Categories
Classical Music Business

How Apple Music Ruined Classical Streaming

Classical music on streaming is a mess

Everyone who pays attention to classical music knows that its presentation on most streaming music services (with the exception of the specialist services IDAGIO and Primephonic) is, in a word, ghastly. This is old news, and has been written about endlessly.

Of course the reason is that streaming services followed the old MP3 standard of supporting only three fields for each track, quite logical for most pop/rock releases: artist name, album title, and track title. Considering that classical music requires composer name, work title, movement name, and one or more individual performers with specific roles (as a basic minimum), it’s difficult to see how to map one set of data to the other. So the streaming platforms were blamed for being inadequate to the task. It was all their fault, and they needed to conform to the needs of classical music aficionados, despite classical music making up only about 1% of the music market. (Need I point out the folly of fighting against tech companies that are constantly optimizing for the 90% use case?)

Apple to the rescue (they thought)

But some streaming platforms, and Apple’s iTunes in particular, tried to listen to the classical labels, and hired musicologists to help sort out the mess. They added capabilities in the form of new classical-related fields and drafted metadata standards and naming conventions to address the problems. And the results were… they made things immeasurably worse.

Wait, what are you talking about? If Apple hired real musicologists and listened to the classical labels, how could they have made things worse? I thought they were the ones leading the charge to fix everything?

In some ways, they are indeed improving things: Apple requires the composer to be entered, with that role, into the “contributor” field, to make sure it’s included and to distinguish it from other artists. They now also require each performer to be entered with their individual roles as well (conductor, pianist, etc.). They also support, and may soon require, additional classical-specific fields such as “work title” and “movement”, and perhaps even some additional details sectioned into their own discrete fields like “key” and “opus number”. These would then comprise the exact fields needed to properly represent most classical metadata.

But most of those fields are not widely supported yet let alone displayed in all contexts, and in the meantime Apple did one other thing that might also have seemed helpful but in reality was not: set standards for the naming of album and track titles. And this is where they destroyed the presentation of classical music on streaming.

How infections spread

Now before I go on, I should point out that not only did Apple thereby ruin the presentation of classical music on their own platform, but they also ruined it on all the other platforms all over the world. Why? Because distributors don’t want to deal with different sets of rules for different DSPs, and labels don’t want to have to upload separate releases for each platform either. So any distributor or label that distributes classical music to Apple Music (which they of course want to do, since Apple is one of the bigger outlets) is highly likely to use the same rules for all other DSPs, and hence the release will follow Apple’s rules everywhere (since Apple’s may be the only rules pertaining to classical, or at any rate are the strictest). A lot of those platforms do not support the “composer” role (or display anything entered into it) or recognize the “work title” field. But they do end up effectively following Apple’s naming conventions for album and track titles, because that’s what gets uploaded to them.

It all came from bats

So what did Apple do, exactly, that was so bad? Here are the two rules in question, one just a bit annoying and silly, and the other profoundly damaging:

  • Album title. Apple says that an album should “include the composers, followed by a colon, followed by the work titles, catalog numbers, or type of works such as Sonatas or Preludes.” Sounds sensible enough, in the singular case that the album contains just a single work by a single composer. Otherwise, this makes no sense at all, as it doesn’t actually provide any helpful information about the specific track we’re looking at. (Plus, it actually sounds like a recipe for some kind of library cataloging system, not for album titles in the music industry. And seems entirely unnecessary on a digital streaming platform, as we can just as easily search directly for individual tracks as search for that information in album titles. Plus that the primary purpose of album titles on streaming is not necessarily as a way to find music, but rather as a label of what you’re already looking at.)

    How does it help to say that the album contains works by several different composers (or many different composers)? Especially if 3/4 of the album title is cut off and never seen, and most of those composer names might not even display? (Remember that often it will be displayed on a small mobile phone screen, or else in a column next to other columns of information, and the field may have a small and fixed number of characters displayed.) It could even be misleading: “Piano Sonatas of Beethoven, Schumann, and Brahms”, but only displaying “Piano Sonatas of Beethoven,”, and the track you’re looking at is by Schumann. In some contexts the album title isn’t displayed at all. The exceptions where this rule is unhelpful dwarf the situations where it might make sense. This makes it obvious that it shouldn’t be a “rule” in the first place.

    In short: the album title field cannot be relied on to convey either the composer’s name or the name of the work on the track we are looking at (or listening to), let alone which movement it is. So why create a rule that such things should be listed in the album title? Apple should have immediately binned “album title” from their list of fields they cared about, and let labels name the album whatever they wanted. The part of their style guide that addresses the album title for classical releases is a distraction and a waste of space, and should be removed.
  • Track titles. Here’s the killer. The album title rule was just silly and useless, but the track title rule is actively destructive. The latest version of the style guide doesn’t say this explicitly, but all the examples make it clear (and various distributors enforce the rule thusly): it is forbidden to include the composer name in the track title.

    So, why would this be problematic? Shouldn’t we put that in the “contributor” field, marked with the “composer” role? Yes, we should. And that’s all fine and dandy and wonderful, except for the rather un-trivial detail that Apple Music themselves do not always display the composer field, and many streaming platforms do not recognize it at all, or ever provide a way to discover its contents.

    The “composer” role of the “contributor” field is a wonderful thing, but it is unreliable. On Apple Music, the “top songs” listing and playlists do not show the composer role. Spotify has similar limitations. Many other platforms don’t even ingest it or provide a way to access it. So, that’s nice, go ahead and add that field and require that it be populated, but at the same time pretend that it doesn’t exist. Because half the time it doesn’t.

    In the real world, anything in the composer role might as well not exist. Thus, the result of excising composer names from track titles is endless lists of classical tracks on all platforms with no indication of who the composer is. Sometimes they display in the album title, but not always, as we just discussed: it’s a lucky coincidence if looking at the album title will reveal the composer of that particular track. Maybe half the time it does, half the time it doesn’t. On some platforms, there is thus no way to ever discover who the composer of most classical tracks is (if they followed Apple’s naming conventions).

    It is difficult to overstate how problematic this is, how much of a disservice it is to both classical music lovers and people with no knowledge of the field. The composer name must be present, always and in every context, closely linked with the name of the work; in fact, they should be considered as an inseparable pair, always conjoined and traveling together.

The reality is that the album title and composer role are both unreliable and cannot be counted on to be informative as to the composer, or even to display in all contexts. Even if we were to put composers in as primary artists (which would be weird and confusing and give a lot of wrong ideas), some platforms only display the first artist listed, which is problematic enough when it comes to performers, but disastrous if it ends up omitting the composer (or conversely, includes the composer but none of the performers).

But the track title is the one thing that always displays, and it thus has a specific role: to fully identify what specific piece of music we are listening to, as succinctly as possible. (The “succinctly” part is because only a small number of characters may be displayed.) Since, as has been pointed out (adequately I hope), it is the only field we can count on to contain this information, perhaps we could consider not intentionally forbidding one of the most mandatory pieces of information from being present here, in the one place where it must exist?

Irrelevant gibberish

Furthermore, Apple’s rules state that the key and opus number must be included in the track title, prior to the movement name. Why, pray tell? Because the rules were written by an anal-retentive pedant with no conception of how this field would be used in the real world? Does anyone care about such information? It is almost always redundant and irrelevant, with a few exceptions where it’s needed to distinguish two otherwise identically-named pieces by the same composer. Otherwise, all it does is clutter things up and make it difficult to read, and, more to the point, takes up valuable characters of which we have precious few to work with. By all means, require them in the “work title” field (or better yet, relegate them to the “key” and “opus number” fields), but not in the track title! Often including the key and opus number means the movement number or name disappears off the end, and then we really can’t tell what we’re looking at. Now not only do we have no idea who the composer is, we also have no idea of the movement.

Let’s say these tracks are on an a two-disc album containing a half dozen French string quartets, each by a different composer. We may now have twenty-some-odd track names that all look exactly the same (aside from some key and opus numbers, which will be meaningless to 99% of even classical music buffs), and tell us nothing but that they are each some movement (not sure which) from some string quartet by… somebody. Thanks for that. Why not just name them all “a classical song, by anonymous” and be done with it?

Honestly, a lot of people just give up. A number of distributors won’t distribute classical to Apple Music at all. Not always because they couldn’t support the metadata, but due to user demand: many of the artists and labels uploading to them rebelled against Apple’s inane rules, and instead uploaded their classical music in the “jazz”, “folk”, or “world music” categories, or similar. (Never mind that a lot of that music really does belong under “contemporary classical” or “new age” rather than traditional classical.) I’m stubborn enough that I want to upload my releases into the proper category, properly named, despite the fact that Apple’s rules make that impossible.

What to do?

Apple’s path forward, in my view, is pretty simple:

  1. Continue working to require the “composer” role and “work title” and “movement” fields for classical releases (in addition to each performer identified by their specific role). If you have all the information you need there, you can ignore the track title and display the more specific fields instead in the appropriate contexts. And you can use that information to provide better search, browsing, and suggestions.
  2. Immediately abandon any standards relating to classical album titles. Let labels call their albums whatever they want. But perhaps suggest that streaming “albums” don’t need to correspond to physical CDs, and suggest instead creating digital albums consisting of single complete works where possible, in which situations Apple’s present naming standard might actually make some sense (as a suggestion, not a rule).
  3. Immediately eliminate fixed requirements for classical track titles, but instead present the suggestion to use the most succinct form of “Composer: Work: Movement” that makes sense to uniquely identify the musical composition in that track using the least space possible. (Notably: keys and opus numbers should be as optional as composer first names or initials, and used only where necessary to prevent confusion.) Make it clear that the track title is a combination of the composer, work title, and movement name in a concise form that provides the needed information even if all other fields are absent (which half the time they are).
  4. Make an announcement that “we screwed up (because we listened to cloistered pedants who didn’t actually know what they were talking about in the context of presenting music to actual people on real streaming audio apps that have real user interfaces displayed on real devices), and we’re sorry, but we’re trying to help fix things now.” And make sure all the distributors know, so that they can promptly adjust their ingestion rules for classical releases, because right now the situation is just impossible. I don’t even want to try. to upload any music until this is fixed, and meanwhile I’m fighting with my distributor, when all they’re doing is carrying out Apple’s dictates.

Sadly, I’m not holding my breath. Anyone who so spectacularly and comprehensively mucked everything up seems unlikely to listen to logic or reason (or admit they were wrong). But I can’t look at myself in the mirror unless I at least give it a shot.

Leave a Reply

Your email address will not be published. Required fields are marked *