Last active
May 21, 2022 03:06
-
-
Save hkva/d9313a119b1c858ab93765d932bddf3f to your computer and use it in GitHub Desktop.
Python script to rip music from YouTube. Includes MP3 metadata, normalizes loudness, and embeds album art
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
#!/usr/bin/env python3 | |
import json | |
import math | |
import os | |
import sys | |
import PIL.Image as img | |
import pyloudnorm as pl | |
import soundfile as sf | |
import yt_dlp as yt | |
# What we want to do: | |
# 1. Accept a video or playlist URL DONE | |
# 2. Extract audio DONE | |
# 3. Normalize audio DONE | |
# 4. Embed mp3 metadata DONE | |
# 5. Embed album art DONE | |
# 5. Save as <ARTIST> - <TITLE>.mp3 DONE | |
# Audio normalizer | |
class NormalizerPP(yt.postprocessor.PostProcessor): | |
def __init__(self, loudness): | |
yt.postprocessor.PostProcessor.__init__(self) | |
self._loudness = loudness | |
def run(self, info): | |
fp = os.path.join(info['__finaldir'], info['filepath']) | |
data, rate = sf.read(fp) | |
meter = pl.Meter(rate) | |
loudness = meter.integrated_loudness(data) | |
self.to_screen( | |
f'Normalizing loudness from {loudness} to {self._loudness}') | |
corrected = pl.normalize.loudness(data, loudness, self._loudness) | |
sf.write(fp, corrected, rate) | |
return [], info | |
# Create square album art | |
class AlbumArtPP(yt.postprocessor.PostProcessor): | |
def run(self, info): | |
# Find thumbnail path | |
thumb_name = next( | |
(t['filepath'] for t in info['thumbnails'] if 'filepath' in t), | |
None) | |
if thumb_name == None: | |
self.to_screen('No thumbnail could be found') | |
return [], info | |
# Open thumbnail | |
fp = os.path.join(info['__finaldir'], thumb_name) | |
im = img.open(fp) | |
# Center crop | |
leftcrop = math.ceil((im.width - im.height) / 2) | |
im = im.crop((leftcrop, 0, im.width - leftcrop, im.height)) | |
self.to_screen(f'Resized to {im.height}x{im.height}') | |
# Save | |
im.save(fp) | |
return [], info | |
class RenamePP(yt.postprocessor.PostProcessor): | |
def run(self, info): | |
newpath = info['filepath'] | |
if 'track' in info and 'artist' in info: | |
track = info['track'] | |
artist = info['artist'] | |
ext = os.path.splitext(info['filepath'])[1] | |
newpath = f'{artist} - {track}{ext}' | |
else: | |
self.to_screen('No metadata') | |
return [], info | |
# Remove path delimiters from name | |
newpath = newpath.replace('/', '-') | |
newpath = newpath.replace('\\', '-') | |
# Remove windows-unfriendly characters from name | |
newpath = newpath.replace('?', '') | |
self.to_screen(f'Renaming to {newpath}') | |
if not os.path.exists(newpath): | |
os.rename( | |
os.path.join(info['__finaldir'], info['filepath']), | |
os.path.join(info['__finaldir'], newpath) | |
) | |
info['filepath'] = newpath | |
else: | |
self.to_screen('File already exists, skipping') | |
return [], info | |
if __name__ == '__main__': | |
if len(sys.argv) != 2: | |
print(f'Usage: {sys.argv[0]} <URL>') | |
exit(1) | |
ytdl_opts = { | |
'format': 'bestaudio', | |
'writethumbnail': 'True', | |
} | |
with yt.YoutubeDL(ytdl_opts) as ytdl: | |
ytdl.add_post_processor(yt.postprocessor.FFmpegExtractAudioPP( | |
preferredcodec='wav' | |
)) | |
ytdl.add_post_processor(NormalizerPP(-15)) | |
ytdl.add_post_processor(yt.postprocessor.FFmpegExtractAudioPP( | |
ytdl, 'mp3' | |
)) | |
ytdl.add_post_processor(yt.postprocessor.FFmpegMetadataPP( | |
ytdl | |
)) | |
ytdl.add_post_processor(AlbumArtPP()) | |
ytdl.add_post_processor(yt.postprocessor.EmbedThumbnailPP( | |
ytdl | |
)) | |
ytdl.add_post_processor(RenamePP()) | |
ytdl.download([sys.argv[1]]) | |
# info = ytdl.extract_info(sys.argv[1]) | |
# print(json.dumps(ytdl.sanitize_info(info), indent=4, sort_keys=True)) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment