soundgen (version 2.3.0)

bandpass: Bandpass/stop filters

Description

Filtering in the frequency domain with FFT-iFFT: low-pass, high-pass, bandpass, and bandstop filters. Similar to ffilter, but here we use FFT instead of STFT - that is, the entire sound is processed at once. This works best for relatively short sounds (seconds), but gives us maximum precision (e.g., for precise notch filtering) and doesn't affect the attack and decay. NAs are accepted and can be interpolated or preserved in the output. Because we don't do STFT, arbitrarily short vectors are also fine as input - for example, we can apply a low-pass filter prior to decimation when changing the sampling rate without aliasing. Note that, unlike pitchSmoothPraat, bandpass applies an abrupt cutoff instead of a smooth gaussian filter.

Usage

bandpass(
  x,
  samplingRate = NULL,
  lwr = NULL,
  upr = NULL,
  action = c("pass", "stop")[1],
  na.rm = TRUE,
  from = NULL,
  to = NULL,
  normalize = FALSE,
  reportEvery = NULL,
  saveAudio = NULL,
  plot = FALSE,
  savePlots = NULL,
  width = 900,
  height = 500,
  units = "px",
  res = NA,
  ...
)

Arguments

x

path to a folder, one or more wav or mp3 files c('file1.wav', 'file2.mp3'), Wave object, numeric vector, or a list of Wave objects or numeric vectors

samplingRate

sampling rate of x (only needed if x is a numeric vector)

lwr, upr

cutoff frequencies, Hz. Specifying just lwr gives a low-pass filter, just upr high-pass filter with action = 'pass' (or vice versa with action = 'stop'). Specifying both lwr and upr a bandpass/bandstop filter, depending on 'action'

action

"pass" = preserve the selected frequency range (bandpass), "stop" = remove the selected frequency range (bandstop)

na.rm

if TRUE, NAs are interpolated, otherwise they are preserved in the output

from

if NULL (default), analyzes the whole sound, otherwise from...to (s)

to

if NULL (default), analyzes the whole sound, otherwise from...to (s)

normalize

if TRUE, resets the output to the original scale (otherwise filtering often reduces the amplitude)

reportEvery

when processing multiple inputs, report estimated time left every ... iterations (NULL = default, NA = don't report)

saveAudio

full path to the folder in which to save audio files (one per detected syllable)

plot

should a spectrogram be plotted? TRUE / FALSE

savePlots

full path to the folder in which to save the plots (NULL = don't save, '' = same folder as audio)

width

graphical parameters for saving plots passed to png

height

graphical parameters for saving plots passed to png

units

graphical parameters for saving plots passed to png

res

graphical parameters for saving plots passed to png

...

other graphical parameters

Details

Algorithm: fill in NAs with constant interpolation at the edges and linear interpolation in the middle; perform FFT; set the frequency ranges to be filtered out to 0; perform inverse FFT; set to the original scale; put the NAs back in.

Examples

# NOT RUN {
# Filter white noise
s1 = fade(c(runif(2000, -1, 1)))
bandpass(s1, 16000, upr = 2000, plot = TRUE)    # low-pass
bandpass(s1, 16000, lwr = 2000, plot = TRUE)    # high-pass
bandpass(s1, 16000, lwr = 1000, upr = 1100, action = 'stop', plot = TRUE) # bandstop
s2 = bandpass(s1, 16000, lwr = 2000, upr = 2100, plot = TRUE) # bandpass
# playme(rep(s2, 5))
spectrogram(s2, 16000)  # more accurate than plotting the spectrum with plot = TRUE

# a short vector with some NAs
x = rnorm(150, 10) + 3 * sin((1:50) / 5)
x[sample(1:length(x), 50)] = NA
plot(x, type = 'l')
points(bandpass(x, samplingRate = 100, upr = 10), type = 'l', col = 'blue')

# }
# NOT RUN {
# precise notch filtering is possible, even in low frequencies
whiteNoise = runif(8000, -1, 1)
s3 = bandpass(whiteNoise, 16000, lwr = 30, upr = 40,
              plot = TRUE, xlim = c(0, 500))
playme(rep(s3, 5))
spectrogram(s3, 16000, windowLength = 150, yScale = 'log')

# compare the same with STFT
s4 = seewave::ffilter(whiteNoise, f = 16000, from = 30, to = 40)
spectrogram(s4, 16000, windowLength = 150, yScale = 'log')
# (note: works better as wl approaches length(s4))

# high-pass all audio files in a folder
bandpass('~/Downloads/temp', saveAudio = '~/Downloads/temp/hp2000/',
         lwr = 2000, savePlots = '~/Downloads/temp/hp2000/')
# }