Since my post on this subject from last year, a few things have happened which have made me rethink about the way I have organised my music collection. In this post, I will go over the changes, and what I used to try to improve my music library.

Server

I was becoming frustrated with Airsonic, with its frequent downtime and sluggishness. I decided to look towards Navidrome, which uses the same streaming protocol, and is in active development.

The changeover was a little difficult as there was no package I could simply install, like I did last time. Instead, I followed these instructions on how to set it up via Docker running on the secondhand Synology NAS I was storing my music collection on.

I created a Shared Folder to store my FLACs, populated it, and simply pointed to it via the instructions.

Listening Clients

I recently found out that SoniXD was no longer being developed, with development moving to Feishin instead. I am not too sure about Feishin just yet, as it is missing many features I like, but I will keep an eye on it. I also looked into Supersonic, a new desktop client for Subsonic protocols, but also found it to be very early days in development.

Substreamer Web

In the meantime, I have started to just use the Navidrome interface from within a web browser, though I recently saw a post on noted.lol about a web client version of Substreamer that caught my eye. It was simple to install on the NAS alongside Navidrome, and quickly loaded the collection, fetching artist information from last.fm including artist image, a biography, top songs, and similar artists.

However, I was unable to edit playlists, so I’ll still need to go back into Navidrome to do that, though aside from that issue, it’s a good “desktop” client for now, which can, of course, be used through any device which doesn’t have a dedicated app and can access a web browser, which might come in handy.

Symfonium

A few months ago, I discovered Symfonium, an Android app accessing music servers running the Subsonic protocol, in a Reddit post. The developer is still regularly updating it, and there is a community forum for feedback and requests too.

Although a paid app, it had a generous free trial, and I managed to save up enough Play Points to get it with a good discount. It’s a beautiful application with highly customisable features and great practicality, it also fetches some data from the internet to improve the listening experience, such as biography and top songs, though artist images and similar artists would also be nice. Furthermore, as with many applications I’ve encountered, I cannot edit playlists on it, which is a shame.

Tagging

I decided with this migration I was to finally sort out my music tags once and for all. It was a tedious process as while there are applications that will automagically fix the tags for you, I have faced two main issues with these in the past.

1. As the majority of the music we are talking about in this collection is Japanese and Korean, the tagging applications, to be fair, did the right thing by changing all the tags into the correct Hiragana/Katakana/Kanji for Japanese or Hangul for Korean. However, I don’t know these languages and finding or identifying the artists, albums or songs I want to listen to would be not only be difficult to recognise, but impossible to search for quickly as my keyboards are in English. As such, I would want the tags to be in a form of Romanisation or an English translation. And I would have to do this by hand.
2. In the past, I have found some tagging applications to be kind of useless with their genres, with some being “Asian Music”, “International” and “World” - not very helpful. As genres vary amongst the tracks within an album, I actually decided to do away with genres, as they were more bother than they were worth. If, however, I could find a decent tagging application with good genre listings, I might consider giving that a go.

MusicBrainz Picard

I decided to use the free MusicBrainz Picard application to manually fix any issues with my tags, not using any of their automatic features. It was a very slow part of the process, but meant the tags were rendered how I liked them. I do think they have decent automatic tagging when needed, I had to hop to the website a couple of times to grab the details of an album when the tags were completely blank, and the edited them afterwards.

File Names

But I had another issue. While the folder structure was relatively neat, with [artist] → [album], the file names themselves were a bit messy, and this affected my playlists, as I could not write the .m3u3 files knowing exactly what the file names would be, as they could be any variations of the following:

01 [song title]
01 - [song title]
01_[song title]
01 [artist] - [song title]
01 [artist] - [album] - [song title]

While MusicBrainz Picard has an export option, it didn’t work out too well for me, so I looked elsewhere.

beets

I heard of beets a long time ago when browsing SynoCommunity, but I was put off my the non-graphical user interface. However, I had since become more used to working within the terminal, and decided to give it a go, as it has an export function.

beets can actually do so much more, and is incredibly extensive in its capabilities in automagically tagging music. Again, not something I want, but something others should consider if they are going to reorganise their music collection.

After setting up my config as so:

directory: /Volumes/Drive/Organised
library: /Volumes/Drive/musiclibrary.db
import:
    copy: yes
    move: no
    write: no
    autotag: no

I simply then ran beet import -A /Volumes/Drive/Fixed to import my tag approved music collection into the “Organised” folder, with a structure based on [artist] → [album] again, and files as 01 [song title].flac

The really neat thing is that if I acquire new music, I can just edit the tags how I want in MusicBrainz Picard, and then run the command again for beets to do its magic and add the music into the library with a perfect file structure.

I then, in turn, copy the music off my local device and onto the NAS for Navidrome to automatically add to the collection, ready for streaming. Although beets can run on my NAS (again, that’s how I found it!), I felt more comfortable running it locally as I am editing my tag locally, and then could view the local versions of the files as a backups from those on the NAS.

Conclusion

With the exception of a stable fully featured desktop client, and a couple of features I would like added to Symfonium, I am relatively happy with this current set up. I love how easy it is to keep organised with a quick adjustment of tags and then running a short command for beets to do its work, and then easily accessible to listen to via Substreamer Web and Symfonium.