Skip to content

Latest commit

 

History

History
110 lines (89 loc) · 2.89 KB

README.md

File metadata and controls

110 lines (89 loc) · 2.89 KB
Raw

Split WAVE audio based on selection table

Andrew Schwenke 29 October 2022

This is a small example workflow using the warbleR and RRaven packages. It takes a WAVE file and Raven selection table as input, and returns one WAVE file (.wav) for each selection, as well as one Raven selection table (.txt file) per selection. The practice data (CC1_20171010_125500.wav) is an audio recording taken at Cedar Creek, south-east Queensland, Australia, and released by Andrew Schwenke under a CC BY 4.0 license.

Input:

Input

Output:

Output

Prepare the data

library(warbleR)
library(Rraven)
library(dplyr)
# with no file tag added, set name from file to true, and provide upper or lower WAV extension format
rvn.dat.1 <-
        imp_raven(
                all.data = TRUE,
                path = "data/typical/",
                name.from.file = TRUE,
                ext.case = "lower"
        )

# rename columns to use with warbler
rvn.dat.2 <-
        rvn.dat.1 %>%
        dplyr::rename(
                selec = Selection,
                start = `Begin Time (s)`,
                end = `End Time (s)`
        )

# generate a selection table
rvn.dat.st <- selection_table(rvn.dat.2, path = "data/typical/")
## checking selections (step 1 of 2):

check_sels(rvn.dat.st, path = "data/typical/")
## all selections are OK

Generate segmented WAVE files

Margin is the time in seconds to add to generated WAVE files before/after each selection. Margin = 0 would generate WAVEs that are exactly the length of the selection annotations.

margin <- 2
cut_sels(
        X = rvn.dat.st,
        path = "data/typical/",
        mar = margin,
        dest.path = "output/cuts/typical"
)
# check if Delta time (s) exists
# create if it doesn't
# modify data frame to reflect what will be the new start and end times for each selection when the WAVE is split 
if(!"Delta Time (s)" %in% colnames(rvn.dat.2))
{
        rvn.dat.3 <-
                rvn.dat.2 %>% dplyr::mutate(`Delta Time (s)` = end - start)
        
        rvn.dat.4 <-
                rvn.dat.3 %>% dplyr::mutate(start = 0 + margin,
                                            end = (start + `Delta Time (s)`))
} else {
        rvn.dat.4 <-
                rvn.dat.2 %>% dplyr::mutate(start = 0 + margin,
                                            end = (start + `Delta Time (s)`))
}

Generate individual selection tables for each cut WAVE

# list the cut WAVE files you generated
cut_wavs <-
        list.files(path = "output/cuts/typical/", pattern = ".wav")
# generate selection tables
for (x in 1:length(rvn.dat.4$selec)){
exp_raven(X = rvn.dat.4[x,], path = "output/cuts/typical/", file.name = cut_wavs[x])
}