Skip to contents

Pitch

Source: R/pitch-wrapper.R
Pitch.Rd

Fundamental frequency (F0) contour of a sound (a Praat Pitch object).

Arguments

.xptr

Not for direct use. External pointer to the underlying C++ Pitch object; set internally when a method returns a new Pitch.

Value

A Pitch object.

Details

Created from a Sound via autocorrelation or cross-correlation pitch tracking. Supports multiple unit systems: hertz, semitones (re 1 Hz, 100 Hz, or custom), mel, and erb; see the Units section for the full set of codes.

Information

  • get_number_of_frames() - number of analysis frames

  • get_time_step() - time step between frames, in seconds

  • count_voiced_frames() - number of frames with voiced (non-zero) pitch

Point queries

  • get_value_at_time(time, unit, interpolate) - F0 at a time point

  • get_values_at_times(times, unit, interpolate) - F0 at a vector of times (batch)

  • get_strength_at_time(time) - strength (voicing likelihood) at a time

  • get_strengths_at_times(times) - strengths at a vector of times (batch)

Statistics

Computed over a time range.

  • get_mean(from_time, to_time, unit) - mean F0

  • get_standard_deviation(from_time, to_time, unit) - standard deviation

  • get_minimum(from_time, to_time, unit, interpolate) - minimum F0

  • get_maximum(from_time, to_time, unit, interpolate) - maximum F0

  • get_quantile(quantile, from_time, to_time, unit) - quantile of F0

  • get_time_of_minimum(...), get_time_of_maximum(...) - time of extremum

Export

  • as_vector(), as_data_frame() - export as vector or data.frame

  • get_times_vector() - frame time points

  • as_matrix() - F0 values as a matrix (frames by candidates)

Transform

  • to_point_process(voicing_threshold, octave_cost, ...) - convert to a PointProcess (glottal pulses)

  • down_to_pitch_tier() - convert to a PitchTier (editable pitch contour)

Units

F0 unit codes: "hertz" (0), "semitones re 1 Hz" (1), "semitones re 100 Hz" (2), "semitones re 200 Hz" (3), "semitones re 440 Hz" (4), "mel" (5), "log hertz" (6), "erb" (7). Default is "hertz".

See also

Sound, PointProcess, PitchTier

Examples

sound <- Sound$create_tone(duration = 1.0, frequency = 200, sampling_rate =
 44100)
pitch <- sound$to_pitch(time_step = 0.01, pitch_floor = 75, pitch_ceiling =
 600)
mean_f0 <- pitch$get_mean(from_time = 0, to_time = 0, unit = "hertz")
min_f0 <- pitch$get_minimum(from_time = 0, to_time = 0, unit = "hertz")
df <- as.data.frame(pitch)

# The same analysis on a recording read from disk
sound <- Sound(system.file("extdata", "test.wav", package = "pladdrr"))
pitch <- sound$to_pitch(time_step = 0.01, pitch_floor = 75, pitch_ceiling =
 600)
n_voiced <- pitch$count_voiced_frames()