Reading MP3 metadata with Python

Music is one of my real passions. If I'm not listening to something, I'm probably singing something to myself all the time. But for the longest time, I never had the time (or money) to have a better listening experience. Since I'm at home, and the pandemic is not over, I decided to understand a little bit better what (and how) I'm listening to music - with Python :]

Quick summary on how I listen to music

So, as millenials do, I have grown listening to cassette tapes, CDs, and eventually MP3s. I remember being 15, using a huge ass pair of headphones (real cheap ones too) with my 128MB MP3 player - and it was awesome.

With time, I grew my MP3 collection, both through good channels (like bandcamp for example), and less... legal channels (like Napster, Kazaa, and Souseek, whic I still use now and then). I can't say my library is big, but it's around 4200 musics in size.

The problem is that a good chunk of it is garbage; not in quality of the music itself, but in the quality of the file compression. And now that I have the money to get some better headphones, some of my music library is just too dank.

As with anything, Python can help me.

Enter mutagen

mutagen is a Python library that helps us to find metadata for our audio files. After discovering it, discovering which MP3 files were too compressed was pretty straightforward.

First, I installed it locally (in my case, it was already installed somehow. I probably fiddled around with it in a distant past):

pip3 install --user mutagen

Then, I wrote a quick and really dirty Python script to tell me which of my MP3 had a bitrate lower than 256kbps. The higher the bitrate, the less it's compressed:

import os
import mutagen


def list_dank_mp3s(folder):
    matches = []
    for root, dirnames, filenames in os.walk(folder):
        for filename in filenames:
            if filename.endswith(".mp3"):
                # I didn't take the time to understand 
                # why it was formatting things wrong
                # so I just hammered the names
                matches.append(os.path.join(root, filename)[:-1] + "3\n")

    bad_mp3s = []
    for item in matches:
        # here, we get the bitrateof the MP3 file
        # there are a LOT of other metadata there,
        # if you want to use it
        bitrate = mutagen.File(item.strip()).info.bitrate
        # if the bitrate doesn't end with 000, it means it's most likely variable
        # if it is, we leave it be. if not, we check if it's below 256kbps
        if str(bitrate).endswith("000") and bitrate / 1000 < 256:
            bad_mp3s.append(os.path.join(os.path.dirname(os.path.abspath(__file__)), item))
    return bad_mp3s

This function gives me a list, that has the paths of the MP3s that I may want to buy or redownload in a better bitrate. I then write the list to a file, and voilà, all my bad sounding MP3s are listed.

---

Do you have any other uses for this kind of library? Do you have any tips on good headphones? Send me a comment down here so we can keep the discussion going! Thanks for your attention, and see you next time :]