Compute minute-scale background sound energy from FLAC recordings using Python

Source: R/SoundEnergyByMinutePython.R

This function reads a sound file manifest in which each row contains paths to one to three sequential FLAC audio files from the same recorder and day. It uses Python's soundfile library via reticulate to extract audio samples directly from FLAC files.

SoundEnergyByMinute.py(
  manifest,
  flac_files,
  output_dir,
  start_filter = "0500",
  y_limits = c(2.3, 4),
  max_minutes = 180,
  parallel = FALSE
)

Arguments

manifest

A data frame containing at least:

  • path — relative path to FLAC directory;

  • file — primary FLAC file name;

  • subsequent.file1, subsequent.file2 — optional FLAC files;

  • area, year, group, date.mmdd;

  • plot, startTime.hhmm.

output_dir

Directory where JPEGs and CSVs will be written.

start_filter

Only rows with startTime.hhmm == start_filter are processed. Default: "0500".

y_limits

Numeric vector of length 2 giving y-axis limits for plots.

max_minutes

Maximum number of minutes to analyze per file set.

parallel

Logical; if TRUE, rows are processed in parallel using future.apply. You must have a plan set (e.g., future::plan()).

input_dir

Root directory containing FLAC files.

Value

Invisibly returns a data frame with columns: area, year, group, date, plot, startTime.hhmm, minute, log10RMSE, movAvg11.

Side effects:

  • JPEG plots written to file.path(output_dir, "noise_jpg_all").

  • group–year CSVs written to output_dir.

Details

It can run sequentially or in parallel (via future.apply) and includes basic error handling for unreadable or corrupted FLAC files.