Music Library Management Chapter Three: Organising - files and folders
November 06, 2012 in digital music by Dan Gravell
It's a truism that the more music you collect, the more it needs to be organised. If you have a small collection there's less chance of inconsistencies creeping into your music library and it is easier to find particular music, whether or not it has been accurately recorded with genre, year of release and similar information.
These mislabelings and inconsistencies affect many aspects of your music collection. I'll discuss tags and album art in future posts, but this serialisation of the Music Library Management e-book covers music files and folder paths.
This is the third in a series of posts serializing the Music Library Management ebook. I'm posting the same content here to make it easier for people to find the information via Google.
Files and folders
The inevitable building blocks of your music library are the files and folders that contain your music. Importantly, because files are the building blocks of your music library, file locations identify your music within music players and so it's important to get your file and folder structure correct from the start. Otherwise, changing the location of files and folders later may lose information linked to the original location of files.
'Files' are a computing concept that have long been used as general storage method for all kinds of data. Music files are computer files in the same way your personal photographs, documents and spreadsheets are. Typically, one file is stored per music track. They are stored within your computers' 'filesystem' which makes them generally accessible to different applications, such as music players. As discussed in chapter two, they can be stored on a computer's hard drive, externally, or remotely on a NAS. Music files are obtained through methods described in Chapter one such as CD ripping or buying and downloading online.
Files are collated within folders. This isn't strictly necessary, but when you consider that you will have one file per track, that would be a lot of files in one folder for even a modest collection. It's easier to collect music files together in folders, named by artist and album. The organisation method by which your tracks are collected together is called your music folder structure. Here's an example music folder structure:
artist/album/tracknumber-trackname
Which would give us (for example):
Blur/Parklife/04-Parklife.flac
When we talk about organising music files and folders, it's this folder structure that we are concerned with. In general, the structure should be imposed to be the same throughout your music library. That said, you may want to vary the rules slightly with compilations which involve different artists.
Why music file and folder names remain important
Back in the 'stone age' early days of digital music, files and folders were actually the only way to identify a track or album of music. Since then, 'tagging' (storing textual metadata within music files, see the section below) has overtaken file naming as the main way of identifying and categorising music. However, file naming still remains important for various reasons.
The first reason file names are important is that music players typically use them to identify music. When you add music to a music player, the music player records the internal tags and the source file for the music, so that when you press Play on Dark Side Of The Moon, the file containing the music for the first track, Speak To Me, is played. In addition, the music player may store its own data related to the file, such as the number of times you play it, or the rating you assign to it.
If you change the names of music files or folders, this identifier vanishes, and so the music player will likely fail to associate the old data with the newly named file. This may or may not be a problem, but it's worth noting. Another place where stable file and folder names become important are playlists which are a list of music files by location. If you change these locations, your playlists will point to non-existent music files.
In addition, file level music browsing is still important when exporting music to different devices, or backing music up. There are applications specially built to synchronise portable music players, but some people still prefer doing it at "a file level", in other words: manually copying the files and folders to the portable music player.
Sources of problems
The source of a lot of problems in digital music organisation is... the source (of the music). That is, because music is sourced from different places the file paths they are delivered in are often inconsistent.
Further problems stem from the source of the metadata ("tags") for the music. This data is populated by different people to different schemes (if, indeed, there's a scheme at all). This is discussed later in more detail in the tagging section. However, because file organisation is generally formed from the tags, inaccuracies and inconsistencies (or absences) in tags means the file paths end up incorrect too.
Most problems relating to music file organisation stem from the synchronisation between tags and filenames. Firstly, whether they are synchronised at all, and secondly whether they continue to be synchronised when tags are changed.
Some tags are changed more than others. For instance 'genre' tags, denoting the genre of a piece of music, is generally changed more often than a band, track or album name. If music files are organised by tags that change often, it's more likely that your file organisation will not reflect internal tags.
However inconsistencies are introduced, the result is a folder of music files that are harder to navigate than they should be. Some albums in lower case titles, some with "The" switched to the end of artist names ("Beatles, The"). This makes synchronising files with your music player more difficult and also using music players that rely on filenames.
Recommendations
My key recommendations for organising music files relate to the problems outlined above. In addition, I would classify music file organisation as one of the few things that is imporant to get correct from the start.
Why is this? The reason is that, as I wrote above, music players depend on file and folder names and use the names to identify music. This means if you decide to change these files and folders later you may lose associated data within your music player. Not a big deal, necessarily, but something to consider.
Another reason is if you have duplicated 'mirrors' of your music collection elsewhere. For instance, you may have copied your library to another hard drive or music player. If this is the case, and you change the filenames in one collection, synchronising the libraries is a pain. A simple combination of the two collections will lead to duplicate music: one with the old file names, one with the new.
So my first recommendation is: choose a file structure (using the further recommendations below) and apply it rigorously.
Synchronise with tags, and keep them synchronised
There are two potential sources for music metadata in your collection. The most widely used and most flexible are the textual tags stored in music files. However, the file and folder paths of your music collection also communicate music information.
Importantly, these two sources should be synchronised. If tags are changed and the files and folders are not, this divergence will make it more difficult to find and navigate your music library. There are limits to the synchronisation, for instance file and folder names typically have rules that pertain to what characters are permitted, but a consistent mapping from tag to file and folder name is still required.
Once they are synchronised, they need to stay that way, otherwise over time the two sources of information will drift apart and it will get more and more difficult to resolve the differences and use your music collection. This can be done by periodically running file and folder renaming routines, or using software that automatically detects any changes and does it for you.
Less is more
The fewer bits of metadata you have in filenames, the better. This is because, for all the problems outlined above, the fewer metadata you have in your file organisation the less likely you are to be bitten by one of the problems.
For instance, the fewer tags you decide to have in a music folder structure, the less likely you are to suffer when you change some tags. If you decide to reorganise your genres and you have wisely decided not to include them in your music folder structure, that's less work for you.
Aim for just 'identifying' aspects for each file you name. Just album name, artist name, track name and number should be enough.
Don't use tags likely to change
Leading on from the above, if you must include extra metadata, make it metadata that doesn't change. Another example of metadata that is unlikely to change, but isn't 'identifying' metadata, is the YEAR tag for an album (denoting the year of a recording).
GENRE, as an example, is more likely to change because the description of a genre and its assignment to music is much more wooly. On the one hand, you may have a set of genres which you want to enforce within your music collection. As your collection grows in different ways this set of genres may change, which means re-assigning certain music to different genres over time.
For file organisation, it's safer to use tags like YEAR than tags like GENRE because the former is less likely to change, which means changing a lot of file paths.
Consider whitespace
This one's a trade-off. The advantage of retaining spaces in file and folder names is that the names look more natural so it's easier to browse your music. The alternative is to replace whitespace with a common character, e.g. the underscore '_'.
Replacing whitespace with, say, underscores is useful when you want to write scripts (such as, in the Windows world, batch or Powershell scripts) to automate some aspect of your music library. If your file or folder names have spaces in them, typically you must employ extra syntax (such as surrounding file paths with quotes) which leads to harder-to-read and write scripts. Potentially, if you aren't careful about how you treat whitespace, scripts can go wrong and lead to unwanted side-effects.
Mind your case
The case of your file and folder names is largely up to you, and it's probably best for readability if they follow the case in the album/artist tags within your music files.
It is worth noting, however, that in a networked music environment with different operating systems case sensitivity can vary from device to device. Windows computers are generally case insensitive, whereas Unix based computers such as a NAS running Linux are case sensitive. This means the NAS could create two folders of music files with the same name but in a different case and the Windows computer would only be able to see one. This is an unlikely situation, but when these incompatibilities occur they can be surprising and difficult to track down.
Thanks to Tambako the Jaguar for the image above.
