As the main aim of mastering is the preparation of material for the manufacture and distribution of a mass market medium, well now take a look at the various types of media that the industry has used, past and present, so that we can better understand how the process of mastering came into being.
Until the 80s, vinyl was the dominant consumer medium, and little has changed in this area since the 50s when the first stereo, long playing vinyl disks were released.
A vinyl disc
In the studio, tape was the preferred medium, and so the audio from 15ips (inches per second) stereo ¼ tape mixes were transferred and cut by lathe onto a master disk or lacquer by the cutting engineer, before making master copies from this for manufacturing. The audio waveform of the song is emulated by the spiral groove cut into a record, and the movement of the stylus follows this. To make sure that the amplitude of the groove wiggle is reasonably constant across all audio frequencies, the lower frequencies (below 500Hz) are reduced and the higher frequencies (above 2kHz) are boosted when the groove is cut. This means that when the record is played back, the opposite characteristics must be applied, i.e. the lower frequencies must be boosted and the higher frequencies reduced. This process was first adopted in the late 50s when the Recording Industry Association of America (RIAA) agreed an international standard - this is why the input to an amplifier for a record deck is often termed the RIAA input, as it contains the EQ circuit necessary to restore the frequency characteristics of the original audio. Because of the physical limitations of vinyl, it was important for the cutting engineer to make sure that both the EQ and dynamics did not make the record distort or jump when trying to reproduce any large peaks or transients, and it was from this process that the mastering engineers role evolved into what it is today.
The compact cassette system was developed by Philips in 1963, originally for recording speech, but it became popular for music once the obstacles of limited bandwidth and poor signal-to-noise ratio had been overcome to a reasonable degree.
The compact cassette
Bandwidth was increased by the use of special magnetic tape formulations, including high-density ferric oxide, chromium dioxide and pure metal compounds. These tapes also allowed signals to be recorded at higher levels, and so went some way to improving the signal to noise ratio, but this issue was only really solved when Dolby Laboratories licensed a consumer version of its noise reduction technology. Their Dolby B system dramatically improved the sound quality of compact cassette tapes, enabling them to rival vinyl. The ability to make Dolby encoded home recordings was also a very attractive feature of the system, and certainly contributed to the widespread acceptance of the compact cassette, sales of which overtook vinyl in 1983. This was only short lived though, as both media were usurped by the Compact Disc shortly after.
With the arrival of the CD, new digital post-production mastering methods were adopted, including the editing of mixes in a DAW (Digital Audio Workstation). The modern procedure involves a finalised master recording being stored on digital media such as an Exabyte 8mm tape cartridge, in the case of an album, along with metadata regarding the track separation, track numbers, and track length. From the tape, a glass master is created, and from this all subsequent copies are taken to produce the completed CDs you see in the shops.
CDs themselves consist of a clear polycarbonate disc, usually 12cm in diameter and 1.2mm thick. The audio data is pressed into the aluminised reflective side of the disc to form the data layer. This is covered in a coating of protective lacquer, onto which the disc label is printed. To play the CD, an infrared laser is focused through the polycarbonate underside to read the audio data stored as pits and lands (see below) in a spiral track spaced 1.6um apart (1um = one millionth of a metre).
A CDs pits and lands
Although CD is still the dominant consumer media for music, MP3 players such as Apples iPod, along with the internet, have provided a way to distribute songs and albums that can cut out the traditional mastering, production and distribution stages. Armed only with ordinary popular sequencing software, you can now mix, master, and upload your material to a website for all to access. Also, low-cost mastering software such as Steinbergs WaveLab and Apples Waveburner enables you to edit and compile your own tracks, as in a mastering studio, and then burn these directly to CD in your computer. Here though, the similarity to the professional mastering process ends, as Exabyte tape is specially formulated to produce virtually no bit-errors, unlike the average CD-R. Even at 1x burning speed, the number of errors can exceed 200 per second, unlike a glass-mastered CD that undergoes bit by bit verification, producing a much more acceptable 20 to 30 errors per second.
Bit-errors aside, doing your own mastering still requires the right knowledge to make use of the tools, and the mastering engineers techniques still apply, which we shall consider next. Probably the most important post-production task is the treatment of the overall dynamics of both individual tracks, and whole albums, as the songs may vary in instrumentation and structure, differing greatly in dynamic content. As we have seen, this can then make it difficult to transfer the audio to the final master medium. To overcome this, an engineer will apply gain reduction with a dynamics processor such as a compressor or limiter: