The almost unstoppable rise in the access model of music use has presented music companies with a fundamental shift in how people listen to music. Although pay-per-track digital stores continue to increase sales year-on-year, relatively new streaming services dominated the headlines in 2009 and will probably do so for the beginning of this year.
However, in all this, little consideration has been given to recorded core-classical music (recordings of orchestras and traditional classical artists rather than crossover classical releases). As a genre it has several elements that require special consideration, particularly when creating the essential “easy to operate” digital interface.
The first stage of digital distribution based on per-track and album purchasing has helped bring core-classical music to a new audience. Able to browse in private, away from the often intimidating environment of the traditional music store, consumers can explore at their own pace. Digital stores are also more able to reflect people’s listening habits. In traditional retail, genres are segmented, not because that is how people listen to music but because that is how the store needs to organize products so people can find what they are looking for.
As physical retailers have increasingly reduced slow-selling, low-margin core-classical music in favor of other entertainment products, only the best-selling core-classical titles are left in stores. Although all core-classical recorded music could be made available online, this is not enough to secure sales. Except for searches for specific titles, browsing music can be more restrictive online than in a physical store – a consumer can see only what is displayed on his monitor at any one time, and if digital music services do not have helpful links, charts, submenus and subgenres, freely browsing a genre without significant background knowledge is almost impossible.
To continue to function, small and artist-owned classical labels need to make new recordings. However, with high recording and production costs coupled with limited marketing budgets and income from new digital services, competing with the larger music companies is difficult. The larger players are easily able to repackage works with almost no production costs and market them to new audiences that aren’t looking for specific versions of works. Moreover, as the aging traditional core-classical consumer base shrinks, the core-classical audience of the future will become younger. With their different approach to the consumption of music, for core-classical music to survive it must be made available on new services.
The online market might be attracting new listeners, but they are often unconcerned with a specific recording of a particular work and will listen to the first one they find. However, smaller music companies are able to compete with the potential influx of back-catalog content from larger labels through upselling. Since deep-catalog items are recorded by artists either deceased or not under contract, upselling beyond the actual copy is difficult. Moreover, a community can be created around an active artist who is connecting to his audience through the many channels now available.
There are several issues associated with making available coreclassical music on access-based services. Advertising-supported services are enduring a period in which they are suffering from their own success – the more consumers use a service, the greater the required payments to rights holders and the more costly they become to operate. Only when advertising and premium subscription revenues cover costs will these services become viable. However, to begin with, these services must attract a large number of users to draw potential advertisers. This requires them to obtain a catalog made up of the most popular genres. Core-classical is often an afterthought. For services built with metadata structured for mainstream genres, ingesting core-classical content has significant challenges. Whereas in pop music there is typically one version of an album, correctly representing each of the hundreds of recordings of classical works is problematic.
Different recordings can have the same orchestra and conductor, and therefore the standard metadata fields are not enough to handle core-classical music. For database-powered recommendation services, basing recommendations on metadata to guide consumers is unreliable. For example, a look at the most played tracks of Beethoven over the past six months on Last.fm lists the fourth-most-played track as Allegro Con Brio. Symphony movements mostly do not have titles, just tempo directions. Therefore, Allegro Con Brio is found in many works. Only the title field is visible in the chart, and it is unable to identify and aggregate different versions of the same piece by recording. This reliance on metadata is a significant challenge for services using core-classical music, especially when it is used as the basis of content organization. Advertising-funded services often play audio advertisements periodically. Aimed at popular-music fans, the tone is often complementary to the music, and since most popular music is heavily compressed, the volume usually matches the volume of the advertising. In core-classical music, however, the dynamic range is far greater, making the ad sound significantly louder than the music around it. Another fundamental difference is that although in pop music one piece is often one track, core-classical pieces can be several tracks. With track length varying from only 10 seconds to almost seven minutes, a service playing ads at the start of each track will be far more disruptive during core-classical. Since single pieces of music are split across many tracks, ads will also often be heard in the middle of a movement or piece of music.
Although music-recommendation services are already considered vital for introducing consumers to new music, their ability to introduce and guide people through the vast world of core-classical music will become one of the most important tools for gaining new audiences. Internet radio with music-recommendation engines enables consumers to explore without having to create a track-by-track playlist, therefore not requiring existing knowledge. As well as metadata issues, database services are also subject to virtuous and vicious circles. Since the most popular acts will be both listened to the most and tagged the most, they will be more prevalent in playlists. Music needs listens and tags to appear in playlists but needs to appear in playlists to be able to gain significant listens and therefore tags.
For example, a look at the most popular artists on Last. fm shows that the top-performing artists are all extremely high-profile acts with recent releases. When services are chart-driven, with the most popular music the most visible and most exposed, the least popular music will remain so. Without a fully functioning service that can support core-classical’s service-unfriendly metadata and tag functionality, it will always be confined to the furthest reaches of the tail at the expense of the heavily marketed and exposed hits. Two variations of the algorithm music-recommendation system have different implications for core-classical music. The algorithm of Pandora is human-based – experts listen to each track and record the specific attributes. The algorithm of Mufin uses computers to automatically analyze each piece of music. Although all services need to ingest content, the extra processing that Pandora needs to undertake manually to attach attributes to each track means additional costs incurred beyond hosting and maintenance. A problem that an automated service has is that using audio to generate recommendations means results are often split across genres and can seem unfocused and almost random in a playlist.
With so many problems and difficulties associated with accommodating core-classical content in second-stage digital music services, opportunities have arisen for tailored core-classical music services. Classical specialist a la carte stores are already established, since many popular download stores are not ideally suited to core-classical music. These services are often prepared to work with smaller organizations and have also increased capabilities. Although it is important to have services that serve core-classical music as best as possible, there are wide-ranging consequences. The first stage of digital distribution placed core-classical music alongside other genres, and although this meant less-than-ideal functionality, it exposed a new audience to core-classical music. However, many younger music consumers are becoming more familiar with on-demand access to content and are foregoing the more-traditional music-consumption route. For these music fans, the continued proliferation of feels-like-free services will leave direct paid-for services seem like the alternative rather than the norm. This has serious implications for core-classical music. If second-stage digital services cannot cater to core-classical music in a way that is financially viable for the service, there is a risk that the move by the mainstream recorded-music market toward the access model will result in core-classical music’s remaining in the pay-per-download model and becoming out of touch with how people listen to music. The ability to reach new audiences is vital for the longterm survival of the recorded core-classical music market, and having core-classical music placed only in specialist stores will limit its reach. This will alienate potential audiences and only serve to increase the stereotypical view of core-classical music as being different from other music, elitist and inaccessible to people below a certain age.