The 3 Cs: Consistency
July 02, 2014 in tags by Dan Gravell
I first posted about the 3 Cs (completeness, correctness, consistency) back in 2011 and since then I refer back to both the concept and that blog post often.
That got me thinking: maybe a more in-depth treatment of the 3 Cs is required! So I've decided to put together three in depth blog articles about the 3 Cs, to be published over the coming weeks. This week we'll start with consistency.
This is the first in a series of blog posts about the 3 Cs: consistency, correctness and completeness.
Consistency comes in multiple forms when applied to digital music tags. Importantly, consistency doesn't just refer to the contents of your tags, it also affects how your tags are formatted and used.
Content consistency
Content consistency is probably the easiest to understand and the most obvious. Consider the tracks with the following ALBUM tags:
| File | ALBUM |
|---|---|
06.Let It Flow.mp3 | Pure Phase |
07.Take Good Care of It.mp3 | Pure Phaze |
08.Born Never Asked.mp3 | Pure Phase |
The difference in spelling in 07.Take Good Care of It.mp3 will likely split this album and mean playing the album "Pure Phase" misses out the aforementioned track. The ALBUM tag fields can be said to be inconsistent.
This is particularly galling and noticeable, but content consistency comes in more subtle forms.
An example of more subtle content inconsistency are artwork sizes. Some artwork in your collection may be high resolution, others lower. In many cases this may not present a problem, but if some software scales up lower resolution artwork, creating a blocky effect, or some software simply refuses to show artwork above a certain resolution, disappointment results.
Syntactic consistency
Syntactic consistency governs how textual tag values are formatted. In most cases (MP4s being an exception) tags are free text fields, so you can complete them as you like. There are numerous fields where this can cause problems.
I think the most common I see are formatting of number series, for example track numbers and media (e.g. disc) numbers. A basic example of this is track number padding; where track numbers are padded such that all tracks in an album are correctly ordered regardless of whether an alphabetic or numeric sort is applied.
TRACK | TRACKNUMBER |
|---|---|
| Let It Flow | 6 |
| Take Good Care of It | 07 |
| Born Never Asked | 8 |
Here, the track numbers are inconsistently formatted; track seven is padded. This will result in tracks being played out of order.
Gaining consistency of track number formats probably only matters within each album; it's only important to have a consistent track number format within the album because you only ever order tracks within the context of their containing release. Compare that with year and date formats.
Year and date tags are different because they pertain to a surrounding release, not the tracks within a release. This means year and date tags should be formatted consistently across your collection. For example, you may adopt a rule that a year tag should be the four digit year, not two digit, and not containing a more specific date.
Year and date tags' consistency should not just be considered with regards their formatting; there's also the semantic consistency to consider...
Semantic consistency
Year and date information tagged inside music files pertain to the time period associated with the surrounding release. But does that mean the year of release, or the year of original release? Consider remastered releases. If some such albums are tagged with the canonical release date of the remastered album, and some are tagged with the release date of the original album, that's an inconsistency that will annoy you when you browse your album by year.
This is an example of semantic inconsistency. This is where the meaning of the tag field differs from music file to music file and album to album. In the YEAR tag case we are applying two different meanings to the tag: one meaning the original recording date, one being the re-release date.
It's not just date and time information. Track artist tag fields are often used in two different ways. One is simply to store the artist information for the track contained within the music file. Another is to store artist metadata for the surrounding release. An example of the latter is with "Various Artists" compilation releases. This second approach is understandable because many music players do not support a separate ALBUM ARTIST tag.
It comes down to the old maxim: The music player is king. You need to model and adapt your music collection in a way that suits your music player(s). You must be pragmatic; it's ok to use a tag field in a non-industry standard way if that way is going to make your life with your music easier!
But the important thing is consistency. It may be appropriate to use a field in a kooky, not-strictly-correct way, but the important thing is applying that way through your entire music folder. That's consistency!
Thanks to kpwerker for the image above.
