Gramex’s director Tuomas Talonpoika examines the importance of metadata in his “Music and the Metadata” -series.
In the field of recorded music, copyright organisations and collective management organisations (CMOs) are undergoing a rapid transition towards becoming international data businesses.
A variety of data types, including both phonogram usage data and rights owner data, are received, processed, and subsequently transformed into a substantial volume of diverse data outputs for the benefit of phonogram rights owners, particularly in relation to compensation and royalty transactions.
One continually expanding factor and commodity in the collective management of phonogram music is the metadata of recorded music.
In the music industry, the term “metadata” is becoming increasingly prevalent, along with other identifiers such as the ISRC code, IPN number and ISNI code.
These identifiers are commonly referred to as Music IDs and typically form part of a larger metadata set. It is often assumed that these identifiers are the same as music metadata. However, in practice, music metadata is much more than just individual identifiers.
It is likely that most readers of this article have encountered what is commonly referred to as music metadata in a variety of contexts related to recorded music and its utilisation. It is essential that all those involved in music production and music ecosystems are aware of the meaning, importance and basic rules of music metadata.
Metadata is a fundamental aspect of the modern music industry, regardless of whether an individual is a professional or an amateur participant.
Concurrently, the commercial value of this music metadata is increasing significantly.
New forms of use and situations that serve the end users of recorded music and music rights owners in general, enabled by technological development, are creating new markets.
Marketing, which is also undergoing digitalisation and rapid growth alongside the development of artificial intelligence, and social media platforms that are constantly evolving, are seeking to make even more efficient use of the metadata that accompanies recorded music.
In technical terms, metadata refers to a range of information embedded in an audio or audiovisual file in digital form. This information is about the content of the file in question or about other things connected to the content in various ways, such as information about rights holders or content styles.
There are numerous ways to classify metadata, but a common model is as follows:
– Descriptive metadata includes information such as the title of the song, the main artist, or the release date of the track on the recording.
– Ownership metadata includes information on the original phonogram producer rights, for example.
Descriptive and proprietary metadata is typically unique information. It is typical for there to be only one title and one set of information about it, or one release date, per track.
The third category of metadata is that of recommendation metadata. By way of illustration, the genre or style description of a song constitutes an example of recommendation metadata.
This metadata is also used to connect the track in question with various recommendation algorithms for end users.
It is also worth noting that the term “finding/searching metadata” is increasingly being used to describe one of the main categories of metadata. This allows users to search for tracks connected to various algorithm functions, either by themselves or by the service itself.
Another frequently utilised classification model for metadata is as follows; the metadata categories can be broadly classified into four main groups: Basic Metadata (e.g. title of song), Technical Metadata (e.g. ISRC code), Creator Metadata (e.g. producer and musician information) and Marketing Metadata (e.g. various publication information and genre).
Metadata can be used to communicate information to end users, for example the identity of the band or artist performing the song.
The metadata is then displayed in an engaging and informative manner for the end user of the recorded music. For instance, it can be seen on the display screen of a music streaming service, where the currently playing artist’s display name (the main artist’s name shown on the streaming service’s display screen) is visible.
Well-structured metadata facilitates the discovery of phonograms on music streaming services. DSPs (Digital Service Providers) have prepared comprehensive instructions and guides to assist in the creation of metadata.
Currently, the most efficient and high-quality use of metadata is concentrated in the DSP sector. However, there is potential for significant growth in the future, particularly in the back office activities of collective management organisations.
The use of recorded music is becoming increasingly prevalent in a variety of settings.
Concurrently, the number of phonogram tracks reported to the copyright organisation is also experiencing a notable increase. This results in a considerable number of phonogram tracks undergoing various identification and alignment processes.
This shift in focus will result in a transformation of the former manual work into a technical process, carried out with the help of technology. Data will be processed and analysed automatically, with the support of various AI-assisted functions.
Metadata is used by organisations in their day-to-day processes when users of recorded music report their use of recorded music.
The organisation’s metadata and associated identifiers can be used to identify the correct recording and the relevant rights holders, for example for the purposes of remuneration and royalty distributions.
It should be noted that the so-called “recorded music usage data” reported by users of recorded music is not directly usable data for the calculation of copyright compensation paid by organisations.
In practice, copyright organisations maintain their own “production track databases,” which often contain more detailed information about the right owners of each individual sound at that time through various technical links.
In certain instances, the original producer of the recording has been able to sell their production rights or the right to collect and receive compensation has been transferred or licensed to a third party. In such instances, the royalties are paid to a party other than the initial, or ‘original’, producer of the recording.
It is crucial to accurately identify the phonogram used by the radio in the initial stage of the usage data process.
To illustrate the above, let’s take an example. The renowned song “Jumpin’ Jack Flash” (writers: Richards and Jagger), made popular by The Rolling Stones, has over 90 recorded versions by The Rolling Stones alone.
The tracks currently have numerous producer rights owners, in addition to various musicians who frequently perform in the background of the band on different recordings. Typically, the studio versions of the song and track feature studio musicians who are specialists in that field, while the live versions feature musicians who are specific to the tour in question. On each new tour, there are often new musicians. For these reasons, it is crucial to have a definitive version in place at all times.
Regarding the accuracy of the metadata, it is important to note that different identifiers, including those that are commonly used, have taken on a new significance and purpose in technical processes.
An illustrative example of this is the utilisation of the ISRC code (International Standard Recording Code) for recorded music and music videos. In accordance with the international regulations governing the use of the ISRC code, each individual recording publication must be assigned a unique and independent code.
In practice, a new and separate ISRC code should therefore be allocated to each slight variation of the recording. The fundamental principle is that a basic remastering does not constitute a new recording in this context.
However, there are instances when this is also applicable. The most common scenario for the new ISRC code is when a band creates both a studio and live version of the same song. In this instance, each version must be assigned a unique ISRC code. It should be noted that the aforementioned 90+ recordings of The Rolling Stones’ “Jumpin’ Jack Flash” each have their own, distinct ISRC code.
It is also possible to encounter instances where the same recording is accompanied by different metadata.
One example of this is where a single song and phonogram has multiple ISRC codes, typically from different sources. This is an unusual occurrence, but not an impossibility.
Such cases and other instances of conflicting metadata can present significant challenges, particularly for collective management organisations. It is often unclear which metadata is correct. Such conflicts are sometimes combined with what are known as rights conflicts or rights disputes.
It is also crucial to consider the manner in which the metadata is initially created.
The information stored in the metadata for a single recording, such as Bon Jovi’s “It’s My Life (Acoustic Version)” and “It’s My Life – Acoustic Version”, may be presented differently due to the varying storage formats, even though it is the same recording.
It is possible that at some point in the future, this recording of “It’s My Life” by Bon Jovi may be combined with the recording of the same name by the artist Dr. Alban. The correct spelling and style can have a significant impact on the processing of data.
Nevertheless, sophisticated rights management systems and the diverse algorithms they employ should be capable of matching and identifying songs through means other than merely their names.
As data will increasingly be read and processed solely by machines and technological solutions in the future, accuracy and adherence to established standards will become crucial. It is possible that different writing and reporting styles of one and the same phonogram can be identified and eventually combined into the same phonogram using, for example, artificial intelligence functions. However, this increases work and costs, and at the same time the probability of various errors also increases. It is therefore crucial to adhere to the various standards and formats that ensure consistent use and utilisation of music metadata.
The most significant standard developed for the international transfer and delivery of metadata in the music industry is DDEX. At present, DDEX is the most widely used standard for transferring recorded music metadata to DSP services. However, the various DDEX standards are also being adopted by other sectors of the industry, including collective management organisations. These metadata standards enhance data uniformity and facilitate data portability across different systems.
As a result, there are also legal questions to be answered regarding liability for the accuracy of the metadata. It is of particular importance that the metadata is correctly ordered and prepared, especially in relation to the aforementioned finding and searching processes.
Once stored and forwarded metadata has been entered, it is very difficult to make changes afterwards. Furthermore, it is more challenging to withdraw metadata once it has been published and distributed. However, those who have financial rights to defend typically take care of their own legal safeguards.
A further key issue is that of establishing who can and should trust the metadata in question.
It is essential that the correct and reliable metadata is created at the source, in other words by the original producer of the recording. The original producer will typically also obtain the ISRC code from the metadata of the phonogram. The original producer is typically in possession of all the requisite knowledge and data, particularly with regard to the so-called “descriptive metadata.”
A key legal question is whether there is a legal obligation to ensure that the information stored in the metadata does not change or is not interpreted in a way that is contrary to its original content.
In practice, this has been secured with the different tools of contract law. This is also linked to the legal question of metadata ownership. As with any other economically significant factor in the music industry, music metadata can also be subject to piracy or other financial abuses.