266 lines
7.6 KiB
Python
Executable File
266 lines
7.6 KiB
Python
Executable File
#!/usr/bin/env python3
|
|
|
|
# multithreading
|
|
import multiprocessing
|
|
|
|
# audio format conversions
|
|
import ffmpy
|
|
|
|
# argument parsing
|
|
import argparse
|
|
|
|
# multiprocessing stuff
|
|
from multiprocessing import Pool
|
|
|
|
# executing some commands
|
|
import subprocess
|
|
|
|
# file/directory handling
|
|
import os
|
|
|
|
# most recent starttime for program
|
|
import time
|
|
|
|
# randomness
|
|
from random import randint
|
|
|
|
# typing hints
|
|
from typing import Any, Optional
|
|
|
|
import tempfile
|
|
|
|
# working with sound files
|
|
import soundfile
|
|
|
|
# loudness normalization
|
|
import pyloudnorm
|
|
|
|
# file copy
|
|
import shutil
|
|
|
|
"""
|
|
Normalize loudness of all music files in a given directory and its subdirectories.
|
|
"""
|
|
|
|
musicfile_extensions = (".flac", ".wav", ".mp3", ".m4a", ".aac", ".opus")
|
|
|
|
|
|
def loudnorm(inputfile: str, outputfile: str):
|
|
"""
|
|
Normalize audio to EBU R 128 standard using pyloudnorm
|
|
|
|
Parameters:
|
|
inputfile (str): Path to input file. Format must be supported by python-soundfile module
|
|
outputfile (str): Path to output file
|
|
"""
|
|
data, rate = soundfile.read(file=inputfile)
|
|
|
|
# measure loudness
|
|
meter = pyloudnorm.Meter(rate=rate)
|
|
loudness = meter.integrated_loudness(data=data)
|
|
|
|
# normalize audio
|
|
file_normalized = pyloudnorm.normalize.loudness(
|
|
data=data, input_loudness=loudness, target_loudness=-30.0
|
|
)
|
|
|
|
# write normalized audio to file
|
|
soundfile.write(file=outputfile, data=file_normalized, samplerate=rate)
|
|
|
|
|
|
def ffmpeg_to_wav(inputfile: str, outputfile: str):
|
|
"""
|
|
Convert a file into .wav for further processing
|
|
|
|
Parameters:
|
|
inputfile (str): Path to input file
|
|
outputfile (str): Path to output file
|
|
"""
|
|
|
|
# convert to wav in temporary directory
|
|
with tempfile.TemporaryDirectory() as tempdir:
|
|
# temporary input file
|
|
temp_input: str = os.path.join(
|
|
tempdir, os.path.splitext(os.path.basename(inputfile))[0] + ".wav"
|
|
)
|
|
|
|
# temporary output file
|
|
temp_output: str = os.path.join(
|
|
tempdir,
|
|
"normalized",
|
|
os.path.splitext(os.path.basename(inputfile))[0] + ".wav",
|
|
)
|
|
os.mkdir(os.path.join(tempdir, "normalized"))
|
|
|
|
# convert audio to wav
|
|
ff = ffmpy.FFmpeg(
|
|
inputs={inputfile: None}, outputs={temp_input: None}, global_options=("-y")
|
|
)
|
|
|
|
subprocess.run(ff.cmd, shell=True, capture_output=True)
|
|
|
|
# normalize loudness
|
|
loudnorm(inputfile=temp_input, outputfile=temp_output)
|
|
|
|
# convert audio back to lossy format
|
|
outputcmd = {
|
|
outputfile: "-c:a libopus" " " "-b:a 192k" " " "-compression_level 10"
|
|
}
|
|
|
|
ff = ffmpy.FFmpeg(
|
|
inputs={temp_output: None}, outputs=outputcmd, global_options=("-y")
|
|
)
|
|
|
|
subprocess.run(ff.cmd, shell=True, capture_output=True)
|
|
|
|
|
|
def ffmpeg_copy_metadata(inputfile: str, outputfile: str):
|
|
"""
|
|
Copy all metadata from the input file to the output file.
|
|
A temporary file is used in an intermediate step
|
|
|
|
Parameters:
|
|
inputfile (str): Path to input file
|
|
outputfile (str): Path to output file
|
|
"""
|
|
|
|
# store output file as temporary file. FFMPEG can't work on files in-place
|
|
with tempfile.NamedTemporaryFile() as temp_audio:
|
|
shutil.copyfile(outputfile, temp_audio.name)
|
|
|
|
inputcmd = {inputfile: None, temp_audio.name: None}
|
|
outputcmd = {outputfile: "-map 1 -c copy -map_metadata 0"}
|
|
|
|
ff = ffmpy.FFmpeg(inputs=inputcmd, outputs=outputcmd, global_options=("-y"))
|
|
|
|
subprocess.run(ff.cmd, shell=True, capture_output=True)
|
|
|
|
|
|
def main(inputfile: str) -> Optional[list[Any]]:
|
|
"""
|
|
Main program loop
|
|
|
|
Parameters:
|
|
inputfile (str): Path to input file
|
|
|
|
Output:
|
|
dynamically normalised audio files (list)
|
|
"""
|
|
|
|
# set output folder to parent path + "normalized"
|
|
outputfolder = os.path.join(os.path.dirname(inputfile), "normalized")
|
|
|
|
# NOTE create output folder
|
|
# because multiple parallel processes are at work here,
|
|
# there might be conflicts with one trying to create the directory although it already exists
|
|
# this while loop makes sure the directory does exist
|
|
# the try/except block ensures the error is caught and (hopefully) doesn't happen again just after with random sleep
|
|
# there's very likely a better way to do this, idk
|
|
while not os.path.isdir(outputfolder):
|
|
try:
|
|
os.mkdir(outputfolder)
|
|
except:
|
|
time.sleep(randint(0, 4))
|
|
|
|
# output file path
|
|
infile_noextension: str = os.path.splitext(os.path.basename(inputfile))[0]
|
|
infile_extension: str = os.path.splitext(os.path.basename(inputfile))[1]
|
|
|
|
match infile_extension:
|
|
case ".flac" | ".wav":
|
|
print("Working on", inputfile)
|
|
outputfile: str = os.path.join(outputfolder, infile_noextension + ".flac")
|
|
# direct conversion start
|
|
loudnorm(inputfile=inputfile, outputfile=outputfile)
|
|
ffmpeg_copy_metadata(inputfile=inputfile, outputfile=outputfile)
|
|
print("Completed", inputfile)
|
|
case ".mp3" | ".m4a" | ".aac" | ".opus":
|
|
print("Working on", inputfile)
|
|
outputfile: str = os.path.join(outputfolder, infile_noextension + ".opus")
|
|
# conversion is started within the ffmpeg_to_wav function
|
|
ffmpeg_to_wav(inputfile=inputfile, outputfile=outputfile)
|
|
ffmpeg_copy_metadata(inputfile=inputfile, outputfile=outputfile)
|
|
print("Completed", inputfile)
|
|
case _:
|
|
print(
|
|
inputfile,
|
|
"does not use a known extension. This error shouldn't be happening actually",
|
|
)
|
|
return
|
|
|
|
|
|
if __name__ == "__main__":
|
|
"""
|
|
Handle arguments and other details for interactive usage
|
|
"""
|
|
|
|
# start time of program
|
|
starttime = time.time()
|
|
|
|
parser = argparse.ArgumentParser(description="")
|
|
|
|
# Input directory
|
|
parser.add_argument(
|
|
"-i", "--input-dir", required=True, type=str, help="Input source directory"
|
|
)
|
|
|
|
# number of cpus/threads to use, defaults to all available
|
|
parser.add_argument(
|
|
"-c",
|
|
"--cpu-count",
|
|
required=False,
|
|
type=int,
|
|
help="Number of cpu cores",
|
|
default=multiprocessing.cpu_count(),
|
|
)
|
|
|
|
# in case you wanted to rerun the conversion for everything
|
|
parser.add_argument(
|
|
"-r",
|
|
"--reset",
|
|
required=False,
|
|
action="store_true",
|
|
help="Rerun conversion for all files",
|
|
)
|
|
|
|
args = parser.parse_args()
|
|
|
|
srcfolder = args.input_dir
|
|
|
|
cpu = args.cpu_count
|
|
|
|
reset = args.reset
|
|
|
|
# file where last run timestamp is stored
|
|
timefile = os.path.join(srcfolder, "run.time")
|
|
|
|
# list of non-linear normalizations
|
|
nonlinear_all: Optional[list[Any]] = []
|
|
|
|
# get time of previous run
|
|
if reset:
|
|
timeprev = 0
|
|
elif os.path.isfile(timefile):
|
|
with open(timefile, "r") as file:
|
|
timeprev = file.read()
|
|
else:
|
|
timeprev = 0
|
|
|
|
musicfiles: list[str] = []
|
|
for root, dirs, files in os.walk(srcfolder):
|
|
# ignore the "normalized" subfolder
|
|
dirs[:] = [d for d in dirs if d not in ["normalized"]]
|
|
for file in files:
|
|
if file.endswith(musicfile_extensions):
|
|
filepath = os.path.join(root, file)
|
|
# only file newer than the last run are added
|
|
if os.path.getmtime(filepath) >= float(timeprev):
|
|
musicfiles.append(os.path.join(root, file))
|
|
|
|
with Pool(cpu) as p:
|
|
nonlinear_all: Optional[list[Any]] = p.map(main, musicfiles)
|
|
|
|
# write this run's time into file
|
|
with open(timefile, "w") as file:
|
|
file.write(str(starttime))
|