Perform Digital Filtering

Description

Finite Impulse Response (FIR) filtering of vectors.

Usage

FIRfilter (sound, from = 0, to = fs/2, fs = 22050, order = 200, 
verify = FALSE, impulse = NULL, pad = TRUE)

Arguments

Details

This function performs lowpass, highpass and bandpass filtering using a windowed-sinc filter. Increasing the filter order decreases the transition region between the passband and the stopband. The magnitude of frequencies in the stopband is usually attenuated by about about 45 dB.

If verify is TRUE, a plot is created which allows the user to inspect the performance of the function.

Value

If a vector is given, the filtered vector is returned.

If a 'sound' object is given, a sound object containing the filtered sound is returned.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

http://en.wikipedia.org/wiki/Sinc_filter

Lyons, R. G. (2004). Understanding Digital Signal Processing (2nd ed.). Prentice Hall.

Examples

## generate random noise
#sound = rnorm (5000, 0, 100)

## implement low, high and band-pass filters
#par (mfrow = c(3,1), mar = c(4,4,1,1))
#snd1 = FIRfilter (sound, from = 400, fs = 1000, verify = T)
#snd2 = FIRfilter (sound, to = 400, fs = 1000, verify = T)
#snd3 = FIRfilter (sound, from = 400, to = 100, fs = 1000, verify = T)

## use filters of different orders (i.e., differing number of taps)
#par (mfrow = c(2,1), mar = c(4,4,1,1))
#snd1 = FIRfilter (sound, to = 400, fs = 1000, order = 50, verify = T)
## higher order filters lead to narrower transition regions
#snd2 = FIRfilter (sound, to = 400, fs = 1000, order = 2000, verify = T)

Formant Filtering

Description

Perform cascade formant filtering of sounds.

Usage

Ffilter (sound, ffs, bwp = 0.06, minbw = 60, fs = 22050, verify = FALSE)

Arguments

Details

This function allows the user to specify one or more formant filters and to pass a signal through said filters. This may be used to create synthetic speech sounds or to modify existing sounds as desired. The given signal is passed through the formant filters in reverse order. Filter bandwidths specify the distance between formant center frequencies and the point at which the output will be 3 dB below peak energy.

Filter bandwidths may be provided in Hz, or as a percentage of the formant frequencies. To set these as a percent of formant frequencies, all values must be less than 1. If these are not provided they are set to 6 percent of the formant center frequencies by default. If only one value is provided, this is assumed to be the desired value for all formants.

Value

A vector representing the filtered sound.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

Klatt, D. H. (1980). Software for a cascade/parallel formant synthesizer. Journal of the Acoustical Society of America 67(3): 971-995.

http://www.fon.hum.uva.nl/praat/manual/Sound__Filter__one_formant____.html

Examples

## uncomment and run
## Generate half a second of white noise
#sound = rnorm (11025, 0, 1)

## pass this through some formant filters
## two static formants
#filtered1 = Ffilter (sound, ffs = c(4000,7000), bw = 500)
## a single moving formant
#filtered2 = Ffilter (sound, ffs = list(3000,8000), bw = 500)
## two moving formants


## inspect the results
#par (mfrow = c(1,2), mar = c(4,4,1,1))
#spectrogram (filtered1, maxfreq = 11025)
#spectrogram (filtered2, maxfreq = 11025)

Probabilistic Sliding Template Model

Description

Classify Vowels Using the Probabilistic Sliding Template Model

Usage

PSTM (ffs, f0, template, winner = TRUE)

Arguments

Details

The classic log-mean normalization method of Nearey (1978) helps compare the vowels produced by different speakers by controlling for the log-mean formant frequency (FF) produced by a speaker. This approach to normalization assumes that variation in the vowel spaces of speakers of the same dialect is primarily according to a single multiplicative parameter. When this speaker-specific scaling parameter is controlled for, differences in the vowel spaces of different speakers are minimized.

The Probabilistic Sliding Template Model (PSTM) of Nearey and Assmann (2007) attempts to predict perceived vowel quality by 'guessing' an appropriate speaker-specific scaling parameter and normalizing vowels using this estimated parameter. 'Method 6' of the PSTM (described in Nearey & Assmann, 2007) is used to estimate the necessary parameter. After normalization, vowels are classified by comparing them to a provided reference template, which can be created using the createtemplate() function included in this package. Normalized or unnormalized vowels may be classified, as long as the same transformations are performed on the data used to create the template and the data being classified.

If no template is passed, the model identifies vowels relative to the vowel system of Edmonton English speakers. For in-depth details regarding the specifics of this model, please see Nearey & Assmann (2007).

Value

If winner = TRUE:

A dataframe with the following columns:

If winner = FALSE:

A list of dataframes, each of which contains information for every candidate vowel category for each token to be classified. Each dataframe has the following columns:

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

Nearey, T. M. (1978). Phonetic Feature Systems for Vowels. PhD thesis, Indiana University Linguistics Club.

Nearey, T. M. & P. F. Assmann. (2007). Pobabilistic 'sliding template' models for indirect vowel normalization. in Experimental Approaches to Phonology, edited by M.J. Sole, P. S., Beddor, and M. Ohala (Oxford University Press, Oxford), pp. 246-269.

Examples

## load Peterson & Barney vowel data
#data (pb52)

## normalize vowel formant frequencies
#normdvowels = normalize (pb52[,7:9], pb52$speaker, pb52$vowel)
#formants = normdvowels[,1:3]
#vowels = pb52$vowel

## make a vowel template based on these frequencies
#template = createtemplate (formants, vowels) 

## first classify only the first three vowels
#ffs = pb52[1:3,c(7:9)] 
#f0 = pb52[1:3,6]

## outputting only the winners, and then the full posterior probabilities
#PSTM (ffs, f0, template)
#PSTM (ffs, f0, template, winner = FALSE)

## now classify all vowels
## uncomment to run
#ffs = pb52[,c(7:9)] 
#f0 = pb52[,6]  
#winner = PSTM (ffs, f0, template)
## with a good degree of accuracy
#table (winner$vowel, pb52$vowel)

Aronson et al. (1996) Hebrew Vowel Data

Description

Formant frequency information for vowels averaged across 6 male speakers.

Usage

data (a96)

Format

A data frame with the following columns:

sex

- A factor indicating speaker sex.

vowel

- The vowel category in x-sampa.

f1

- A numeric vector indcating the vowel F1 in Hz.

f2

- A numeric vector indcating the vowel F2 in Hz.

f3

- A numeric vector indcating the vowel F3 in Hz.

f4

- A numeric vector indcating the vowel F4 in Hz.

References

Aronson, L., Rosenhouse, J. Rosenhouse, G. & Podoshin, L. (1996). An acoustic analysis of modern Hebrew vowels and voiced consonants. Journal of Phonetics 24. 283-193.

Examples

#data(a96)
#vowelplot (a96$f1, a96$f2, a96$vowel, logaxes = 'xy', xsampa = TRUE)

Bradlow (1995) Spanish Vowel Data

Description

Formant frequency information for vowels averaged across 4 male speakers.

Usage

data (b95)

Format

A data frame with the following columns:

sex

- A factor indicating speaker sex.

vowel

- The vowel category in x-sampa.

f1

- A numeric vector indcating the vowel F1 in Hz.

f2

- A numeric vector indcating the vowel F2 in Hz.

References

Bradlow, A. R. (1995). A comparative acoustic study of English and Spanish vowels. Journal of the Acoustical Society of America 97. 1916-1924.

Examples

#data(b95)
#vowelplot (b95$f1, b95$f2, b95$vowel, logaxes = 'xy', xsampa = TRUE)

Combinations and Permutations

Description

Calculate the number of combinations or permutations for a number of objects.

Usage

combocalc (objects, choose, order = FALSE, repetition = TRUE)

Arguments

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

# combinations, no repetition
combocalc (10, 4, order = FALSE, repetition = FALSE)
# combinations, with repetition
combocalc (10, 4, order = FALSE, repetition = TRUE)
# permutations, no repetition
combocalc (10, 4, order = TRUE, repetition = FALSE)
# permutations, with repetition
combocalc (10, 4, order = TRUE, repetition = TRUE)

Create an LDA Template

Description

Create a linear discriminant analysis template.

Usage

createtemplate (features, classes)

Arguments

Details

This function finds the location of the mean of each class in an n-dimensional space, where each dimension corresponds to one of n columns in 'features'. In addition, the pooled, within-category covariance matrix is found.

The name for each vowel category is stored as the rownames of the 'means' element.

The function plot() is defined for template objects and allows the user to view the location of and expected variation around the different vowel categories in the formant space.

This information may be used in conjunction with the PSTM() function, included in this package. The mean and covariance matrices provided by this function may also be useful for the ldclassify() function provided in this package.

Value

A 'template' object, a list containing the elements:

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

Nearey, T. M. & P. F. Assmann. (2007). Pobabilistic 'sliding template' models for indirect vowel normalization. in Experimental Approaches to Phonology, edited by M.J. Sole, P. S., Beddor, and M. Ohala (Oxford University Press, Oxford), pp. 246-269.

Examples

#data (pb52)            ## load the Peterson and Barney vowels.
## normalize them.
#normdvowels = normalize (pb52[,7:9], pb52$speaker, pb52$vowel)  
#formants = normdvowels[,1:3]
#vowels = pb52$vowel

## create a vowel template with the normalized formant frequencies
## and information about the vowel category.
#template = createtemplate (formants, vowels)

## and inspect with plot()
#plot (template, xsampa = TRUE)

Error bars.

Description

This functions adds error bars to a plot or, optionally, plots points and added error bars.

Usage

errorbars (x, y, top, bottom = top, length = .2, add = TRUE, ...)

Arguments

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

Inspired by a similar function first seen here:

http://monkeysuncle.stanford.edu/?p=485

Examples

## add bars to existing plots 
#plot (1:10, 1:10, col = 4, pch = 18)
#errorbars (1:10, 1:10, .5, col = 4)

## or create a new plot
#errorbars (1:10, 1:10, 2, add = FALSE)

Fant (1973) Swedish Vowel Data

Description

Formant frequency information for vowels averaged across 50 male speakers.

Usage

data (f73)

Format

A data frame with the following columns:

sex

- A factor indicating speaker sex.

vowel

- The vowel category in x-sampa.

f1

- A numeric vector indcating the vowel F1 in Hz.

f2

- A numeric vector indcating the vowel F2 in Hz.

f3

- A numeric vector indcating the vowel F3 in Hz.

f4

- A numeric vector indcating the vowel F4 in Hz.

References

Fant, G. (1973). Speech sounds and features. Cambridge, MA: MIT Press.

Examples

#data(f73)
#vowelplot (f73$f1, f73$f2, f73$vowel, logaxes = 'xy', xsampa = TRUE)

Fourakis et al. (1999) Greek Vowel Data

Description

Formant frequency information for vowels averaged across 5 male speakers.

Usage

data (f99)

Format

A data frame with the following columns:

sex

- A factor indicating speaker sex.

vowel

- The vowel category in x-sampa.

f1

- A numeric vector indcating the vowel F1 in Hz.

f2

- A numeric vector indcating the vowel F2 in Hz.

References

Fourakis, M., Botinis, A. & Katsaiti, M. (1999). Acoustic characteristics of Greek vowels. Phonetica, 56. 28-43.

Examples

#data(f99)
#vowelplot (f99$f1, f99$f2, f99$vowel, logaxes = 'xy', xsampa = TRUE)

Fast Autocorrelation

Description

Compute the ACF of a signal.

Usage

fastacf  (signal, lag.max = length(signal), window = 'hann',
          show = TRUE, correct = FALSE)

Arguments

Details

The autocorrelation function is calculated using the inverse Fourier transform applied to the power spectrum. This leads to much faster calculation times than the acf() function included in the typical R installation. Corrections for window type and length are carried out as described in Boersma (1993).

Value

A dataframe with the following columns:

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

Boersma, P., (1993). Accurate short-term analysis of the fundamental frequency and the harmonics-to-noise ratio of a sampled sound. Proc. Instit. Phon. Sci. 17: 97-110.

Examples

## Uncomment and run the code below to see the speed advantage.
## Raising the n makes the difference even more pronounced.
#n = 25000
#system.time ({
#acf (rnorm (n), plot = F, lag.max = n)
#})

#system.time ({
#fastacf (rnorm (n), plot = F, lag.max = n)
#})

Find Formants

Description

Find formants given a sound or set of LPC coefficients.

Usage

findformants (sound, fs = 10000, coeffs = NULL, maxbw = 600, 
minformant = 200, verify = TRUE, showbws = FALSE, 
showrejected = TRUE)

Arguments

Details

Formant frequencies are found analytically using the formulas provided in Snell (1993). If Verify = TRUE, the estimated frequency response, formant locations, and a pole-zero plot of the estimated filter are presented. Accepted formants are presented in 5 colors (which are reused if there are more than 5 formants), while rejected formants are presented in black.

Value

A dataframe with the following elements is returned:

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

Snell, R.(1993). "Formant location from LPC analysis data", IEEE Transactions on Speech and Audio Processing, 1(2), pp. 129-134.

Examples

## make a synthetic vowel with a known set of formant frequencies
## and bandwidths
#sound = vowelsynth (ffs = c(500,1500,2500,3500,4500),
#                    fbw = c(30, 90, 150, 210, 270), f0 = 100)

## compare different plotting options					
#findformants (sound)
#findformants (sound, showrejected = FALSE)
#findformants (sound, showbws = TRUE)

Formant Tracking

Description

Create a formant track for a sound.

Usage

formanttrack (sound, timestep = 5, windowlength = 30, 
formants = 5, cutoff = 5000, minformant = 200, maxbw = 600, 
fs = 22050, show = TRUE, periodicity = .5, returnbw = FALSE)

Arguments

Details

This function works by repeatedly calling findformants(), and periodicity is established using pitchtrack(), both of which are included in this package. When no formants are found, or if less than the desired number of formants are found, a value of zero is returned for that formant, at that time point.

Tracked formants are presented over a greyscale spectrogram if show is TRUE. When plotting, different colors are used for each formant to allow the user to distinguish these.

Value

A dataframe with the following elements is returned:

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

#data (sound)
#par (mfrow = c(2,1), mar = c(4,4,1,1))
#formanttrack (sound)
#formanttrack (sound, periodicity = 0)

Frequency Response

Description

Find the frequency response of a digital filter

Usage

freqresponse (b, a, fs = 0, add = FALSE, show = TRUE,
              steps = 1000,...)

Arguments

Details

This function plots (and optionally returns) the frequency response for the digital filter whose transfer function is determined by the numerator and denominator filter coefficients given in b and a.

Value

A dataframe with two columns (frequency and response) that can be used to redraw the frequency response if required. The 'response' value corresponds to dB. magnitude below peak.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

Lyons, R. G. (2004). Understanding Digital Signal Processing (2nd ed.). Prentice Hall.

Examples

## make a synthetic vowel with a known set of formant frequencies
#sound = vowelsynth (ffs = c(500,1500,2500,3500,4500), 
#fbw = c(30, 80, 150, 200, 220),f0 = 100, dur = 100)

#plot (sound)

## let the LPC function estimate the filter used to generate the vowel
#coeffs = lpc (sound, show = FALSE)

## compare frequency response of estimated filter to vowel spectrum
#spectralslice (sound, col = 4, preemphasisf = 50)
#freqresponse (1, coeffs, add = TRUE, fs = 10000)

## generate a sinc function
#filt = sinc (seq (-15,15,1/2), normalized = TRUE)
## treat it as a low-pass FIR filter and inspect its frequency response
#freqresponse (filt, 1)

Hillenbrand et al. (1995) Vowel Data

Description

Formant frequency, f0 and duration information for vowels collected from 139 speakers in the Hillenbrand et al. (1995) data. Speaker numbers have been modified to be uniquely identifying numbers. Data has been simplified so that only "steady state" formant frequency measures are given. Missing F2 values (n = 10) and F3 values (n = 41) have been imputed using the imputeformants() function included in this package.

Usage

data (h95)

Format

A data frame with 1668 observations on the following 9 variables:

type

- A factor with levels b g m w representing speaker type: boy, girl, man and woman.

speaker

- A numeric vector indicating a uniquely identifying speaker number.

vowel

- The vowel category in x-sampa

dur

- A numeric vector indicating the duration of the vowel in milliseconds.

f0

- A numeric vector indcating the vowel f0 in Hz.

f1

- A numeric vector indcating the vowel F1 in Hz.

f2

- A numeric vector indcating the vowel F2 in Hz.

f3

- A numeric vector indcating the vowel F3 in Hz.

Source

The data was created from data provided on Dr. Hillenbrand's personal website:

http://homepages.wmich.edu/~hillenbr/voweldata.html

References

Hillenbrand, J.M., Getty, L.A., Clark, M.J., and Wheeler, K. (1995). "Acoustic characteristics of American English vowels," Journal of the Acoustical Society of America, 97, 3099-3111.

Examples

#data(h95)
#vowelplot (h95$f1, h95$f2, h95$vowel, logaxes = 'xy', ellipses = TRUE,
#xsampa = TRUE)

Hotelling's T2 Test

Description

Hotelling's T2 test for one and two samples.

Usage

hotelling.test(matrix1, matrix2 = NULL)

Arguments

Details

If a single matrix is provided, this function tests the alternative hypothesis that all column means are not equal to zero. If a second matrix is provided, the alternative hypothesis to be tested is that the group means are not all equal. The statistic is tested using an F-distribution which assumes that the matrices represent (roughly) multivariate normal variables.

This function is only designed for multivariate tests of location. If a univariate test is desired, please use a t-test.

Value

An object of class 'Hotelling.test', a list containing the elements:

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

Hotelling, H. (1931). The generalization of Student's ratio. Annals of Mathematical Statistics 2 (3): 360-378.

http://en.wikipedia.org/wiki/Hotelling's_T-squared_distribution

Examples

## load Peterson & Barney data
#data (pb52)

## separate the Peterson & Barney vowels by speaker
## gender and age (child vs. adult)
#men = pb52[pb52$sex == 'm' & pb52$type == 'm',]
#women = pb52[pb52$sex == 'f' & pb52$type == 'w',]
#boys = pb52[pb52$sex == 'm' & pb52$type == 'c',]
#girls = pb52[pb52$sex == 'f' & pb52$type == 'c',]

## fit 4 separate models which predict F1 frequency 
## on the basis of vowel category. 
#men = rcr (f1 ~ vowel, men$speaker, men)
#women = rcr (f1 ~ vowel, women$speaker, women)
#boys = rcr (f1 ~ vowel, boys$speaker, boys)
#girls = rcr (f1 ~ vowel, girls$speaker, girls)

## A Hotelling T2 test indicates that there are 
## significant differences in F1 frequency 
## based on vowel category between males and females
#hotelling.test (men$coefficients, women$coefficients)

## but no significant differences based on the same 
## criteria between boys and girls.
#hotelling.test (boys$coefficients, girls$coefficients)

Impute Missing Formant Values

Description

Impute missing formant values using a least-squares approximation.

Usage

imputeformants (ffs, speaker, vowel)

Arguments

Details

This function finds the least-squares approximation for each missing value based on the assumption that each formant for a given vowel differs between-speakers solely on the basis on a speaker-specific multiplicative parameter. This assumption is well supported in the literature (Nearey 1978, Nearey & Assmann 2007, Turner et al. 2009). This parameter would be most closely related to gross speaker vocal-tract length.

Value

A vector containing each original formant frequency and imputed formant frequencies where appropriate.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

Nearey, T. M. (1978). Phonetic Feature Systems for Vowels. PhD thesis, Indiana University Linguistics Club.

Nearey, T. M. & P. F. Assmann. (2007). Pobabilistic 'sliding template' models for indirect vowel normalization. in Experimental Approaches to Phonology, edited by M.J. Sole, P. S., Beddor, and M. Ohala (Oxford University Press, Oxford), pp. 246-269.

Turner, R. E., Walters, T. C., Monaghan, J. J. M., & Patterson, R. D. (2009). A statistical, formant-pattern model for segregating vowel type and vocal-tract length in developmental formant data. The Journal of the Acoustical Society of America, 125(4), 2374. doi:10.1121/1.3079772

Examples

#data (h95)

## Select F2 values from the tenth speaker in H95 data
## set the first 5 values to "missing"
#ffs = h95$f2
#ffs[h95$speaker == 10][1:5] = 0
#speaker = h95$speaker
#vowel = h95$vowel

## impute these missing values
#imputedf2 = imputeformants (ffs, speaker, vowel)

## resulting in a very close approximation of the original values
#plot (imputedf2[h95$speaker == 10], h95$f2[h95$speaker == 10])
#abline (0, 1, col = 2)

Interpolation

Description

Piece-wise cubic or linear spline interpolation.

Usage

interpolate (y, x = 1:length(y), steps = 20, increment = -1, 
show = FALSE, type = 'cubic', ...)

Arguments

Details

By default, this function performs a 'natural' cubic spline interpolation between the points provided by the user. Optionally, a linear interpolation between the points may be carried out.

Value

A dataframe with columns corresponding to the x and y dimensions of the interpolated points is returned.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

http://en.wikipedia.org/wiki/Spline_interpolation

Examples

## generate ten random points
#y = rnorm (10, 0, 5)
#interpolate (y, show = TRUE)  ## plot a cubic interpolation
#linear = interpolate (y, type = 'linear')   
## and compare to a linear interpolation
#lines (linear, col = 2) 

Linear Discriminant Boundary

Description

Given two mean vectors and a covariance matrix (and optional prior probabilities), this function will return the slope and intercept of the boundary line between the two categories.

Usage

ldboundary (mean1, mean2, covariance, prior1 = .5, prior2 = .5, add = F, ...)

Arguments

Value

The slope and intercept of the boundary line are returned.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

http://en.wikipedia.org/wiki/Linear_discriminant_analysis https://onlinecourses.science.psu.edu/stat557/book/export/html/35

Examples

## create two groups with the same covariance patterns
#group1 = rmvtnorm (200, means= c(0,0), k=2, sigma = -.4)
#group2 = rmvtnorm (200, means= c(3,3), k=2, sigma = -.4)
#covariance = (var (group1) + var (group2)) / 2

## plot groups and boundary line between categories.
#plot (group1, col = 2, pch = 16, ylim = c(-2,5), xlim = c(-2,5))
#points (group2, col = 4, pch = 16)
#ldboundary (c(0,0), c(3,3), covariance, add = TRUE)

Linear Discriminant Classification

Description

Classify items using linear discriminant analysis.

Usage

ldclassify (data, means, covariance, template = NULL, posterior = 'no')

Arguments

Details

This function classifies the items described in the data matrix by comparing them to the reference patterns for the different candidate categories represented in the means matrix. The category with the minimum Mahalanobis distance to the observed pattern (i.e., a given row of the data matrix) is selected as the winner. Mahalanobis distances are found with using the covariance matrix provided to the function.

Mean and covariance matrices can be made easily for data using the createtemplate() function included in this package.

Value

A vector of winning categories is returned. If winning posteriors are desired, these are returned in a second column. All posteriors are returned in separate columns for each category.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

## load Peterson & Barney vowel data
#data (pb52)

## normalize vowel formant frequencies
#normdvowels = normalize (pb52[,7:9], pb52$speaker, pb52$vowel)
#formants = normdvowels[,1:3]
#vowels = pb52$vowel

## make a vowel template based on these frequencies
#template = createtemplate (formants, vowels) 

## classify vowels
#answers = ldclassify (formants, template$means, template$covariance)
## compare to known categories
#table (answers, vowels)

Load WAV files into R

Description

A function which allows WAV files to be loaded into R.

Usage

loadsound (filename)

Arguments

Details

The function is only compatible with 8 and 16 bit mono WAV files. The function returns a 'sound object'. Many of the functions included in this package interact with 'sound' objects.

Value

An object of class 'sound', a list containing the elements:

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

https://ccrma.stanford.edu/courses/422/projects/WaveFormat/

Examples

## Use the command below to select a WAV file to load into R
## sound = loadsound ()

## sound
## plot (sound)
## spectrogram (sound)

Load Table

Description

Load text table data quickly.

Usage

loadtable (...)

Arguments

Details

This function is a wrapper for read.table() for those times when you just don't feel like typing a filename. The function opens up a file selection dialog box allowing the user to select the file containing the data.

Value

Returns the output of read.table().

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

## uncomment and run
# data = loadtable ()
# head (data)

Linear Predictive Coding

Description

Predict autoregressive filter coefficients.

Usage

lpc (sound, order = round(fs/1000)+3, fs = 10000, show = FALSE, 
add = FALSE, preemph = TRUE)

Arguments

Details

LPC coefficients are estimated using the autocorrelation method. The signal is windowed with a Hanning window prior to analysis.

Value

A vector containing the LPC coefficients is returned.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

## make a synthetic vowel with a known set of formant frequencies
#sound = vowelsynth (ffs = c(500,1500,2500,3500,4500), 
#fbw = c(30, 80, 150, 200, 220),f0 = 100, dur = 250)

## let the LPC function estimate the filter used to generate the vowel
#coeffs = lpc (sound, show = TRUE)

Create a Digital Filter

Description

Design a Finite Impulse Response (FIR) Filter.

Usage

makeFIR (frequency, dB, order = 200, signal = NULL, window = 'hann', 
         verify = FALSE, interpolation = 'linear')

Arguments

Details

Designs Type I FIR filters of odd length (even order). If an odd order is provided, 1 is added to the order. Filters are designed using the window-design method. The filter frequency response is defined at evenly-spaced locations determined by the filter order and the sampling frequency. If frequency specifications do not fall exactly on these points, the nearest appropriate location is used. This design method may lead to 'undesirable' behaviour between specified frequencies. This can be minimized by increasing the filter order and selecting an appropriate window function.

Value

If output = TRUE, the impulse response of the filter specified by the user is returned.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

Lyons, R. G. (2004). Understanding Digital Signal Processing (2nd ed.). Prentice Hall.

Examples

## specify a filter with an arbitrary response
#frequency = c(0, 500, 502, 5000, 5002, 7000, 7002, 11025)
#dB = c(0, 0, -50,  -50, -10,  -10, -70, -70)

## create the filter and verify that the frequency response is as desired
#testfilter = makeFIR (frequency, dB, verify = TRUE, order = 1500)
#spectralslice (testfilter, padding = 1000)


## create the filter and verify that the frequency response is as desired
#makeFIR (frequency, dB, verify = TRUE, order = 300, signal = rnorm (1400))

Make a 'sound' object

Description

Create a 'sound' object from a numeric vector.

Usage

makesound (sound, filename, fs = 22050)

Arguments

Details

This function can make a vector into a 'sound' object. If a filename is not set, the filename defaults to 'sound.wav' where 'sound' indicates the name of the sound variable that was passed to the function. The benefit of working with 'sound' objects is that they carry their sampling frequency and filename (as well as some other information) with them.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

https://ccrma.stanford.edu/courses/422/projects/WaveFormat/

Examples

data (sound)
## take only the first 10000 samples from a 'sound' object
tmp = sound$sound[1:10000]
## and make a new 'sound' object
tmp = makesound (tmp, fs = 22050)
tmp

## get ready to make two plots with thin margins
#multiplot (2); par (mar = c(4,4,1,1));
## and show a spectrogram of the original
#spectrogram (sound)       

## and the new, truncated version
#spectrogram (tmp)         

Plot with variable panel sizes

Description

Create plots with columns or rows of unequal sizes.

Usage

multiplot (n = 2, type = 'r', sizes = rep (1/n, n), show = FALSE)

Arguments

Details

This function is essentially a wrapper for the layout() function, which allows the user to create multi-panel figures in which each row or column is of a varying height/width.

Please note that small rows or columns might result in an error related to figure margins being too large when you try to create a plot. These may be reduced with the 'mar' parameter for the par() function, which sometimes solves the problem.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

## uncomment and run

data (sound)
## run this instead to select your own sound for this demo.
# sound = loadsound ()       

#par (mar = c(4,4,1,1))
#multiplot (n = 3, sizes = c(.25, .5, .25))  

#plot (sound)
#spectrogram (sound, dynamicrange = 50, maxfreq = 7000)
#spectralslice (sound)

Normalize Vowels

Description

Function to normalize vowels using one of several methods.

Usage

normalize(formants, speakers, vowels, method = 'neareyE', 
          corners = NULL)

Arguments

Details

This function normalizes vowels based on provided formant frequencies (FFs). The available methods are:

Nearey formant-extrinsic log-mean ('neareyE'): This method finds the logarithmic-mean FF across all vowels produced by a speaker, and subtracts this value from the log-transformed FFs representing each vowel.

Nearey formant-intrinsic log-mean ('neareyI'): This method finds the logarithmic-mean for each formant independently across all vowels produced by a speaker. The log-mean for each individual formant is then subtracted from the log-transformed FF representing each vowel.

Lobanov ('lobanov'): This method finds the mean and standard deviation for each formant. FFs are then standardized (in the statistical sense) using these estimated parameters for each speaker, for each formant.

Watt and Fabricius ('wandf'): This method requires the user to provided point vowels representing the frontmost and highest vowel, and the lowest (and, ideally central) vowel in a vowel system. An estimate of the centroid of the vowel system is calculated based on these values. Normalized FFs are then expressed as the ratio of observed FFs to the estimated centroids, independently for F1 and F2.

For both Nearey methods, and the Lobanov method, the average is found for each vowel category within-speaker before calculating the overall mean. As a result, the data from each speaker may contain unequal numbers of each vowel category. However, all speakers must be represented by the same vowel categories or the result will be (possibly) subtle differences in normalized vowel spaces dues to the possibly differing estimates of means and stadard deviations of the different formants.

Value

A dataframe with the same numbers of rows as the formant data provided and the following columns:

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

Lobanov, B. M. (1971). Classification of Russian vowels spoken by different listeners. Journal of the Acoustical Society of America 49:606-08.

Nearey, T. M. (1978). Phonetic Feature Systems for Vowels. PhD thesis, Indiana University Linguistics Club.

Watt, D. and Fabricius, A. (2002). Evaluation of a technique for improving the mapping of multiple speakers' vowel spaces in the F1 ~ F2 plane. In D. Nelson, Leeds Working Papers in Linguistics and Phonetics 9:159-73.

Examples

## normalize all Peterson & Barney (1952) vowels using each method.
#data (pb52)

#neareyE = normalize (pb52[,7:9], pb52$speaker, pb52$vowel,
#method = 'neareyE')
#neareyI = normalize (pb52[,7:9], pb52$speaker, pb52$vowel,
#method = 'neareyI')
#lobanov = normalize (pb52[,7:9], pb52$speaker, pb52$vowel,
#method = 'lobanov')
#wandf = normalize (pb52[,7:9], pb52$speaker, pb52$vowel,
#method = 'wandf', corners =  c('i','A'))

## compare normalization methods using vowelplot().
#par (mfrow = c(2,2), mar = c(4,4,3,1))
#vowelplot (neareyE[,1], neareyE[,2], neareyE$vowel, alternateAxes = TRUE, 
#  pointType = 16, main = 'neareyE', ellipses = TRUE)
#vowelplot (neareyI[,1], neareyI[,2], neareyI$vowel, alternateAxes = TRUE, 
#  pointType = 16, main = 'neareyI', ellipses = TRUE)
#vowelplot (lobanov[,1], lobanov[,2], lobanov$vowel, alternateAxes = TRUE, 
#  pointType = 16, main = 'lobanov', ellipses = TRUE)
#vowelplot (wandf[,1], wandf[,2], wandf$vowel, alternateAxes = TRUE, 
#  pointType = 16, main = 'wandf', ellipses = TRUE)

Compare Normalization Methods

Description

Compare the effectiveness of different normalization methods.

Usage

normalize.compare (normd)

Arguments

Details

This function provides a relatively straightforward way to compare the effectiveness of different normalization methods based on the assumption that a good normalization method maximizes the separation between different vowel categories and minimizes the variation within a single vowel category. Minimizing the variation within a single vowel category means that the vowel spaces of different speakers are maximally similar.

This function provides two measures to compare the performance of normalization methods:

1) The square root of the average Mahalanobis distance between vowel categories is found for all pairs of vowel categories. This value indicates the average separation of vowel categories with respect to within-category error and the covariance patterns of the formant frequencies. A higher value indicates a better performing normalization algorithm.

2) The percentage of correct classifications using a linear discriminant model trained on the normalized formant-frequencies using the given category-labels.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

## load the Peterson and Barney vowels
#data (pb52)

## normalize using several different methods
#neareyE = normalize (pb52[,7:9], pb52$speaker, pb52$vowel, 
#method = 'neareyE')
#neareyI = normalize (pb52[,7:9], pb52$speaker, pb52$vowel, 
#method = 'neareyI')
#lobanov = normalize (pb52[,7:9], pb52$speaker, pb52$vowel, 
#method = 'lobanov')
#wandf = normalize (pb52[,7:9], pb52$speaker, pb52$vowel, 
#method = 'wandf', corners =  c('i','A'))
#normd = list (pb52, neareyE, neareyI, lobanov, wandf)

## compare outcome of methods (and unnormalized vowels)
## uncomment to run
#normalize.compare (normd)

Number of Unique Elements

Description

Find the number of unique elements in a vector.

Usage

ntypes (vector) 

Arguments

Details

A simple function that converts a vector to a factor, and finds the number of levels. This provides the number of unique elements in a vector, something I find I frequently need.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

data (pb52)
## the number of unique vowel categories.
ntypes (pb52$vowel)

Pols et al. (1973) Dutch Vowel Data

Description

Formant frequency information for vowels averaged across 24 male speakers.

Usage

data (p73)

Format

A data frame with the following columns:

sex

- A factor indicating speaker sex.

vowel

- The vowel category in x-sampa.

f1

- A numeric vector indcating the vowel F1 in Hz.

f2

- A numeric vector indcating the vowel F2 in Hz.

f3

- A numeric vector indcating the vowel F3 in Hz.

References

Pols, L. C. W., Tromp, H. R. C., & Plomp, R. (1973). Frequency analysis of Dutch vowels from 50 male speakers. Journal of the Acoustical Society of America, 53. 1093-1101.

Examples

data(p73)
#vowelplot (p73$f1, p73$f2, p73$vowel, logaxes = 'xy', ellipses = TRUE, 
#xsampa = TRUE)

Peterson & Barney (1952) Vowel Data

Description

Formant frequency and f0 information for vowels collected from 76 speakers in the Peterson & Barney (1952) data.

Usage

data (pb52)

Format

A data frame with 1520 observations on the following 9 variables:

type

- A factor with levels c m w representing speaker type: child, man and woman.

sex

- A factor with levels f m representing speaker gender: male and female.

speaker

- A numeric vector indicating a uniquely identifying speaker number.

vowel

- The vowel category in x-sampa.

repetition

- A numeric vector indicating the repetition number.

f0

- A numeric vector indcating the vowel f0 in Hz.

f1

- A numeric vector indcating the vowel F1 in Hz.

f2

- A numeric vector indcating the vowel F2 in Hz.

f3

- A numeric vector indcating the vowel F3 in Hz.

Source

The data was created from tables provided within Praat:

http://www.fon.hum.uva.nl/praat/

References

Peterson, G.E. & Barney (1952). Control methods used in a study of the vowels. Journal of the Acoustical Society of America 24: 175-184.

Boersma, Paul & Weenink, David (2012). Praat: doing phonetics by computer [Computer program]. Version 5.3.19, retrieved 24 June 2012 from http://www.praat.org/

Examples

data(pb52)
#vowelplot (pb52$f1, pb52$f2, pb52$vowel, logaxes = 'xy', ellipses = TRUE, 
#xsampa = TRUE)

Find the Peaks

Description

Locate the peaks in a numeric vector.

Usage

peakfind (x, show = TRUE)

Arguments

Details

This function looks for peaks by finding elements whose value is greater than both the elements that surround it. If no peaks are found, a value of zero is returned.

Value

A vector indicating the location (position in the vector) of peaks in the vector.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

#data (sound)
#sound2 = sound$sound[10000:11000] 
#spectrum = spectralslice (sound2, padding = 0, output = TRUE, show = TRUE)
#peakfind (spectrum[,2])

Plot Phasors

Description

Plot phasors representing one or more complex-valued variables.

Usage

phasor (num, scaled = TRUE, add = FALSE, circle = FALSE, 
        xlim, ylim, ...)

Arguments

Details

Complex-valued variables may be created using the complex() function.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

#noise = rnorm (100)
#phasors = fft(noise)

#par (mfrow = c(1,3))
#phasor (phasors)
#phasor (phasors, circle = TRUE)
#phasor (phasors, scaled = FALSE)

Pick IPA Symbols

Description

Select IPA symbols that you wish to include in a plot.

Usage

pickIPA (vowels, n = 0, xsampa = FALSE, description = FALSE, 
         verify = TRUE)

Arguments

Details

This is an interactive function that allows the user to select IPA symbols for plotting using a chart. The values returned by this function may only be passed to the 'pch' parameter within plotting functions. At the moment it has only been implemented for vowel sounds.

If a vowels vector is given, the function finds the number of categories in the vector. The user is then prompted to select the IPA symbol corresponding to each category by clicking on the correct location on the plot.

If a vowels vector is not provided, the function allows the user to select any desired number of symbols, and returns these in the same order as indicated by the user.

**There may be issues when exporting figures to PDF using IPA font. Exporting plots directly as images works 'out of the box'**

Value

A list with the following columns (some of which are optional):

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

http://en.wikipedia.org/wiki/X-SAMPA

Examples

##uncomment to run
#vowels = pickIPA (n = 3)
#plot (c(1,2,3), c(1,2,3), pch = vowels)

## select vowels in the order displayed in the console
## to obtain symbols and descriptions of the vowel categories 
## in the Peterson and Barney data. 

# data (pb52)
# tmp = pickIPA (pb52$vowel, description = TRUE, xsampa = TRUE)
# tmp

Pitch Tracking

Description

Create a pitch track for a sound.

Usage

pitchtrack (sound, f0range = c(60,400), timestep = 2, fs = 22050, minacf = .5,
correction = TRUE, show = TRUE, windowlength = 50, addtospect = FALSE)

Arguments

Details

Pitch tracking is carried out using a simplified version of the algorithm described in Boersma (1993), including corrections for lag value and window function. When plotting pitch tracks, the points sizes are proportional to autocorrelation values.

Value

A dataframe with the following columns:

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

Boersma, P., (1993). Accurate short-term analysis of the fundamental frequency and the harmonics-to-noise ratio of a sampled sound. Proc. Instit. Phon. Sci. 17: 97-110.

Examples

#data (sound)               ## use the example 'sound' object provided
#sound = loadsound()       ## or run this line to use you own sound

## to generate a pitch track
#output = pitchtrack (sound)

## to add a pitch to a spectrogram
#spectrogram (sound)
#pitchtrack (sound, addtospect = TRUE)
 

Play Sounds

Description

Play sounds in R using VLC Player.

Usage

playsound  (sound, path = 'default', fs = 10000, erase = TRUE)

Arguments

Details

The sound is written out as a .wav file and VLC is called from the command line to play the file. The file is optionally erased after the fact. If the path is selected by the user, this is returned for use in future calls. Obviously, this function relies on VLC being installed. With the appropriate path, it seems like this function should work on Linux and OSX, though it has only been tested on a Windows computer.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

#  ## make a 1 second, 100 Hz tone.
#  tone = sinusoid (freqs = 250, dur = 1000, fs = 1000)[,2]
#  ## play it in VLC player
#  playsound (q[,2], fs = 1000, erase = FALSE)

Pole-zero Plots

Description

Generate a Pole-zero plot from filter coefficients.

Usage

polezero (b, a, ...)

Arguments

Details

This function plots filter poles (x) and zeros (o) based on the given coefficients.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

http://en.wikipedia.org/wiki/Pole

Examples

## example of a typical single-zero preemphasis filter
#a = 1
#b = c(1, -.94)
#polezero (b, a)

#example of a complex-pole formant-style filter
#a = c(1, -.3, .2)
#b = c(1)
#polezero (b, a)

Power tracking

Description

Create a power track for a sound.

Usage

powertrack (sound, timestep = 5, windowlength = 30, 
            fs = 22050, show = TRUE, zeromax = TRUE, ...)

Arguments

Details

Returns the average power in the signal in a section as determined by the parameters of the function. A Hann window is applied to each section prior to analysis. Sections with zero power are ignored and not returned.

Value

A dataframe with the following columns:

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

## plot the waveform and power of a given sound.
#data (sound)
# sound = loadsound()        ## run this line to use you own sound

## compare waveform and power
#par (mfrow = c(2,1), mar = c(4,4,1,1));
#plot (sound)
#powertrack (sound)

Add Preemphasis

Description

Single-zero preemphasis filter.

Usage

preemphasis (input, cutoff = 50, fs = 22050, verify = FALSE, 
coeff = 0)

Arguments

Value

The modified sound is returned. If a 'sound' object another 'sound' object is returned.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

http://www.fon.hum.uva.nl/praat/manual/Sound__Filter__pre-emphasis____.html

Examples

#signal = sinusoid (c(100, 200, 400, 800, 1600), fs = 4000, 
#dur = 100, sum = TRUE)[,7]

#preemphasis (signal, verify = TRUE, fs = 4000, cutoff = 50)

Welch's Power Spectral Density Estimate

Description

Calculates a power spectral density estimate using Welch's method.

Usage

pwelch (sound, points = 0, overlap = 0, padding = 0, 
window = 'hamming', show = TRUE, fs = 1, preemphasisf = 0, 
zeromax = TRUE, type,...)

Arguments

Details

This function divides the signal into a number of equally-sized windows, finds the power spectrum of each one, and then finds the average across all windowed sections.

Value

The following columns:

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

## make a sine wave
#sinewave = sinusoid (f = 300, fs = 1000, dur = 2000, sum = FALSE)

## add noise
#sinewave = sinewave[,2] + rnorm (length (sinewave[,2]), 0, 3)

## compare the results of pwelch() and spectralslice()
#par (mfrow = c(2,1), mar = c(4,4,1,1))
#spectralslice (sinewave, fs = 1000)
#pwelch (sinewave, points = 400, fs = 1000)

Random Coefficients Regression

Description

Carry out a random coefficients regression (rcr) using repeated calls to glm, individually for the data from each participant/data cluster.

Usage

rcr(formula, participants, dataframe, ...)

Arguments

Details

This function fits a model to the data from each participant individually using repeated calls to glm(). Significance testing is then carried out on the coefficients fit for each participant using the methods established in Gumpertz & Pantula (1989) and Lorch & Myers (1990).

In perceptual experiments there is frequently a high number of data points collected from each participant, and the data collected from each participant is balanced by design. In these situations rcr performs comparably to mixed-effects models. In the event that only a small number of observations are made from each listener, or the data is not balanced, rcr may not be appropriate.

A call to summary() on an rcr object performs a one-sample t-test on each coefficient to test whether it is significantly different from zero.

A call to anova() on an rcr object performs a one-sample t-test in the case of single coefficients, and a one-sample Hotelling T2 test in the event that multiple coefficients are associated with a single factor, to test that they are not all equal to zero.

A call to plot() on an rcr object displays the density corresponding to the distribution of all fitted coefficients. These are compared to a normal distribution with the same mean and standard deviation.

Value

An object of class 'rcr', a list containing the elements:

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

Gumpertz, M., & Pantula, S. G. (1989). A Simple Approach to Inference in Random Coefficient Models. The American Statistician, 43(4), 203-210.

Lorch, T. F. & Myers, J. L. (1990). Regression analyses of repeated measures data in cognitive research. J. Exp. Psychol. Learn. Mem. Cogn. 16: 149-157.

Examples

#data (pb52)

## runs an rcr model on the Peterson & Barney (1952) vowels to test 
## for the predictive value of the speaker's f0 and F3
#rcr.model = rcr (f1 ~ f0 * f3, pb52$speaker, pb52)
#rcr.model

## test for the significance of each individual coefficient 
#summary (rcr.model)

## a similar analysis can be run using vowel category as the predictor
#rcr.model = rcr (f1 ~ vowel, pb52$speaker, pb52)
#rcr.model

## here, summary() tests each coefficient individually
#summary (rcr.model)

## while anova() tests associated coefficients together
#anova (rcr.model)

Reduce Fractions

Description

Reduce fractions to lowest terms using Euclid's Algorithm.

Usage

reduce.fraction (ratio)

Arguments

Value

A vector containing the elements of the reduced fraction. The first element is the numerator of a ratio, and the second element is the denominator.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

## an easy one
reduce.fraction (c(100, 200))

## irreducible
reduce.fraction (c(140, 201))

## a hard one
reduce.fraction (c(140, 203))

Resample a Sound

Description

Resample using sinc interpolation.

Usage

resample (sound, newfs, oldfs, precision = 50, filterorder = 200, synthfilter = FALSE)

Arguments

Value

The resampled vector is returned. If a 'sound' object is passed, the resampled sound is returned as an object.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

#data (sound)
## downsample and then upsample the sound back to 
## its original sampling frequency
#downsamped = resample (sound, 11025)
#upsamped = resample (downsamped, 22050)

# compare a part of the waveforms for all three sounds
#par (mfrow = c(3,1), mar = c(4,4,1,1))
#plot (sound$sound[1:14000], type = 'l')
#plot (upsamped$sound[1:14000], type = 'l', col = 2)
#plot (downsamped$sound[1:7000], type = 'l', col = 4)

Random multivariate normal variables

Description

Draw vectors from a multivariate normal distribution.

Usage

rmvtnorm (n = 1, k = 2, means = rep (0, k), sigma = diag (k))

Arguments

Details

If means and sigma are not specified, a standard normal distribution is assumed along every dimensions, and dimensions are assumed to be uncorrelated. If the number of dimensions is not specified, a bivariate distribution is assumed.

Value

A matrix with rows equal to n and columns equal to k, where each row indicates a single observation, and each column represents a different dimension.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

## Examples of draws from different bivariate normal distributions
## and standard deviation ellipses drawn to fit them.
#par (mfrow = c(2,2))
#draws = rmvtnorm (n = 1000, k = 2, sigma = .3)
#plot (draws)
#sdellipse (draws, stdev = 3, lwd = 3, col = 2)

#draws = rmvtnorm (n = 1000, k = 2, sigma = -.3)
#plot (draws)
#sdellipse (draws, stdev = 3, lwd = 3, col = 2)

#draws = rmvtnorm (n = 1000, k = 2, sigma = -.7)
#plot (draws)
#sdellipse (draws, stdev = 3, lwd = 3, col = 2)

#draws = rmvtnorm (n = 1000, k = 2, sigma = .7)
#plot (draws)
#sdellipse (draws, stdev = 3, lwd = 3, col = 2)

Rotate

Description

Rotate 2D and complex-valued observations. The output is of the same kind as the input.

Usage

rotate (xy, angle, degrees = FALSE, origin = TRUE)

Arguments

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

## rotate points in a 2D space 
#mat = cbind (1:100, 100:1)
#rotate (mat, pi/2)

## rotate complex-valued numbers
#complx = complex (real = mat[,1], imaginary = mat[,2])
#rotate (complx, pi/2)

Standard deviation Ellipse

Description

Draw standard deviation ellipses around a group of observations.

Usage

sdellipse (points, stdev = 1.96, density = .01,  
add = TRUE, show = TRUE, means = NULL, se = FALSE, ...)

Arguments

Details

This function may be used in 2 different ways: 1) To draw standard deviation ellipses around a set of observations (if 'means' is not specified) 2) To draw ellipses and circles on plots (if 'means' is specified).

Value

A matrix is returned where the first column indicate x-axis ellipse coordinates and the second column indicates y-axis ellipse coordinates.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

## Examples of draws from different bivariate normal distributions
## and standard deviation ellipses drawn to fit them.
#par (mfrow = c(2,2))
#draws = rmvtnorm (n = 1000, k = 2, sigma = .3)
#plot (draws)
#sdellipse (draws, stdev = 3, lwd = 3, col = 2)

#draws = rmvtnorm (n = 1000, k = 2, sigma = -.3)
#plot (draws)
#sdellipse (draws, stdev = 3, lwd = 3, col = 2)

#draws = rmvtnorm (n = 1000, k = 2, sigma = -.7)
#plot (draws)
#sdellipse (draws, stdev = 3, lwd = 3, col = 2)

#draws = rmvtnorm (n = 1000, k = 2, sigma = .7)
#plot (draws)
#sdellipse (draws, stdev = 3, lwd = 3, col = 2)

## alternatively, a covariance matrix may be specified directly. 
#par (mfrow = c(1,1))
#sdellipse (matrix(c(1,.5,.5,1),2,2), means = c(0,0), 
#add = FALSE, stdev = 1)
#sdellipse (matrix(c(1,-.5,-.5,1),2,2), means = c(0,0), 
#add = TRUE, stdev = 1)

Select Slices

Description

Select spectral slices from a spectrogram.

Usage

selectslice (specobject, n = 1, plot = TRUE, ...)

Arguments

Details

This function allows the user to select a given number of time points on a spectrogram, and to retrieve spectral slices at those time points. The spectrogram object is plotted, and the user must click on the spectrogram a given number of times at the desired points. The analysis parameters of the slices will reflect the parameters used when creating the spectrogram. Setting the 'plot' parameter to FALSE is useful if you want one set of analysis parameters for the visually presented spectrogram and another set for the spectral slices.

Value

A dataframe with one column for each spectral slice selected. Row names indicate frequencies, column names indicate times in milliseconds. Values are decibels below peak.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

## uncomment to run

# sound = vowelsynth ()
# spect = spectrogram (sound)
# slices = selectslice (spect, n = 3)

Sinc Function

Description

Sample the sinc function at given points

Usage

sinc (x, normalized = FALSE)

Arguments

Details

The formula for the unnormalized sinc function is y = sin(x)/x.

The formula for the normalized sinc function is y = sin(x*pi)/(x*pi).

Value

A vector containing y-axis values for each specified x-axis value is returned.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

http://en.wikipedia.org/wiki/Sinc_function

Lyons, R. G. (2004). Understanding Digital Signal Processing (2nd ed.). Prentice Hall.

Examples

#x = seq(-20,20,.1)
## generate both forms of the sinc function between -20 and 20
#y1 = sinc (x)
#y2 = sinc (x, normalized = TRUE)
## the unnormalized sinc function has zero crossing at 
## integer multiples of pi
#plot (x,y1, type = 'b')
## the normalized sinc function has zero crossing at integers
#lines (x,y2, type = 'b', col = 4)
#abline (h = 0)

Create Sinusoids

Description

Create and display one or more sinusoids.

Usage

sinusoid (freqs, amps = rep(1, length(freqs)), dur = 50, 
phases = rep(0, length(freqs)), fs = 10000, sum = FALSE, 
show = FALSE, colors = NULL)

Arguments

Details

A number of sinusoids are generated, and optionally plotted and/or returned. The number of frequencies specified must equal the number of amplitudes and initial phases.

Value

A dataframe with the following columns:

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

## several waves, and the sum
#sum = sinusoid (freqs = c(100,200,300), amps = c(1,3,2), 
#sum = TRUE, show = TRUE)

## no sum, different phase shifts
#nosum = sinusoid (freqs = c(100,200,300), amps = c(1,3,2), 
#phases = c(pi/2, 0, pi/4), sum = FALSE)

Snip/Zoom

Description

Select a subsection of a sound or spectrogram object.

Usage

snip  (object, show = TRUE)

Arguments

Details

The input object is plotted and the user must click on two points on the plot. If show = TRUE, this function allows the user to zoom in on a subsection of the object. If the output is assigned to a variable, a new object is created and returned that contains only the data in between the two designated points.

Value

A 'sound' or 'spectrogram' object which interacts with several functions included in this package.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

#data (sound)
## the example below is commented because examples cannot 
## require user interaction. simply uncomment the lines below
## and select two points on the plot. 
# clipped = snipsound (sound)
# spectrogram (clipped)
# spectralslice (clipped)

Sound object

Description

An example of a sound object, the phrase 'This is a spectrogram', produced by the author of this package. This sound object may be inspected with spectrogram() and plotted with plot(). Sections of the individual samples representing the waveform (found in the "sound" element of a sound object) can be passed to spectralslice() to see a spectral slice. Several functions included in this package may also be used to modify or manipulate 'sound' objects.

Usage

data (sound)

Format

This sound object has the same properties as all sound objects. These may be inspected by using the print() function.

Examples

## uncomment and run

#data (sound)
#par (mar = c(4,4,1,1))
#multiplot (n = 3, sizes = c(.25, .5, .25))  

#plot (sound)
#spectrogram (sound, dynamicrange = 50, maxfreq = 7000)

Spectral Slice

Description

A function to plot the power spectrum of a vector representing a sound wave.

Usage

spectralslice (sound, padding = length(sound) * 2, fs = 1,  
show = TRUE, add = FALSE, window = "kaiser", 
windowparameter = 3, zeromax = TRUE, preemphasisf = 0, type, 
line = FALSE, removeDC = TRUE, ...)

Arguments

Value

A dataframe with the following elements is returned:

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

## synthesize schwa-like vowel 
#vowel = vowelsynth (ffs = c(500,1500,2500,3500,4500))$sound
    
## compare window lengths
#par (mfrow = c(3,1))
#spectralslice (vowel[500:550], fs = 10000)
#spectralslice (vowel[500:1000], fs = 10000)
     
## line spectrum
#spectralslice (vowel[500:600], padding = 0, line = TRUE, fs = 10000)

Create Spectrograms

Description

Create and display spectrograms.

Usage

spectrogram (sound, fs = 22050, windowlength = 5, 
timestep = -1000, padding = 10, 
preemphasisf = 50, maxfreq = 5000, colors = TRUE, 
dynamicrange = 50, nlevels = dynamicrange, maintitle = "", 
show = TRUE, window = 'kaiser', windowparameter = 3, 
quality = FALSE)

Arguments

Details

This function is used to create and plot spectrograms. The user may specify all analysis parameters, in addition to the colors used to display the spectrogram.

The function optionally returns a spectrogram object for which the plot() function is defined.

Value

If output is set to TRUE, an object of class 'spectrogram', a list containing the elements:

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

## uncoment and run. 
#data (sound)             ## use the example 'sound' object provided
#sound = loadsound()       ## or run this line to use you own sound

#spectrogram (sound)
# spectrogram (sound, quality = TRUE)

Filtering by Synthesis

Description

Use Fourier synthesis to recreate signal without undesired frequency components.

Usage

synthfilter (sound, band = c(0,fs/4), fs = 1, verify = FALSE, attenuation = 0)

Arguments

Details

This function performs lowpass, highpass and bandpass filtering by performing an FFT on the entire signal, zeroing out coefficients representing undesired frequency components, and performing an IFFT on the result. This approach may not be appropriate for some applications, but it is useful in some cases. This may be slow for long signals.

If verify is TRUE, a plot is created which allows the user to inspect the performance of the function.

Value

If a vector is given, the filtered vector is returned. If a 'sound' object is given, a sound object containing the filtered sound is returned.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

## uncomment and run
##  white noise
#noise = rnorm(5000)

## low-pass filter
#synthfilter (noise, band = c(0, .25), verify = TRUE)

##  band-pass filter
#synthfilter (noise, band = c(.15, .25), verify = TRUE)

##  high-pass filter
#synthfilter (noise, band = c(.35, .5), verify = TRUE)

Thomson (2007) Vowel Data

Description

Mean formant frequency information for the vowels of male and female speakers from Edmonton English.

Usage

data (pb52)

Format

A data frame with 1520 observations on the following 9 variables:

vowel

- The vowel category in x-sampa.

gender

- A factor with levels f m representing speaker gender: male and female.

f1

- A numeric vector indcating the vowel F1 in Hz.

f2

- A numeric vector indcating the vowel F2 in Hz.

f3

- A numeric vector indcating the vowel F3 in Hz.

References

Thomson, R. (2007). Modeling L1/L2 interactions in the perception and production of English vowels by Mandarin L1 speakers: A training study. PhD dissertation, University of Alberta.

Examples

#data(t07)
#vowelplot (t07$f1, t07$f2, t07$vowel, logaxes = 'xy', meansOnly = TRUE, 
#xsampa = TRUE)

Information about Vowel Data Sets

Description

Contains a brief description and listing of all vowel data sets included in this package.

Usage

data (voweldata)

Format

A data frame with the following information:

name

- The name of the data set.

language

- The language represented in the data set.

nspeakers

- The number of speakers.

sexes

- The sexes of the speakers, m (male) and female (f).

nformants

- The number of formants represented, between 2 and 4.

individual

- If 'yes', individual speaker measurements are given, if 'no', measurements are averaged within gender.

References

At the moment 9 vowel data sets are included in this package. See individual data set pages for citation information.

Examples

#data(voweldata)

Plot Vowels

Description

A flexible function that can create a wide variety of vowel plots (including IPA symbols).

Usage

vowelplot (f1s, f2s, labels = 0, xrange = NULL, yrange = NULL, 
meansOnly = FALSE, ellipses = FALSE, ellipsesd = 1.96, add = FALSE, 
pointType = 0, colors = NULL, logaxes = '', defaultPlot = TRUE, 
alternateAxes = FALSE, xsampa = FALSE, ...)

Arguments

Details

*** This function has been deprecated and is no longer being developed. I recommend using vplot() instead, which is more flexible and gives the user more control over plotting. ***

This function now includes functionality to easily generate vowel plots using IPA symbols. This relies on category labels being specified in x-sampa. Alternatively, the required plotting values for IPA symbols may be selected using the pickIPA() function included in this package, and then passed to the 'pointType' parameter.

There may be issues when exporting figures to PDF using IPA font. Exporting plots directly as images works 'out of the box'.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

http://en.wikipedia.org/wiki/X-SAMPA

Examples

## A few examples of some vowel plots. 

#data (pb52)
#par (mfrow = c(1,4), mar = c(4.2,4.2,1,1))

# standard layout with linear axes
#vowelplot (pb52$f1, pb52$f2, pb52$vowel, xsampa = TRUE)

# alternate layout with log axes
#vowelplot (pb52$f1, pb52$f2, pb52$vowel, logaxes = 'xy', 
#alternateAxes = TRUE, xsampa = TRUE)

# category means only 
#vowelplot (pb52$f1, pb52$f2, pb52$vowel, logaxes = 'xy', 
#meansOnly = TRUE, xsampa = TRUE)

# category means only with standard deviation ellipses
#vowelplot (pb52$f1, pb52$f2, pb52$vowel, logaxes = 'xy', meansOnly = TRUE,
#           ellipses = TRUE, xsampa = TRUE)

Vowel Synthesis

Description

Create synthetic vowels using a cascade formant synthesizer.

Usage

vowelsynth (ffs = c(270, 2200, 2800, 3400, 4400), fbw = 0.06, 
dur = 300, f0 = c(120,100),fs = 10000, verify = FALSE, 
returnsound = TRUE, noise1 = .001, noise2 = .01, power = NULL)

Arguments

Details

This function is a Klatt-style cascade formant synthesizer that is intended to create synthetic vowel sounds. The voice source is generated using the KLGLOTT88 method described in Klatt (1988).

If the vowels sound too 'robotic', there may be formants too close to the Nyquist frequency. Conversely, if the vowels sound too 'muffled', you may need to add more formants, and make sure they extend closer to the Nyquist frequency. Vowels sound more natural when f0 is not static.

Formant bandwidths may be provided in Hz, or as a percentage of the formant frequencies. To set these as a percent of formant frequencies, all values must be less than 1. If these are not provided they are set to 6 percent of the formant center frequencies by default. If only one value is provided, this is assumed to be the desired value for all formants.

Value

A vector or 'sound' object representing the filtered sound.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

Klatt, D. H. (1980). Software for a cascade/parallel formant synthesizer. Journal of the Acoustical Society of America 67(3): 971-995.

Klatt, D. H. (1988). Klattalk: The conversion of English text to speech. Unpublished Manuscript. Massachusetts Institute of Technology, Cambridge, MA. Chapter 3.

Examples

## uncomment to run

## The following examples are based on my vowels 
#i = vowelsynth (returnsound = FALSE, f0 = c(125,105))
#a = vowelsynth (ffs = c(700, 1300, 2300, 3400, 4400), 
#returnsound = FALSE, f0 = c(125,105))
#e = vowelsynth (ffs = c(400, 2000, 2600, 3400, 4400), 
#returnsound = FALSE, f0 = c(125,105))
#o = vowelsynth (ffs = c(400, 900, 2300, 3400, 4400), 
#returnsound = FALSE, f0 = c(125,105))
#u = vowelsynth (ffs = c(300, 750, 2300, 3400, 4400), 
#returnsound = FALSE, f0 = c(125,105))

#silence = rep(0, 1000)
#vowels = c(a, silence, e, silence, i, silence, o, silence, u)

#writesound (vowels, filename = 'vowels.wav', fs = 10000)

## an example of a synthetic diphthong
#ei = vowelsynth (ffs = list(c(400, 2000, 2600, 3400, 4400), 
#c(270, 2200, 2800, 3400, 4400)), f0 = c(125,105))
#writesound (ei)
#spectrogram (ei, pause = FALSE)

Plot Vowels

Description

A flexible function that can create a wide variety of vowel plots (including IPA symbols).

Usage

vplot (x, y, labels = NULL, colors = NULL, points = NULL, 
meansonly = FALSE, ellipsesd = 0, add = FALSE, alternateaxes = FALSE,
xsampa = FALSE, logaxes = FALSE, ...)

Arguments

Details

*** This function replaces the older vowelplot() function, which has been deprecated. ***

A flexible vowel plotting function, including functionality to easily generate vowel plots using IPA symbols. This relies on category labels being specified in x-sampa (the required plotting values for IPA symbols may be selected using the pickIPA() function included in this package).

Default parameter values are set for the plot, but these may all be overridden using the standard plotting parameters.

There may be issues when exporting figures to PDF using IPA font. Exporting plots directly as images works 'out of the box'.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

http://en.wikipedia.org/wiki/X-SAMPA

Examples

## A few examples of some vowel plots. 

## load the Peterson and Barney data
#data (pb52)
#pb52 = pb52[pb52$type=='m',]  ## use only the males

#par (mfrow = c(3,2), mar = c(4.2,4.2,1,1))

# standard layout with linear axes
#vplot (pb52$f1, pb52$f2, pb52$vowel, xsampa = TRUE)

# alternate layout with log axes
#vplot (pb52$f1, pb52$f2, pb52$vowel, logaxes = TRUE, 
#           alternateaxes = TRUE, xsampa = TRUE)

# category means only 
#vplot (pb52$f1, pb52$f2, pb52$vowel, logaxes = TRUE, 
#           meansonly = TRUE, xsampa = TRUE, cex = 3)

# category means only with standard deviation ellipses
#vplot (pb52$f1, pb52$f2, pb52$vowel, logaxes = FALSE, 
#       meansonly = TRUE, ellipsesd = 2, xsampa = TRUE)

# same as above, with alternate axes
#vplot (pb52$f1, pb52$f2, pb52$vowel, logaxes = TRUE, 
#       meansonly = TRUE, ellipsesd = 2, xsampa = TRUE, 
#	   alternateaxes = TRUE)

# individual points with standard deviation ellipses
# and alternate axes
#vplot (pb52$f1, pb52$f2, pb52$vowel, logaxes = TRUE, 
#	   meansonly = FALSE, ellipsesd = 2, xsampa = TRUE, 
#	   alternateaxes = TRUE)

Window Function

Description

Generates a window function of a given type and length.

Usage

windowfunc (npoints, type = 'hann', parameter = -1)

Arguments

Details

A window function is generated, of the type and length specified by the user. This is returned as a numeric vector.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

http://en.wikipedia.org/wiki/Window_function

Examples

#par (mfrow = c(1,4))
#plot (windowfunc (100, 'hann'))
#plot (windowfunc (100, 'bartlett'))
#plot (windowfunc (100, 'kaiser', parameter = 2))
#plot (windowfunc (100, 'kaiser', parameter = 6))

Write out a WAV file

Description

Create a WAV file from a numeric vector or 'sound' object.

Usage

writesound (samples, filename = '', fs = 22050)

Arguments

Details

This function generates single channel (mono), 16-bit WAV sound files at a desired sampling frequency. If a 'sound' object is passed, the filename and sampling frequency do not need to be set. If a filename is not set, the filename defaults to 'samples.wav' where 'samples' indicates the name of the samples variable that was passed to the function.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

https://ccrma.stanford.edu/courses/422/projects/WaveFormat/

Examples

## generate a sine wave with a frequency of 1000 Hz
## sampled at a frequency of 10000 Hz
#x = seq (0,.1, 1/10000)
#snd = sin (2*pi*1000*x)
#plot (snd[1:100], type = 'b')

## write out sine wave as a WAV file
# writesound (snd, filename = '1khz.wav', fs = 10000)

## if no filename is provided, this sound will be called 'snd.wav'
# writesound (snd, fs = 10000)

Convert X-Sampa to IPA

Description

Convert a vector of x-sampa vowel descriptions to values that can be used to plot IPA.

Usage

xsampatoIPA (vowels, chart = FALSE, verify = FALSE)

Arguments

Details

This function converts x-sampa to values that can be used to plot IPA symbols by passing them to the 'pch' parameter within plotting functions. At the moment it has only been implemented for vowel sounds. The function also generate a figure comparing IPA to x-sampa vowel representations that may be useful to some users.

**There may be issues when exporting figures to PDF using IPA font. Exporting plots directly as images works 'out of the box'**

Value

A vector of the same length as 'vowels', with the values required to plot the desired vowels.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

References

http://en.wikipedia.org/wiki/X-SAMPA

Examples

## compare x-sampa and IPA vpwel charts
#xsampatoIPA ()
## some examples
#IPA = xsampatoIPA (c('I','3','e','Q'))
#plot (1:5, pch = IPA, cex = 2)

Yang (1996) Korean Vowel Data

Description

Formant frequency information for vowels averaged across 60 male and female speakers.

Usage

data (y96)

Format

A data frame with the following columns:

sex

- A factor indicating speaker sex.

vowel

- The vowel category in x-sampa.

f1

- A numeric vector indcating the vowel F1 in Hz.

f2

- A numeric vector indcating the vowel F2 in Hz.

f3

- A numeric vector indcating the vowel F3 in Hz.

References

Yang, B. (1996). A comparative study of American English and Korean vowels produced by male and female speakers. Journal of Phonetics, 24. 245-261.

Examples

#data(y96)
#vowelplot (y96$f1, y96$f2, y96$vowel, logaxes = 'xy', xsampa = TRUE)

Zero Vector/Matrix

Description

Returns a vector or 2D matrix of zeros of the same shape as the input.

Usage

zeros (x,y=0)

Arguments

Details

If a scalar value is given, a vector of that many zeros is returned. If the y parameter is also specified, a matrix is returned with x rows and y columns.

If a vector is given, a vector of zeros of the same length is returned.

If a 2D matrix or dataframe is given, a matrix of zeros with the name number of rows and columns is returned.

Author(s)

Santiago Barreda <sbarreda@ucdavis.edu>

Examples

## A vector 
zeros (5)
## A matrix
zeros (5, 3)

## Copy an existing vector
zeros (1:10)

## Copy an existing matrix
samplematrix = 1:10 
zeros (samplematrix)