Skip to contents

Formant frequencies and bandwidths via Praat's Burg method

trk_formant_burg.Rd

Tracks formant frequencies (F1–F5), bandwidths (B1–B5), and optionally spectral intensities (L1–L5) using Praat's Burg LPC algorithm via pladdrr. Optional HMM-based formant tracking smooths trajectories across time. Prefer this over trk_formant_forest when Praat-compatible Burg estimates are needed.

Usage

trk_formant_burg(listOfFiles, beginTime = 0, endTime = 0, timeStep = 0.005, number_of_formants = 5, maxHzFormant = 5500, windowLength = 0.025, pre_emphasis = 50, track_formants = FALSE, number_of_tracks = 3, reference_F1 = 550, reference_F2 = 1650, reference_F3 = 2750, reference_F4 = 3850, reference_F5 = 4950, frequency_cost = 1, bandwidth_cost = 1, transition_cost = 1, windowShape = "Gaussian1", relativeWidth = 1, include_intensity = TRUE, spectrogram_resolution = 40, toFile = TRUE, explicitExt = "pfm", outputDirectory = NULL, verbose = TRUE)

Arguments

listOfFiles

Character vector of audio file paths. Any format supported by av is accepted; non-native inputs are transcoded automatically.

timeStep

Numeric. Frame shift in seconds; sets output frame rate (1 / timeStep Hz). Default 0.005 s (200 Hz).

number_of_formants

Integer. Number of formant candidates to find (up to 5). Default 5.

maxHzFormant

Numeric. Formant ceiling in Hz; typically 5500 for female, 5000 for male speakers. Default 5500 Hz.

windowLength

Numeric. LPC analysis window length in seconds. Default 0.025 s.

pre_emphasis

Numeric. Pre-emphasis onset frequency in Hz. Default 50 Hz.

track_formants

Logical. Apply HMM-based formant tracking to smooth trajectories. Default FALSE.

number_of_tracks

Integer. Number of tracks to retain when track_formants = TRUE. Default 3.

reference_F1

Numeric. Reference F1 frequency for HMM tracking in Hz. Default 550 Hz.

reference_F2

Numeric. Reference F2 frequency for HMM tracking in Hz. Default 1650 Hz.

reference_F3

Numeric. Reference F3 frequency for HMM tracking in Hz. Default 2750 Hz.

reference_F4

Numeric. Reference F4 frequency for HMM tracking in Hz. Default 3850 Hz.

reference_F5

Numeric. Reference F5 frequency for HMM tracking in Hz. Default 4950 Hz.

frequency_cost

Numeric. HMM cost weight for deviation from reference frequencies. Default 1.0.

bandwidth_cost

Numeric. HMM cost weight for formant bandwidth. Default 1.0.

transition_cost

Numeric. HMM cost weight for frame-to-frame transitions. Default 1.0.

windowShape

Character. Window shape applied before analysis. Default "Gaussian1".

relativeWidth

Numeric. Relative width of the analysis window. Default 1.0.

include_intensity

Logical. Extract spectral intensity at each formant frequency (L1–L5 tracks via spectrogram lookup). Default TRUE.

spectrogram_resolution

Numeric. Frequency resolution of the spectrogram used for intensity extraction in Hz. Default 40 Hz.

toFile

Logical. If TRUE, write SSFF output files and return the count written (invisibly). If FALSE, return an AsspDataObj. Default TRUE.

explicitExt

Character. Output file extension. Default "pfm".

beginTime

Start time for the extracted portion in seconds. Default: NULL (beginning of signal). Note: uses beginTime/endTime (seconds) matching DSP function conventions, unlike read_audio() which uses begin/end.

endTime

The end time of the section of the sound files that should be analysed (in seconds). Use 0 for end of file.

outputDirectory

The directory where the slice file should be stored. If not defiled (NULL), the sparse slice file will placed in the same folder as the media file.

verbose

Logical. Show a progress bar (sequential path) or a progress-aware parallel apply (pbapply/pbmcapply, if installed).

Value

If toFile = FALSE: an AsspDataObj with tracks:

fm1–fm5

REAL32, Hz, n_frames x 1. Formant frequencies F1–F5. 0 encodes an undefined (unvoiced) frame.

bw1–bw5

REAL32, Hz, n_frames x 1. Formant bandwidths B1–B5. 0 encodes an undefined frame.

L1–L5

REAL32, Pa^2/Hz, n_frames x 1. Spectral power at each formant frequency (only when include_intensity = TRUE).

Frame rate: 1 / timeStep Hz (default 200 Hz). If toFile = TRUE: integer count of files written, returned invisibly.

Details

Uses Praat's Burg all-pole LPC estimator. Spectral intensity tracks (L1–L5) are extracted from a separate spectrogram and require pladdrr >= 4.8.20; earlier versions may crash with include_intensity = TRUE. HMM tracking (track_formants = TRUE) may not be available in all pladdrr builds; a warning is issued and untracked formants are returned.

Examples

if (FALSE) { # \dontrun{
# Analyze formants for a single file
formants <- trk_formant_burg("speech.wav", toFile = FALSE)

# Batch process multiple files with tracking
files <- c("speech1.wav", "speech2.wav")
trk_formant_burg(files, track_formants = TRUE, toFile = TRUE)

# With custom formant ceiling (e.g., male speaker)
formants <- trk_formant_burg("speech.wav",
                         maxHzFormant = 5000,
                         number_of_formants = 4,
                         toFile = FALSE)
} # }