Vocal Tremor Analysis Using pladdrr

Analyzes vocal tremor from sustained vowel recordings using pladdrr's Praat bindings. Extracts 18 measures of frequency and amplitude tremor based on Brückl (2012) autocorrelation algorithm.

Usage

lst_voice_tremor(listOfFiles, beginTime = 0, endTime = 0, analysisTimeStep = 0.015, minPitch = 60, maxPitch = 350, silenceThreshold = 0.03, voicingThreshold = 0.3, octaveCost = 0.01, octaveJumpCost = 0.35, voicedUnvoicedCost = 0.14, minTremorFreq = 1.5, maxTremorFreq = 15, tremorMagThresh = 0.01, tremorCyclicalThresh = 0.15, freqTremorOctaveCost = 0.01, ampTremorOctaveCost = 0.01, nanAsZero = FALSE, toFile = FALSE, return_jstf = FALSE, explicitExt = "pvt", outputDirectory = NULL, verbose = TRUE)

Arguments

listOfFiles

Character vector with path(s) to audio file(s)

beginTime

Numeric. Start time in seconds (default 0)

endTime

Numeric. End time in seconds (0 = end of file)

analysisTimeStep

Numeric. Time step for analysis in seconds (default 0.015)

minPitch

Numeric. Minimum pitch for extraction in Hz (default 60)

maxPitch

Numeric. Maximum pitch for extraction in Hz (default 350)

silenceThreshold

Numeric. Threshold for silence detection (default 0.03)

voicingThreshold

Numeric. Threshold for voicing detection (default 0.3)

octaveCost

Numeric. Cost for octave jumps in pitch tracking (default 0.01)

octaveJumpCost

Numeric. Cost for large octave jumps (default 0.35)

voicedUnvoicedCost

Numeric. Cost for voiced/unvoiced transitions (default 0.14)

minTremorFreq

Numeric. Minimum tremor frequency in Hz (default 1.5)

maxTremorFreq

Numeric. Maximum tremor frequency in Hz (default 15)

tremorMagThresh

Numeric. Threshold for contour magnitude (default 0.01)

tremorCyclicalThresh

Numeric. Threshold for cyclicality (default 0.15)

freqTremorOctaveCost

Numeric. Octave cost for frequency tremor (default 0.01)

ampTremorOctaveCost

Numeric. Octave cost for amplitude tremor (default 0.01)

nanAsZero

Logical. Convert undefined measurements to zeros (default FALSE)

toFile

Logical. If TRUE, write results to JSTF file. Default FALSE.

explicitExt

Character. File extension for output. Default "pvt".

outputDirectory

Character. Output directory path. Default NULL (use input directory).

verbose

Logical. Print progress messages (default TRUE)

return_jstf

Logical. Return JsonTrackObj instead of data.frame? Default FALSE. When both toFile and return_jstf are TRUE, the file is written AND the object returned.

Value

If toFile=FALSE (default), a data.frame (or list of data.frames for multiple files) with 18 tremor measurements. If toFile=TRUE, invisibly returns the path(s) to the written JSTF file(s).

Each result contains 18 columns:

FCoM

Frequency contour magnitude

FTrC

Frequency tremor cyclicality (0-1)

FMoN

Number of frequency modulation candidates

FTrF

Frequency tremor frequency (Hz)

FTrI

Frequency tremor intensity index (percent)

FTrP

Frequency tremor power index

FTrCIP

Frequency tremor cyclicality-intensity product

FTrPS

Frequency tremor product sum

FCoHNR

Frequency contour HNR (dB)

ACoM

Amplitude contour magnitude

ATrC

Amplitude tremor cyclicality (0-1)

AMoN

Number of amplitude modulation candidates

ATrF

Amplitude tremor frequency (Hz)

ATrI

Amplitude tremor intensity index (percent)

ATrP

Amplitude tremor power index

ATrCIP

Amplitude tremor cyclicality-intensity product

ATrPS

Amplitude tremor product sum

ACoHNR

Amplitude contour HNR (dB)

Details

This function processes sustained phonations to detect tremor characteristics in both pitch (frequency) and intensity (amplitude) contours. It applies Gaussian1 windowing and uses autocorrelation-based analysis to identify tremor frequency, intensity, and cyclicality.

References

(Brückl 2012)

Examples

if (FALSE) { # \dontrun{
# Analyze sustained vowel
result <- lst_voice_tremor("sustained_vowel.wav")
print(result$FTrF)  # Frequency tremor frequency
print(result$FTrI)  # Frequency tremor intensity

# Write to JSTF file
lst_voice_tremor("sustained_vowel.wav", toFile = TRUE)
track <- read_track("sustained_vowel.pvt")
df <- as.data.frame(track)
} # }