Import sound files and select signals

Bioacoustics in R with `warbleR`

Bioacoustics research encompasses a wide range of questions, study systems and methods, including the software used for analyses. The warbleR and Rraven packages leverage the flexibility of the R environment to offer a broad and accessible bioinformatics tool set. These packages fundamentally rely upon two types of data to perform bioacoustics analyses in R:

Sound files: Recordings in wav or mp3 format, either from your own research or open-access databases like xeno-canto
Selection tables: Selection tables contain the temporal coordinates (start and end points) of selected acoustic signals within recordings

Package repositories

These packages are both available on CRAN: warbleR, Rraven, as well as on GitHub: warbleR, Rraven. The GitHub repository will always contain the latest functions and updates. You can also check out an article in Methods in Ecology and Evolution documenting the warbleR package ^[1].

We welcome all users to provide feedback, contribute updates or new functions and report bugs to warbleR’s GitHub repository.

Please note that warbleR and Rraven use functions from the seewave, monitoR, tuneR and dtw packages internally. warbleR and Rraven have been designed to make bioacoustics analyses more accessible to R users, and such analyses would not be possible without the tools provided by the packages above. These packages should be given credit when using warbleR and Rraven by including citations in publications as appropriate (e.g. citation("seewave")).

Parallel processing in `warbleR`

Parallel processing, or using multiple cores on your machine, can greatly speed up analyses. All iterative warbleR functions now have parallel processing for Linux, Mac and Windows operating systems. These functions also contain progress bars to visualize progress during normal or parallel processing. See ^[1] for more details about improved running time using parallel processing.

Vignette introduction

Below we present a case study of microgeographic vocal variation in long-billed hermit hummingbirds, Phaethornis longirostris. Variation at small geographic scales has been already described in this species ^[2]. Our goal is to search for visible differences in song structure within a site, and then determine whether underlying differences in acoustic parameters are representative of spectrographic distinctiveness. In this vignette, we will demonstrate how to:

Prepare for bioacoustics analyses by downloading warbleR and Rraven
Use Rraven to import Raven selection tables for your own recordings
Obtain recordings from the open-access database xeno-canto
Select signals using warbleR functions

This vignette can be run without an advanced understanding of R, as long as you know how to run code in your console. However, knowing more about basic R coding would be very helpful to modify the code for your research questions.

For more details about function arguments, input or output, read the documentation for the function in question (e.g. ?query_xc).

Prepare for analyses

Install and load packages

First, we need to install and load warbleR and Rraven. You will need an R version .1 and seewave version .1. Also, users using UNIX machines (Linux or Mac operating systems), may need to install fftw3, pkg-config and libsndfile on their machines prior to installing warbleR. These external packages will need to be installed through a UNIX terminal. Installing these packages lies outside the scope of this vignette, but you can find more information on Google.

### Install packages from CRAN
# Note that if you install from CRAN, then don't run the code to install from GitHub below, and vice versa
install.packages("warbleR")
install.packages("Rraven")

### Alternatively, install warbleR and Rraven from GitHub repositories, which contain the latest updates
# Run this ONLY if devtools is not already installed
install.packages("devtools")

# Load devtools to access the install_github function
library(devtools)

# Install packages from GitHub
# install_github("maRce10/warbleR")
# install_github("maRce10/Rraven")
# install_github("maRce10/NatureSounds")

# Load warbleR and Rraven into your global environment
X <- c("warbleR", "Rraven")
invisible(lapply(X, library, character.only = TRUE))

This vignette series will not always include all available warbleR functions, as existing functions are updated and new functions are added. To see all functions available in this package:

# The package must be loaded in your working environment
ls("package:warbleR")

Make a new directory and set your working directory

# Create a new directory and set your working directory (assuming that you are in your /home/username directory)
dir.create(file.path(getwd(),"warbleR_example"))
setwd(file.path(getwd(),"warbleR_example"))

# Check your location
getwd()

Import selection tables

Rraven is an interface between Raven and R that allows you to import selection tables for your own recordings. This is very useful if you prefer to select signals in recordings outside of R. Once you have selection tables imported into R and the corresponding sound files in your working directory, you can move on to making spectrograms or performing analyses (see the next vignette in this series).

The sound files and selection tables loaded here correspond to male long-billed hermit hummingbird songs recorded at La Selva Biological Station in Costa Rica. Later, we will add to this data set by searching for more recordings on the xeno-canto open-access database.

Check out the Rraven package documentation for more functions and information (although you will need Raven or Syrinx installed on your computer for some functions).

# Load Raven example selection tables
data("selection_files")

# Write out Raven example selection tables as physical files
out <- lapply(1:2, function(x) 
  writeLines(selection_files[[x]], con = names(selection_files)[x]))

# Write example sound files out as physical .wav files
data(list = c("Phae.long1", "Phae.long2"))

writeWave(Phae.long1, "Phae.long1.wav")
writeWave(Phae.long2, "Phae.long2.wav")

# Import selections
sels <- imp_raven(all.data = FALSE, freq.cols = FALSE, warbler.format = TRUE)
str(sels)

# Write out the imported selections as a .csv for later use
write.csv(sels, "Raven_sels.csv", row.names = FALSE)

Make your data frame into an object of class `selection table`

Downstream warbleR functions require selection tables in order to run correctly. Use the function selection_table to convert your data frame into an object of class selection_table. In future versions of warbleR, all functions will require selection table objects of class selection_table.

sels <- selection_table(X = sels)
str(sels)
class(sels)

Obtain metadata and recordings from xeno-canto

The open-access xeno-canto database is an excellent source of sound files across avian species. You can query this database by a species or genus of interest. The function query_xc has two types of output:

Metadata of recordings: geographic coordinates, recording quality, recordist, type of signal, etc.
Sound files: Sound files in mp3 format are returned if the argument download is set to TRUE.

We recommend downloading metadata first from xeno-canto, as this data can be filtered in R to more efficiently download recordings (e.g. only those relevant to your question).

Here, we will query the xeno-canto database to download more Phaethornis longirostris sound files for our question of how male songs vary at a microgeographic scale.

# Query xeno-canto for all Phaethornis recordings (e.g., by genus)
Phae <- query_xc(qword = "Phaethornis", download = FALSE) 

# Check out the structure of resulting the data frame
str(Phae)

'data.frame':   899 obs. of  36 variables:
 $ Recording_ID     : int  406968 403856 403854 387711 275252 261696 261695 261694 261693 261692 ...
 $ Genus            : chr  "Phaethornis" "Phaethornis" "Phaethornis" "Phaethornis" ...
 $ Specific_epithet : chr  "yaruqui" "yaruqui" "yaruqui" "yaruqui" ...
 $ Subspecies       : chr  "" "" "" "" ...
 $ English_name     : chr  "White-whiskered Hermit" "White-whiskered Hermit" "White-whiskered Hermit" "White-whiskered Hermit" ...
 $ Recordist        : chr  "Myornis" "Myornis" "Myornis" "Jerome Fischer" ...
 $ Country          : chr  "Colombia" "Colombia" "Colombia" "Ecuador" ...
 $ Locality         : chr  "La Chocoana, Guachalito, Nuquí, Chocó" "La Chocoana, Guachalito, Nuquí, Chocó" "La Chocoana, Guachalito, Nuquí, Chocó" "Amagusa Reserve Pichincha" ...
 $ Latitude         : num  5.627 5.627 5.627 0.161 0.75 ...
 $ Longitude        : num  -77.4 -77.4 -77.4 -78.9 -78.9 ...
 $ Vocalization_type: chr  "flight call" "flight call" "flight call" "song" ...
 $ Audio_file       : chr  "//www.xeno-canto.org/406968/download" "//www.xeno-canto.org/403856/download" "//www.xeno-canto.org/403854/download" "//www.xeno-canto.org/387711/download" ...
 $ License          : chr  "//creativecommons.org/licenses/by-nc-sa/4.0/" "//creativecommons.org/licenses/by-nc-sa/4.0/" "//creativecommons.org/licenses/by-nc-sa/4.0/" "//creativecommons.org/licenses/by-nc-sa/4.0/" ...
 $ Url              : chr  "//www.xeno-canto.org/406968" "//www.xeno-canto.org/403856" "//www.xeno-canto.org/403854" "//www.xeno-canto.org/387711" ...
 $ Quality          : chr  "A" "A" "A" "A" ...
 $ Time             : chr  "?" "?" "?" "09:00" ...
 $ Date             : chr  "2018-08-00" "2017-08-00" "2017-08-00" "2017-09-24" ...
 $ Altitude         : chr  "30" "10" "10" "1300" ...
 $ Spectrogram_small: chr  "//www.xeno-canto.org/sounds/uploaded/OXSIOLJJUP/ffts/XC406968-small.png" "//www.xeno-canto.org/sounds/uploaded/OXSIOLJJUP/ffts/XC403856-small.png" "//www.xeno-canto.org/sounds/uploaded/OXSIOLJJUP/ffts/XC403854-small.png" "//www.xeno-canto.org/sounds/uploaded/JPBSNBUUEF/ffts/XC387711-small.png" ...
 $ Spectrogram_med  : chr  "//www.xeno-canto.org/sounds/uploaded/OXSIOLJJUP/ffts/XC406968-med.png" "//www.xeno-canto.org/sounds/uploaded/OXSIOLJJUP/ffts/XC403856-med.png" "//www.xeno-canto.org/sounds/uploaded/OXSIOLJJUP/ffts/XC403854-med.png" "//www.xeno-canto.org/sounds/uploaded/JPBSNBUUEF/ffts/XC387711-med.png" ...
 $ Spectrogram_large: chr  "//www.xeno-canto.org/sounds/uploaded/OXSIOLJJUP/ffts/XC406968-large.png" "//www.xeno-canto.org/sounds/uploaded/OXSIOLJJUP/ffts/XC403856-large.png" "//www.xeno-canto.org/sounds/uploaded/OXSIOLJJUP/ffts/XC403854-large.png" "//www.xeno-canto.org/sounds/uploaded/JPBSNBUUEF/ffts/XC387711-large.png" ...
 $ Spectrogram_full : chr  "//www.xeno-canto.org/sounds/uploaded/OXSIOLJJUP/ffts/XC406968-full.png" "//www.xeno-canto.org/sounds/uploaded/OXSIOLJJUP/ffts/XC403856-full.png" "//www.xeno-canto.org/sounds/uploaded/OXSIOLJJUP/ffts/XC403854-full.png" "//www.xeno-canto.org/sounds/uploaded/JPBSNBUUEF/ffts/XC387711-full.png" ...
 $ Length           : chr  "0:02" "0:04" "0:01" "0:50" ...
 $ Uploaded         : chr  "2018-03-23" "2018-02-28" "2018-02-28" "2017-09-27" ...
 $ Other_species    : chr  "" "Coereba flaveola" "" "" ...
 $ Remarks          : chr  "In forest." "Forest border." "Forest border." "" ...
 $ Bird_seen        : chr  "yes" "yes" "yes" "yes" ...
 $ Playback_used    : chr  "no" "no" "no" "no" ...
 $ Other_species1   : chr  NA NA NA NA ...
 $ Other_species2   : chr  NA NA NA NA ...
 $ Other_species3   : chr  NA NA NA NA ...
 $ Other_species4   : chr  NA NA NA NA ...
 $ Other_species5   : chr  NA NA NA NA ...
 $ Other_species6   : chr  NA NA NA NA ...
 $ Other_species7   : chr  NA NA NA NA ...
 $ Other_species8   : chr  NA NA NA NA ...

# Query xeno-canto for all Phaethornis longirostris recordings
Phae.lon <- query_xc(qword = "Phaethornis longirostris", download = FALSE) 

# Check out the structure of resulting the data frame
str(Phae.lon)

'data.frame':   85 obs. of  31 variables:
 $ Recording_ID     : int  497036 495384 433645 402755 355350 282529 274377 271499 154138 154129 ...
 $ Genus            : chr  "Phaethornis" "Phaethornis" "Phaethornis" "Phaethornis" ...
 $ Specific_epithet : chr  "longirostris" "longirostris" "longirostris" "longirostris" ...
 $ Subspecies       : chr  "" "cephalus" "" "" ...
 $ English_name     : chr  "Long-billed Hermit" "Long-billed Hermit" "Long-billed Hermit" "Long-billed Hermit" ...
 $ Recordist        : chr  "Jerome Fischer" "Guy Kirwan" "Oscar Campbell" "Marilyn Castillo" ...
 $ Country          : chr  "Panama" "Panama" "Panama" "Mexico" ...
 $ Locality         : chr  "Achiote Road, Colón Province" "Panama Rainforest Discovery Centre" "Panama Rainforest Discovery Centre" "Boca de Chajul, Marqués de Comillas, Chiapas" ...
 $ Latitude         : num  9.2 9.13 9.13 16.13 8.94 ...
 $ Longitude        : num  -80 -79.7 -79.7 -90.9 -78.5 ...
 $ Vocalization_type: chr  "song" "call" "song" "alarm call" ...
 $ Audio_file       : chr  "//www.xeno-canto.org/497036/download" "//www.xeno-canto.org/495384/download" "//www.xeno-canto.org/433645/download" "//www.xeno-canto.org/402755/download" ...
 $ License          : chr  "//creativecommons.org/licenses/by-nc-sa/4.0/" "//creativecommons.org/licenses/by-nc-sa/4.0/" "//creativecommons.org/licenses/by-nc-sa/4.0/" "//creativecommons.org/licenses/by-nc-nd/4.0/" ...
 $ Url              : chr  "//www.xeno-canto.org/497036" "//www.xeno-canto.org/495384" "//www.xeno-canto.org/433645" "//www.xeno-canto.org/402755" ...
 $ Quality          : chr  "no score" "A" "A" "A" ...
 $ Time             : chr  "09:30" "09:00" "08:30" "07:30" ...
 $ Date             : chr  "2019-02-10" "2019-07-04" "2018-07-10" "2016-01-18" ...
 $ Altitude         : chr  "30" "70" "70" "180" ...
 $ Spectrogram_small: chr  "//www.xeno-canto.org/sounds/uploaded/JPBSNBUUEF/ffts/XC497036-small.png" "//www.xeno-canto.org/sounds/uploaded/BZVYBRUAAE/ffts/XC495384-small.png" "//www.xeno-canto.org/sounds/uploaded/RFRTVEHIZX/ffts/XC433645-small.png" "//www.xeno-canto.org/sounds/uploaded/CHRIDPJVMH/ffts/XC402755-small.png" ...
 $ Spectrogram_med  : chr  "//www.xeno-canto.org/sounds/uploaded/JPBSNBUUEF/ffts/XC497036-med.png" "//www.xeno-canto.org/sounds/uploaded/BZVYBRUAAE/ffts/XC495384-med.png" "//www.xeno-canto.org/sounds/uploaded/RFRTVEHIZX/ffts/XC433645-med.png" "//www.xeno-canto.org/sounds/uploaded/CHRIDPJVMH/ffts/XC402755-med.png" ...
 $ Spectrogram_large: chr  "//www.xeno-canto.org/sounds/uploaded/JPBSNBUUEF/ffts/XC497036-large.png" "//www.xeno-canto.org/sounds/uploaded/BZVYBRUAAE/ffts/XC495384-large.png" "//www.xeno-canto.org/sounds/uploaded/RFRTVEHIZX/ffts/XC433645-large.png" "//www.xeno-canto.org/sounds/uploaded/CHRIDPJVMH/ffts/XC402755-large.png" ...
 $ Spectrogram_full : chr  "//www.xeno-canto.org/sounds/uploaded/JPBSNBUUEF/ffts/XC497036-full.png" "//www.xeno-canto.org/sounds/uploaded/BZVYBRUAAE/ffts/XC495384-full.png" "//www.xeno-canto.org/sounds/uploaded/RFRTVEHIZX/ffts/XC433645-full.png" "//www.xeno-canto.org/sounds/uploaded/CHRIDPJVMH/ffts/XC402755-full.png" ...
 $ Length           : chr  "0:27" "0:49" "0:58" "0:03" ...
 $ Uploaded         : chr  "2019-09-13" "2019-09-01" "2018-09-09" "2018-02-14" ...
 $ Other_species    : chr  "" "" "" "" ...
 $ Remarks          : chr  "" "Lek with up to three different individuals. Focal bird was perched 1.5 m above ground on a 45-degree angle twig." "Male seen on lek; 3 feet above ground in gap in undergrowth. Calling incessantly and quivering tail as doing so"| __truncated__ "" ...
 $ Bird_seen        : chr  "yes" "yes" "yes" "yes" ...
 $ Playback_used    : chr  "no" "no" "no" "no" ...
 $ Other_species1   : chr  NA NA NA NA ...
 $ Other_species2   : chr  NA NA NA NA ...
 $ Other_species3   : chr  NA NA NA NA ...

You can then use the function map_xc to visualize the geographic spread of the queried recordings. map_xc will create an image file of a map per species in your current directory if img = TRUE. If img = FALSE, maps will be displayed in the graphics device.

# Image type default is jpeg, but tiff files have better resolution

# When the data frame contains multiple species, this will yield one map per species
map_xc(X = Phae, img = TRUE, it = "tiff") # all species in the genus
map_xc(X = Phae.lon, img = FALSE) # a single species

Filter xeno-canto recordings by quality, signal type and locality

In most cases, you will need to filter the xeno-canto metadata by type of signal you want to analyze. When you subset the metadata, you can input the filtered metadata back into query_xc to download only the selected recordings. There are many ways to filter data in R, and the example below can be modified to fit your own data.

Here, before downloading the sound files themselves from xeno-canto, we want to ensure that we select high quality sound files that contain songs (rather than other acoustic signal types) and were also recorded at La Selva Biological Station in Costa Rica.

# How many recordings are available for Phaethornis longirostris?
nrow(Phae.lon)

[1] 85

# How many signal types exist in the xeno-canto metadata?
unique(Phae.lon$Vocalization_type)

 [1] "song"              "call"              "alarm call"        "song at lek"       "lekking"           "flight call"       "calls"            
 [8] "300"               "song, wing whirrs" "lek, song"

# How many recordings per signal type?
table(Phae.lon$Vocalization_type)


              300        alarm call              call             calls       flight call         lek, song           lekking              song 
                1                 1                 6                 1                 2                 2                 1                68 
      song at lek song, wing whirrs 
                2                 1

# Filter the metadata to select the signals we want to retain

# First by quality
Phae.lon <- Phae.lon[Phae.lon$Quality == "A", ]
nrow(Phae.lon)

[1] 12

# Then by signal type
Phae.lon.song <- Phae.lon[grep("song", Phae.lon$Vocalization_type, ignore.case = TRUE), ]
nrow(Phae.lon.song)

[1] 9

# Finally by locality
Phae.lon.LS <- Phae.lon.song[grep("La Selva Biological Station, Sarapiqui, Heredia", Phae.lon.song$Locality, ignore.case = FALSE), ]

# Check resulting data frame, 6 recordings remain
str(Phae.lon.LS)

'data.frame':   3 obs. of  31 variables:
 $ Recording_ID     : int  154138 154129 154072
 $ Genus            : chr  "Phaethornis" "Phaethornis" "Phaethornis"
 $ Specific_epithet : chr  "longirostris" "longirostris" "longirostris"
 $ Subspecies       : chr  "" "" ""
 $ English_name     : chr  "Long-billed Hermit" "Long-billed Hermit" "Long-billed Hermit"
 $ Recordist        : chr  "Marcelo Araya-Salas" "Marcelo Araya-Salas" "Marcelo Araya-Salas"
 $ Country          : chr  "Costa Rica" "Costa Rica" "Costa Rica"
 $ Locality         : chr  "La Selva Biological Station, Sarapiqui, Heredia" "La Selva Biological Station, Sarapiqui, Heredia" "La Selva Biological Station, Sarapiqui, Heredia"
 $ Latitude         : num  10.4 10.4 10.4
 $ Longitude        : num  -84 -84 -84
 $ Vocalization_type: chr  "song" "song" "song"
 $ Audio_file       : chr  "//www.xeno-canto.org/154138/download" "//www.xeno-canto.org/154129/download" "//www.xeno-canto.org/154072/download"
 $ License          : chr  "//creativecommons.org/licenses/by-nc-sa/3.0/" "//creativecommons.org/licenses/by-nc-sa/3.0/" "//creativecommons.org/licenses/by-nc-sa/3.0/"
 $ Url              : chr  "//www.xeno-canto.org/154138" "//www.xeno-canto.org/154129" "//www.xeno-canto.org/154072"
 $ Quality          : chr  "A" "A" "A"
 $ Time             : chr  "14:18" "10:09" "7:05"
 $ Date             : chr  "2010-05-21" "2010-05-21" "2010-05-28"
 $ Altitude         : chr  "" "" ""
 $ Spectrogram_small: chr  "//www.xeno-canto.org/sounds/uploaded/EMCWQLLKEW/ffts/XC154138-small.png" "//www.xeno-canto.org/sounds/uploaded/EMCWQLLKEW/ffts/XC154129-small.png" "//www.xeno-canto.org/sounds/uploaded/EMCWQLLKEW/ffts/XC154072-small.png"
 $ Spectrogram_med  : chr  "//www.xeno-canto.org/sounds/uploaded/EMCWQLLKEW/ffts/XC154138-med.png" "//www.xeno-canto.org/sounds/uploaded/EMCWQLLKEW/ffts/XC154129-med.png" "//www.xeno-canto.org/sounds/uploaded/EMCWQLLKEW/ffts/XC154072-med.png"
 $ Spectrogram_large: chr  "//www.xeno-canto.org/sounds/uploaded/EMCWQLLKEW/ffts/XC154138-large.png" "//www.xeno-canto.org/sounds/uploaded/EMCWQLLKEW/ffts/XC154129-large.png" "//www.xeno-canto.org/sounds/uploaded/EMCWQLLKEW/ffts/XC154072-large.png"
 $ Spectrogram_full : chr  "//www.xeno-canto.org/sounds/uploaded/EMCWQLLKEW/ffts/XC154138-full.png" "//www.xeno-canto.org/sounds/uploaded/EMCWQLLKEW/ffts/XC154129-full.png" "//www.xeno-canto.org/sounds/uploaded/EMCWQLLKEW/ffts/XC154072-full.png"
 $ Length           : chr  "2:58" "2:25" "3:09"
 $ Uploaded         : chr  "2013-11-11" "2013-11-11" "2013-11-11"
 $ Other_species    : chr  "" "" ""
 $ Remarks          : chr  "Recording equipment: Marantz Pmd 660+sennheiser ME67. Comments: individuo-AK; lek-Sura. Primera parte de grabac"| __truncated__ "Recording equipment: Marantz Pmd 660+sennheiser ME67. Comments: individuo-AK; lek-Sura. video 511, revisar. 9 m"| __truncated__ "Recording equipment: Marantz Pmd 660+sennheiser ME67. Comments: individuo-VA; lek-CCL. en percha ded VA, probab"| __truncated__
 $ Bird_seen        : chr  "yes" "yes" "yes"
 $ Playback_used    : chr  "no" "no" "no"
 $ Other_species1   : chr  NA NA NA
 $ Other_species2   : chr  NA NA NA
 $ Other_species3   : chr  NA NA NA

We can check if the location coordinates make sense (all recordings should be from a single place in Costa Rica) by making a map of these recordings using map_xc.

# map in the RStudio graphics device (img = FALSE)
map_xc(Phae.lon.LS, img = FALSE)

Once you’re sure you want the recordings, use query_xc to download the files. Also, save the metadata as a .csv file.

# Download sound files
query_xc(X = Phae.lon.LS) 

# Save the metadata object as a .csv file 
write.csv(Phae.lon.LS, "Phae_lon.LS.csv", row.names = FALSE)

Convert xeno-canto mp3 recordings to wav format

xeno-canto maintains recordings in mp3 format due to file size restrictions. However, we require wav format for all downstream analyses. Compression from wav to mp3 and back involves information losses, but recordings that have undergone this transformation have been successfully used in research ^[3].

To convert mp3 to wav, we can use the warbleR function mp32wav, which relies on a underlying function from the tuneR package. This function does not always work (and it remains unclear as to why!). This bug should be fixed in future versions of tuneR. If RStudio aborts when running mp32wav, use an mp3 to wav converter online, or download the open source software Audacity (available for Mac, Linux and Windows users).

After mp3 files have been converted, we need to check that the wav files are not corrupted and can be read into RStudio (some wav files can’t be read due to format or permission issues).

# Always check you're in the right directory beforehand
# getwd()

# here we are downsampling the original sampling rate of 44.1 kHz to speed up downstream analyses in the vignette series
mp32wav(samp.rate = 22.05) 

# Use checkwavs to see if wav files can be read
check_wavs()

A note on combining data from different sources

We now have .wav files for existing recordings ( Phae.long1.wav through Phae.long4.wav, representing recordings made in the field) as well as 6 recordings downloaded from xeno-canto. The existing Phae.long*.wav recordings have associated selection tables that were made in Raven, but the xeno-canto have no selection tables, as we have not parsed these sound files to select signals within them.

Depending on your question(s), you can combine your own sound files and those from xeno-canto into a single data set (after ground-truthing). This is made possible by the fact that warbleR functions will read in all sound files present in your working directory.

For the main case study in this vignette, we will move forwards with only the xeno-canto sound files. We will use the example sound files when demonstrating warbleR functions that are not mandatory for the case study but may be useful for your own workflow (e.g. consolidate below).

To continue the workflow, remove all example wav files from your working directory

# Make sure you are in the right working directory
# Note that all the example sound files begin with the pattern "Phae.long"
wavs <- list.files(pattern = "wav$")
wavs

rm <- wavs[grep("Phae.long", wavs)]

file.remove(rm)

# Check that the right wav files were removed
# Only xeno-cant wav files should remain
list.files(pattern = "wav$")

Consolidate sound files across various directories

Since warbleR handles sound files in working directories, it’s good practice to keep sound files associated with the same project in a single directory. If you’re someone who likes to make a new directory for every batch of recordings or new analysis associated with the same project, you may find the consolidate function useful.

In case you have your own recordings in wav format and have skipped previous sections, you must specify the location of the sound files you will use prior to running downstream functions by setting your working directory again.

# For this example, set your working directory to an empty temporary directory
setwd(tempdir())

# Here we will simulate the problem of having files scattered in multiple directories

# Load .wav file examples from the NatureSounds package
data(list = c("Phae.long1", "Phae.long2", "Phae.long3"))

# Create first folder inside the temporary directory and write new .wav files inside this new folder
dir.create("folder1")
writeWave(Phae.long1, file.path("folder1","Phae_long1.wav"))
writeWave(Phae.long2, file.path("folder1","Phae_long2.wav"))

# Create second folder inside the temporary directory and write new .wav files inside this second new folder
dir.create("folder2")
writeWave(Phae.long3, file.path("folder2","Phae_long3.wav"))

# Consolidate the scattered files into a single folder, and make a .csv file that contains metadata (location, old and new names in the case that files were renamed)
invisible(consolidate(path = tempdir(), save.csv = TRUE))

list.files(path = "./consolidated_folder")

# set your working directory back to "/home/user/warbleR_example" for the rest of the vignette, or to whatever working directory you were using originally

Make long spectrograms of whole recordings

full_spectrograms produces image files with spectrograms of whole sound files split into multiple rows. It is a useful tool for filtering by visual inspection.

full_spectrograms allows you to visually inspect the quality of the recording (e.g. amount of background noise) or the type, number, and completeness of the vocalizations of interest. You can discard the image files and recordings that you no longer want to analyze.

First, adjust the function arguments as needed. We can work on a subset of the recordings by specifying their names with the flist argument.

# Create a vector of all the recordings in the directory
wavs <- list.files(pattern = "wav$")

# Print this object to see all sound files
# 6 sound files from xeno-canto
wavs 

# Select a subset of recordings to explore full_spectrograms() arguments 
# Based on the list of wav files we created above
sub <- wavs[c(1, 5)]

# How long are these files? this will determine number of pages returned by full_spectrograms
duration_wavs(sub)

# ovlp = 10 to speed up function
# tiff image files are better quality and are faster to produce
full_spectrograms(flist = sub, ovlp = 10, it = "tiff")

# We can zoom in on the frequency axis by changing flim, 
# the number of seconds per row, and number of rows
full_spectrograms(flist = sub, flim = c(2, 10), sxrow = 6, rows = 15, ovlp = 10, it = "tiff")

Once satisfied with the argument settings we can make long spectrograms for all the sound files.

# Make long spectrograms for the xeno-canto sound files
full_spectrograms(flim = c(2, 10), ovlp = 10, sxrow = 6, rows = 15, it = "jpeg", flist = fl)

# Concatenate full_spectrograms image files into a single PDF per recording
# full_spectrograms images must be jpegs to do this
full_spectrograms2pdf(keep.img = FALSE, overwrite = TRUE)

The pdf image files (in the working directory) for the xeno-canto recordings should look like this:

The sound file name and page number are placed in the top right corner. The dimensions of the image are made to letter paper size for printing and subsequent visual inspection.

Recording 154123 has a lot of background noise. Delete the wav file for this recording to remove it from subsequent analyses.

Select signals in warbleR

warbleR provides a function for selecting acoustic signals within recordings. auto_detec automatically detects the start and end of signals in sound files based on amplitude, duration, and frequency range attributes.

Both functions are fastest with shorter recordings, but there are ways to deal with larger recordings (an hour long or more). In this section we have expanded on some important function arguments, but check out the function documentation for more information.

Automatically detect signals with `auto_detec`

auto_detec has 2 types of output:** + data frame with recording name, selection, start and end times. The last two are temporal coordinates that will be passed on to downstream functions to measure acoustic parameters + a spectrogram per recording, with red dotted lines marking the start and end of each detected signal, saved in your working directory

Check out the auto_detec documentation for more information. The argument threshold is one of the most important detection parameters, as well as information about signal frequency range and duration. Phaethornis longirostris songs have frequencies between 2 and 10 kHz and durations between 0.05 and 0.5 s.

If you need to detect all or most of the signals within the recording, play around with different arguments to increase detection accuracy. Since you may need to do several rounds of optimization, we recommend using subsets of your recordings to speed up the process. If the species you study produces stereotyped signals, like Phaethornis longirostris, just a few signals are needed per recording, and a low-accuracy detection could yield enough selections.

auto_detec does not replace visual inspection of selected signals. Ensure that you set aside the time to inspect all selected signals for accuracy. You will often need to run detection functions several times, and in the process you will get to know your signals better (if you don’t already).

# Select a subset of sound files
# Reinitialize the wav object
wavs <- list.files(pattern = ".wav$", ignore.case = TRUE)

# Set a seed so we all have the same results
set.seed(1)
sub <- wavs[sample(1:length(wavs), 3)]

# Run auto_detec() on subset of recordings
# The data frame object output is printed to the console, we are not saving this in an object yet, since we are just playing around with argument settings
# you can run this in parallel to speed up computation time
auto_detec(flist = sub, bp = c(1, 10), threshold = 10, mindur = 0.05, maxdur = 0.5, envt="abs", ssmooth = 300, ls = TRUE, res = 100, flim = c(1, 12), wl = 300, set = TRUE, sxrow = 6, rows = 15, redo = FALSE)

Check out the image files in your working directory. Note that some songs were correctly detected but other undesired sounds were also detected. In most cases, the undesired selections have a shorter duration than our target signals.

We won’t save the auto_detec ouput in an object until we’re satisfied with the detection. To improve our detection we should play around with argument values. Also note that the image files produced by auto_detec contain the values used for the different arguments, which can help you better compare between runs. Below are some detection parameters that work well for these Phaethornis longirostris recordings:

auto_detec(flist = sub, bp = c(2, 10), threshold = 20, mindur = 0.09, maxdur = 0.22, envt = "abs", ssmooth = 900, ls = TRUE, res = 100, flim= c(1, 12), wl = 300, set =TRUE, sxrow = 6, rows = 15, redo = TRUE, it = "tiff", img = TRUE, smadj = "end")

This seems to provide a good detection for most recordings (recording ID 154161):

Once we’re satisfied with the detection, we can run the auto_detec on all the recordings, removing the argument flist (so auto_detec runs over all wav files in the working directory). We will also save the temporal output in an object.

Phae.ad <- auto_detec(bp = c(2, 10), threshold = 20, mindur = 0.09, maxdur = 0.22, envt = "abs", ssmooth = 900, ls = TRUE, res = 100, flim = c(2, 10), wl = 300, set =TRUE, sxrow = 6, rows = 15, redo = TRUE, it = "tiff", img = TRUE, smadj = "end")

Let’s look at the number of selections per sound file:

table(Phae.ad$sound.files)

Use SNR to filter automatically selected signals

Signal-to-noise ratio (SNR) can be a useful filter for automated signal detection. When background noise is detected as a signal it will have a low SNR, and this characteristic can be used to remove background noise from the auto_detec selection table. SNR = 1 means the signal and background noise have the same amplitude, so signals with SNR <= 1 are poor quality. SNR calculations can also be used for different purposes throughout your analysis workflow.

Optimize SNR measurements

snr_spectrograms is a function in the family of spectrogram creators that allows you to pick a margin for measuring noise. These margins are very important for calculating SNR, especially when working with signals separated by short gaps (e.g. duets).

# A margin that's too large causes other signals to be included in the noise measurement
# Re-initialize X as needed, for either auto_detec output

# Try this with 10% of the selections first
# Set a seed first, so we all have the same results
set.seed(5)

X <- Phae.ad[sample(1:nrow(Phae.ad),(nrow(Phae.ad)*0.05)), ]
nrow(X)

snr_spectrograms(X = X, flim = c(2, 10), snrmar = 0.5, mar = 0.7, it = "jpeg")

Check out the image files in your working directory. This margin overlaps neighboring signals, so a smaller margin would be better.

# This smaller margin is better
snr_spectrograms(X = X, flim = c(2, 10), snrmar = 0.04, mar = 0.7, it = "jpeg")

Calculate SNR for automatically selected signals

Once we’ve picked an SNR margin we can move forward with the SNR calculation. We will measure SNR on every other selection to speed up the process.

Phae.snr <- sig2noise(X = Phae.ad[seq(1, nrow(Phae.ad), 2), ], mar = 0.04)

As we just need a few songs to characterize individuals (here sound files are equivalent to different individuals), we can choose selections with the highest SNR per sound file. In this example, we will choose 5 selections per recording with the highest SNRs.

Phae.hisnr <- Phae.snr[ave(-Phae.snr$SNR, Phae.snr$sound.files, FUN = rank) <= 5, ]

# save the selections as a physical file
write.csv(Phae.hisnr, "Phae_hisnr.csv", row.names = FALSE)

# Double check the number of selection per sound files
# Only the xeno-canto sound files will have 5 selections, the other sound files started off with less than 5 selections
table(Phae.hisnr$sound.files)

Next vignette: Visual inspection and signal classification

Here we have given examples of how to begin the warbleR workflow. Note that there are many different ways to begin the workflow, depending on your question and source of data. After running the code in this first vignette, you should now have an idea of:

the type of data used in warbleR (sound files and selections)
how to import Raven selection tables for your own sound files
how to obtain open-access xeno-canto sound files
how to create long spectrograms of recordings for visual inspection
how to select signals within sound files in warbleR
- automatic selection
- filtering automatically selected signals using SNR
- manual selection

The next vignette will cover the second phase of the warbleR workflow, which includes methods to visualize signals for quality control and classification.

Citation

Please cite warbleR when you use the package:

Araya-Salas, M. and Smith-Vidaurre, G. (2017), warbleR: an R package to streamline analysis of animal acoustic signals. Methods Ecol Evol. 8, 184-191.

Reporting bugs

Please report any bugs here.

References

Araya-Salas, M. and G. Smith-Vidaurre. 2016. warbleR: an R package to streamline analysis of animal acoustic signals. Methods in Ecology and Evolution. doi: 10.1111/2041-210X.12624
Araya-Salas, M. and T. Wright. 2013. Open-ended song learning in a hummingbird. Biology Letters. 9 (5). doi: 10.1098/rsbl.2013.0625
Medina-Garcia, Angela, M. Araya-Salas, and T. Wright. 2015. Does vocal learning accelerate acoustic diversification? Evolution of contact calls in Neotropical parrots. Journal of Evolutionary Biology. doi: 10.1111/jeb.12694

warbleR: Import sound files and select signals

Marcelo Araya-Salas, PhD

Grace Smith-Vidaurre

“2022-02-23”

Bioacoustics in R with `warbleR`

Package repositories

Parallel processing in `warbleR`

Vignette introduction

Prepare for analyses

Install and load packages

Make a new directory and set your working directory

Import selection tables

Make your data frame into an object of class `selection table`

Obtain metadata and recordings from xeno-canto

Filter xeno-canto recordings by quality, signal type and locality

Convert xeno-canto mp3 recordings to wav format

A note on combining data from different sources

Consolidate sound files across various directories

Make long spectrograms of whole recordings

Select signals in warbleR

Automatically detect signals with `auto_detec`

Use SNR to filter automatically selected signals

Optimize SNR measurements

Calculate SNR for automatically selected signals

Next vignette: Visual inspection and signal classification

Citation

Reporting bugs

References

warbleR: Import sound files and select signals

Marcelo Araya-Salas, PhD

Grace Smith-Vidaurre

“2022-02-23”

Bioacoustics in R with warbleR

Package repositories

Parallel processing in warbleR

Vignette introduction

Prepare for analyses

Install and load packages

Make a new directory and set your working directory

Import selection tables

Make your data frame into an object of class selection table

Obtain metadata and recordings from xeno-canto

Filter xeno-canto recordings by quality, signal type and locality

Convert xeno-canto mp3 recordings to wav format

A note on combining data from different sources

Consolidate sound files across various directories

Make long spectrograms of whole recordings

Select signals in warbleR

Automatically detect signals with auto_detec

Use SNR to filter automatically selected signals

Optimize SNR measurements

Calculate SNR for automatically selected signals

Next vignette: Visual inspection and signal classification

Citation

Reporting bugs

References

Bioacoustics in R with `warbleR`

Parallel processing in `warbleR`

Make your data frame into an object of class `selection table`

Automatically detect signals with `auto_detec`