It would be great if someone could test if the audiotools backend still works after these changes. I haven't got that backend to work on my system so haven't tested it myself.

Nice! This looks great already. If we can confirm that the audiotools backend still works, I think we should merge!

Nice! This looks great already. If we can confirm that the audiotools backend still works, I think we should merge!

Okay, I tried to get the python audio tools backend to work again. However i noticed that I get the same error in the master branch. I don't know if I'm doing anything wrong or if this is a bug. I've used my changes without problems using the bs1770gain backend. Should I create a bug or am I doing something obviously wrong?

command: beet replaygain --force

system: Arch Linux

audiotools-config:
Default verbosity level : normal

extraction arguments
Format Readable Writable Default Quality
────── ──────── ──────── ───────────────
aiff yes yes
alac yes yes
au yes yes
flac yes yes 8
m4a yes yes 100
mp2 yes yes 192
mp3 yes yes 2
ogg yes yes 3
opus yes yes 10
tta yes yes
wav yes yes
wv yes yes standard

Format : %(track_number)2.2d - %(track_name)s.%(suffix)s
Default simultaneous jobs : 1
Add ReplayGain by default : no

ID3 arguments
ID3v2 tag version : ID3v2.3
ID3v2 digit padding : not padded ("1", "2", …)
ID3v1 tag version : ID3v1.1

CD lookup arguments
Use MusicBrainz service : no
Default MusicBrainz server : musicbrainz.org
Default MusicBrainz port : 80

      Use FreeDB service : no
   Default FreeDB server : us.freedb.org
     Default FreeDB port : 80

system arguments
Default CD-ROM device : /dev/cdrom
CD-ROM sample read offset : 0
CD-ROM sample write offset : 0
Filesystem text encoding : utf-8
Audio output : PulseAudio, ALSA, NULL

beets config:
plugins: replaygain

replaygain:
backend: audiotools

Nice! This looks great already. If we can confirm that the audiotools backend still works, I think we should merge!

Okay, I tried to get the python audio tools backend to work again. However i noticed that I get the same error in the master branch. I don't know if I'm doing anything wrong or if this is a bug. I've used my changes without problems using the bs1770gain backend. Should I create a bug or am I doing something obviously wrong?

command: beet replaygain --force

system: Arch Linux

audiotools-config:
Default verbosity level : normal

extraction arguments
Format Readable Writable Default Quality
────── ──────── ──────── ───────────────
aiff yes yes
alac yes yes
au yes yes
flac yes yes 8
m4a yes yes 100
mp2 yes yes 192
mp3 yes yes 2
ogg yes yes 3
opus yes yes 10
tta yes yes
wav yes yes
wv yes yes standard

Format : %(track_number)2.2d - %(track_name)s.%(suffix)s
Default simultaneous jobs : 1
Add ReplayGain by default : no

ID3 arguments
ID3v2 tag version : ID3v2.3
ID3v2 digit padding : not padded ("1", "2", …)
ID3v1 tag version : ID3v1.1

CD lookup arguments
Use MusicBrainz service : no
Default MusicBrainz server : musicbrainz.org
Default MusicBrainz port : 80

      Use FreeDB service : no
   Default FreeDB server : us.freedb.org
     Default FreeDB port : 80

system arguments
Default CD-ROM device : /dev/cdrom
CD-ROM sample read offset : 0
CD-ROM sample write offset : 0
Filesystem text encoding : utf-8
Audio output : PulseAudio, ALSA, NULL

beets config:
plugins: replaygain

replaygain:
backend: audiotools

And the error:
replaygain: analyzing Pink Floyd - The Wall - In the Flesh?
Traceback (most recent call last):
File "/usr/bin/beet", line 11, in
load_entry_point('beets==1.5.0', 'console_scripts', 'beet')()
File "/usr/lib/python3.7/site-packages/beets-1.5.0-py3.7.egg/beets/ui/init.py", line 1267, in main
_raw_main(args)
File "/usr/lib/python3.7/site-packages/beets-1.5.0-py3.7.egg/beets/ui/init.py", line 1254, in _raw_main
subcommand.func(lib, suboptions, subargs)
File "/usr/lib/python3.7/site-packages/beets-1.5.0-py3.7.egg/beetsplug/replaygain.py", line 1052, in func
self.handle_track(item, write, force)
File "/usr/lib/python3.7/site-packages/beets-1.5.0-py3.7.egg/beetsplug/replaygain.py", line 1003, in handle_track
track_gains = backend_instance.compute_track_gain([item])
File "/usr/lib/python3.7/site-packages/beets-1.5.0-py3.7.egg/beetsplug/replaygain.py", line 752, in compute_track_gain
return [self._compute_track_gain(item) for item in items]
File "/usr/lib/python3.7/site-packages/beets-1.5.0-py3.7.egg/beetsplug/replaygain.py", line 752, in
return [self._compute_track_gain(item) for item in items]
File "/usr/lib/python3.7/site-packages/beets-1.5.0-py3.7.egg/beetsplug/replaygain.py", line 781, in _compute_track_gain
rg_track_gain, rg_track_peak = self._title_gain(rg, audiofile)
File "/usr/lib/python3.7/site-packages/beets-1.5.0-py3.7.egg/beetsplug/replaygain.py", line 764, in _title_gain
return rg.title_gain(audiofile.to_pcm())
File "/usr/lib/python3.7/site-packages/audiotools/mp3.py", line 217, in to_pcm
return MP3Decoder(self.filename)
TypeError: argument 1 must be str, not bytes

Aha! If this is broken on master, it's probably just a bug (or, more accurately, an API change in the library we're invoking). If this seems to work otherwise, I think it's ready to merge—and we can talk about fixing the error in a separate issue.

Aha! If this is broken on master, it's probably just a bug (or, more accurately, an API change in the library we're invoking). If this seems to work otherwise, I think it's ready to merge—and we can talk about fixing the error in a separate issue.

Nice! I have used it frequently and it has worked well for me. I have created an issue at #3305 for the audiotools error.

Awesome! Thank you for your help with this!!! 🚀 ✨