Sound recording history

During the early period of sound recording, first with cylinders (Edison, 1877) and a little later with early discs (Berliner, 1887) and discs of the coarse groove era a performance or event was recorded as a one-off take during a recording session, and without the possibility to improve or edit the content of that particular recording.

To improve on a take an entire repeat performance was necessary. Each separate recording was, therefore, a separate sound recorded event. With the introduction of discs, matrix numbers were usually allocated by the recording company and often etched or stamped into the wax. The wax is the name given to the area between the grooves and the label on disc formats. Matrix numbers were used to identify master recordings. They were often added to the stamper during the production process, and are usually visible on the final pressing which was generally mass-produced for publication and sale to the public.

On occasions when another take was considered necessary to improve on the performance, a new matrix number would be allocated to the next disc, or a new take number added to the existing matrix number. Take numbers were applied incrementally and appear after the matrix number, sometimes preceded with a hyphen. As with matrix numbers, take numbers may be numeric or alphabetic.

Each recording company applied its own matrix and take number system, and without a good knowledge of the company in question, it is sometimes difficult to identify a matrix and take number. The following made up examples, however, illustrate some of the types of systems used by record companies to apply take numbers:


A1234-1 A1234-2 A1234-3
A1234A A1234B A1234C
1234 1234-A 1234-B
1234 1235 1236
1234 1234-x 1234-xx


Occasionally special symbols such as an x in a circle, triangle, or square have been used to indicate takes. (Note that with some record companies though, these types of geometric symbols, or other characters in circles located after or near the matrix number, were used to indicate the type of electrical system used for making the recording rather than the take.)

The best performance was then selected for publication and the corresponding matrix number (and possibly take number) would usually appear on the final product. With record companies selling and buying businesses, the record label and catalogue number on the different publications could change for the same recording over time. The matrix and take numbers (almost without exception) would, however, remain constant for that particular sound recording.

For discs produced during the coarse groove era then, the matrix and take numbers usually uniquely identify the particular recorded performance or event.

The opportunity to edit the audio content of a sound recording came as early as 1898 with wire recordings. However, the techniques, associated with editing wire were developed more extensively with the introduction of tape in 1934 (which became commercially available from 1937). Audio tape presented the possibility of editing by splicing for the first time. Today, there are many editing techniques: multi-tracking, patching, mixing and remixing the tracks (with the possibility of also including additional new tracks made in a later recording session).

Digital technology also allows sampling and a kind of 'virtual reality' by superimposing various recordings to create a new work through the ear and hand of the sound engineer.

The sound which is eventually fixed is what is then recognised as the recording. Another remixed or digitally remastered version, or an edited extract, used for instance as an audio clip in an interactive multimedia product, or as an example during a lecture, radio programme, interview, etc., is effectively a new recording for the purposes of identifying it usefully within an audiovisual collection.

Consequently, the usefulness of the matrix number to positively identify a particular sound recording has waned for post coarse groove era audio formats. Its usefulness to identify the side of an unlabelled disc through the production process has, however, continued through the microgroove era at least.

In the 1980s the International Standard Recording Code (ISRC) was developed by the International Federation of Phonographic Industries (IFPI) and standardised as ISO 3901:1986. This code is similar to ISBN in structure and purpose and is designed to uniquely identify each individual recording. It has been implemented in many countries. Other standard numbering systems under development at the time of writing and which are relevant to audiovisual archives are: International Standard Work Code (ISWC) to identify "any musical composition from songs to symphonies to advertising jingles"; and the International Standard Audiovisual Number (ISAN) to identify "audiovisual works from television shows to films to multimedia product … transmitted to worldwide audiences via broadcast, satellite, cable, on-line services and the Internet". These are being developed by International Confederation of Authors' and Composers' Societies (CISAC) as part of the Common Information System (CIS). Unique identifiers for persons as contributors to sound recordings (i.e. as authors or performers) are also under development.

Such numbering systems will be essential metadata components within fully automated archives, such as digital mass storage systems, but are not so vital to the traditional approach addressed by these rules. In any case, the numbers are intended to be digitally encoded on the recording and will therefore be invisible to the human eye without specialised decoding software. However, some record companies have printed ISRCs on the packaging, and these should be recorded in Area 8, e.g. DE-A14-93-514-00.