MP3 format
(Should be located in same folder as Matlab script)
Output:
1. Audio_File_Stats.csv
Contains 4 columns:
- Filename: name of audio file
- Duration: duration of audio file in seconds
- Intensity: total energy (RMS amplitude) of audio file computed with the MIR toolbox: http://bit.ly/mirtoolbox
- SF_Energy: Proportion of spectral energy in the singer’s formant (SF) 2500-3500 Hz region
Audio_File_Stats.csv was used to produce (in Excel) the file Stimulus_Audio_Stats.csv, which serves as input for Stimulus_Audio_Analysis.R (see below).
The rightmost column in Stimulus_Audio_Stats.csv is headed ‘Code’ and includes an identifier that links to our original pool of excerpts (but is otherwise non relevant).
2. Text files containing data for plots of time-averaged spectra (Figures 1A & 1B in manuscript)
[chorale_spec_raw.txt; chorale_spec.txt; fugue_spec_raw.txt; fugue_spec.txt]
3. Figures: Generates Figures 1A & 1B for manuscript
Fugue_SF.png
Chorale_SF.png
————
Step 2: Statistical analysis of audio features in R
R script:
Stimulus_Audio_Analysis.R
Input:
Stimulus_Audio_Stats.csv
Contains 11 columns:
- Filename: Name of audio file, specifying piece (Chorale or Fugue), excerpt number (1-6 for each piece), and condition (Girls Present or Absent)
- Duration: Duration of audio file in seconds
- Intensity: Total energy (RMS amplitude) of audio file computed with the MIR toolbox: http://bit.ly/mirtoolbox
- SF_Energy: Proportion of spectral energy in the singer’s formant (SF) 2500-3500 Hz region
- Dur_Effect: Difference in duration for excerpts sung with girls present vs absent (Present - Absent)
- Int_Effect: Difference in total energy (RMS amplitude) for excerpts sung with girls present vs absent (Present - Absent)
- SF_Effect: Difference in SF spectral energy for excerpts sung with girls present vs absent (Present - Absent)
- Piece: Chorale or Fugue
- Excerpt: Excerpt number 1-6 for each piece
- Girls: Condition, i.e., girls Present vs Absent
- Code: Original code for excerpt in author PEK’s database (included for cross-referencing only)
Output:
1. Summary tables of ANOVAs for Duration\, Intensity\, and SF_Energy data\, reported in Supplementary Material
2. Plot for Figure 1C in manuscript
<br>
—————————————————————————
SECTION 2: ONLINE LISTENER RATINGS ANALYSES
—————————————————————————
Data files for sensitivity study & preference study
Data for each study are in a separate .CSV file:
Sensitivity_data.csv
Preference_data.csv
Column headings and contents
- Participant_ID_number: Non-consecutive because some participants did not meet criteria for successful task completion
- Age_years: Possible range 12-90
- Sex: Female, Male
- Musical_training: Musician, Non-musician
- Choir_experience: Yes, No
- Cultural_background: European, Non-European, Mixed
- Chorale_%SF_selected: Percentage of the 6 Chorale stimulus items in which excerpt with an enhanced singer’s formant was selected
- Fugue_%SF_selected: Percentage of 6 Fugue stimulus items in which excerpt with an enhanced singer’s formant was selected
- Total_%SF_selected: Percentage of 12 total stimulus items in which excerpt with an enhanced singer’s formant was selected
- Chorale_item_1_SF_selected: Excerpt with an enhanced singer’s formant was (1) or was not (0) selected for Chorale excerpt 1
- Chorale_item_2_SF_selected: Excerpt with an enhanced singer’s formant was (1) or was not (0) selected for Chorale excerpt 2
- Chorale_item_3_SF_selected: Excerpt with an enhanced singer’s formant was (1) or was not (0) selected for Chorale excerpt 3
- Chorale_item_4_SF_selected: Excerpt with an enhanced singer’s formant was (1) or was not (0) selected for Chorale excerpt 4
- Chorale_item_5_SF_selected: Excerpt with an enhanced singer’s formant was (1) or was not (0) selected for Chorale excerpt 5
- Chorale_item_6_SF_selected: Excerpt with an enhanced singer’s formant was (1) or was not (0) selected for Chorale excerpt 6
- Fugue_item_1_SF_selected: Excerpt with an enhanced singer’s formant was (1) or was not (0) selected for Fugue excerpt 1
- Fugue_item_2_SF_selected: Excerpt with an enhanced singer’s formant was (1) or was not (0) selected for Fugue excerpt 2
- Fugue_item_3_SF_selected: Excerpt with an enhanced singer’s formant was (1) or was not (0) selected for Fugue excerpt 3
- Fugue_item_4_SF_selected: Excerpt with an enhanced singer’s formant was (1) or was not (0) selected for Fugue excerpt 4
- Fugue_item_5_SF_selected: Excerpt with an enhanced singer’s formant was (1) or was not (0) selected for Fugue excerpt 5
- Fugue_item_6_SF_selected: Excerpt with an enhanced singer’s formant was (1) or was not (0) selected for Fugue excerpt 6
————
Statistical analysis of sensitivity & preference data
Contents are separate data files (.CSV) and R scripts for analysis for the Sensitivity study and the Preference study.
Data files and R scripts should be in the same directory.
The path to this directory needs to be specified in each R script.
Output includes statistical values and plots reported in the main manuscript and Supplementary Material.
Sensitivity data analysis files
Data file: Sensitivity_data.csv
R script for analysis: Sensitivity_analysis_script.R
Preference data analysis files
Data file: Preference_data.csv
R script for analysis: Preference_analysis_script.R