What is mastering?
Mastering is the process of 'finalising' a track or a group of tracks ready for duplication on CD, Vinyl or any other audio format.
This firstly involves processing the audio of a finished mix to make it sound as good as possible on a wide range of stereo systems; to give it a 'shared' character if it is part of a group of tracks (such as an album or EP) by ensuring the volumes and Eqs of the tracks are similar and to bring the volume up to a level which gives the most impact without sacrificing dynamic range or introducing unwanted distortion.
Secondly, when working with a group of tracks for release, it involves arranging those tracks in order with the correct gaps or fades between them, inserting track metadata such as ISRC codes, track titles and so on and burning a master CD to a 'red-book' industry standard.
So why can't I do most of this in the mix?
In the days of vinyl, mastering was a necessity. The audio would be fed from a tape machine into a special lathe, like a reverse record player, and cut directly into an acetate master which would be used to stamp metal negative discs, these would then be used to stamp the final records. Due to the limitations of vinyl (in terms of small dynamic and frequency range), the audio would have to go through some processing before reaching the cutting machine. Too much bass, or bass out of phase between left and right channels could cause a stylus to jump out of a record. Too quiet a level and a recording could sound quiet and scratchy. Because of this, some compression was often required. EQ compensation would be needed to help the sound of the record match the original recording as closely as possible.
With digital recording, this kind of processing is not absolutely necessary to produce a CD, and if needed, it is possible to do most of it in the master bus of a final mix. However, mastering is still done for CD production as a separate process, and mastering studios are always busy and command very high prices. There must be a reason for this. Firstly, in practical terms, if you were mastering a group of tracks for release on an album, it would be difficult to switch quickly between full mixes to compare the eq curve and volume of each one. It is far easier to do this, and to make the necessary alterations, when you are dealing with a group of finished mixes together. Also many high-end computer based mastering plugins require a great deal of processing power – using these alongside other effects on a mix can cause problems with slowdowns, audio latency and crashing.
Perhaps more importantly though, a mastering engineer will offer a fresh set of ears on a mix as a whole, rather than a collection of tracks. A good mastering engineer will have ears trained to hear small differences in eq and volume between tracks, the know-how to smooth over the aspects of a track that don't sound quite so good, and to enhance the aspects that do. He or she will also use expensive specialist hardware and software dedicated to the mastering process.
What kind of processing is normally used on the audio and why?
Normally in mastering, processing tools will be used that will EQ a track, alter a track's volume in real time (dynamics processing), enhance the stereo image of a recording, and ensure (when mastering for digital formats) that the level never goes beyond the maximum permitted. Sometimes these tools are combined - for instance a 'multiband compressor' will split the track into different frequency 'bands' and apply independently controllable dynamic processing to each one. Mid-Side stereo tools will allow an engineer to separate the 'centre' of a mix from the left and right, and to process them seperately, giving a great deal of control over a track's stereo image.
What is a compressor / limiter / dynamic processor and why would it be used in mastering?
Generally, mastering engineers will only use compression sparingly as it can have a big impact on the sound of a mix. A mastering engineer's job is to enhance the sound of a track, rather than to mess with the balance of a mix.
'Compression' seems like a counter-intuitive term – the word describes volume being squashed down, yet the result of using a compressor is normally louder sounding audio. This is because actually 'compressing' a sound in volume is only the first part of what a compressor does- the second part is to re-amplify that 'compressed' signal.
To illustrate this, you can imagine you are listening to a piece of classical music on a hifi with your hand on the volume control of the amplifier. The music has some very quiet passages and some loud passages. You set the volume to a comfortable level for listening to the quiet passages, then a loud passage comes along and you react by quickly turning the level down so the volume is kept the same as before. The quiet bit comes along again, and you once again turn the volume back up to match. The actual volume of the piece you have experienced has been the same throughout, despite the dynamics of the original audio varying. This is what the first part of a compressor does - continuously monitoring the level of audio and adjusting its amplification. Most compressors have parameters for adjusting the speed of the reaction to volume changes as well as how loud the volume has to reach before making any adjustments (known respectively as 'attack', 'release' and 'threshold'. Now imagine that you the output of the amplifier you have been adjusting hooked up to a second amplifier. The volume control on this is set to a constant level of your choosing. If you turn it up, while still 'compressing' the first hifi, the overall level of the piece of music will sound consistently louder. On a compressor, this is known as the Output Gain stage. A compressor that reacts at a very fast speed to changes in volume, and always turns the volume down to a flat level beyond a certain point, preventing the level of audio from ever exceeding this point is commonly known as a Limiter. Any automatic processing of the volume of audio this way in hardware or software is known as 'Dynamic Processing'.
Compression can be used in mastering to increase the overall apparent volume of a track without pushing it beyond the level at which it will distort on analogue or digital playback equipment. It does this by basically making the loud bits softer, like a hand on a volume control, and bringing the result back up to a consistant level.
My mixes never sound as loud as stuff I hear on commercial CDs. Can mastering fix this? Is compression / limiting involved?
Yes, and sometimes. In recent years there has been a lot of controversy about record labels and bands competing with each other to produce the 'loudest' CD. Obviously, in audio, 'loudness' is a relative term. No matter how 'loud' the level of audio information is (up to the maximum) on a CD, the listener can always turn the volume up or down on a stereo. So why all the fuss? Firstly, it is important to remember that the dynamic range (range between quietest and loudest sounds) on a CD is wide, but finite. The race to squeeze the absolute maximum volume from this limited dynamic range through use of audio limiting and other tools probably started with the idea that if one record label could get its track to sound louder on the radio than another, it would have more impact on listeners, be heard from further away and sell more copies. It likely also came from the feeling that louder music sounds better - you can hear more detail in a track at higher volume, and it carries more emotional impact. Because of the maximum volume limit on a CD (if it goes beyond this you end up with horrible distortion), mastering engineers were increasingly asked to use their tools to give the illusion of volume by squeezing the entire dynamic range of a mix into a smaller, loud area. This means less range between loud and soft sounds on a mastered track which often leads to less clarity and impact in a track (how can there be loud with no soft to compare it to?) In some cases it even leads to distortion, when mastering for volume has been pushed to extremes. Most of the time the public will only ever hear the end result and not realise that excessive pushing of volume up to this limit could be having a detrimental effect on music, which infuriates many people who know the extent to which audio is being frequently distorted by the mastering process.
On the other hand, loudness of a mix is a creative choice for an artist or a label, and a mastering engineer will and should do what he or she is paid to do. Sometimes, 'brick walling' of a track is an inherent part of its sound- particularly in hiphop, dance and heavy rock, and the race to create ever louder tracks has helped to define these genres over the last twenty years.
There are also tools available now (which I use) that can enhance apparent volume of a track using a number of methods besides compression / limiting. This means that a track can be made to sound louder whilst retaining the apparent dynamic fluctuations between loud and soft (thus preserving the impact of quick 'transient' sounds such as snare drums). These tools generally work by introducing a certain amount of overdrive-type distortion to specific frequencies in a mix, and use complicated algorythms to control dynamics.
Ultimately the loudness of a track is a decision that should be left with the artist. If I am asked to master a track to make it as loud as possible, I am happy to do that. If I am asked to use my own judgement, like most mastering engineers I will err on the side of caution and bring the volume to a level consistent with commercial releases of a similar genre, but no louder.
My mix has horrible boxy vocal sounds / thin drums / the guitar is too loud and I can't afford more studio time. Can I fix this with mastering?
Mastering should really be used to enhance what is already good in a track rather than to fix problems. This is mainly because when you cover over one problem in a mix, another will usually appear as a result (like air bubbles in wallpaper). Having said that, there are a lot of things that can be done to a mix to widen, fatten, hide or bring out specific parts, take out certain frequencies and even tune individual notes in harmonies / guitar chords (though this is not advised). You can do a lot, but there are never any guarantees that it will make a mix sound better, and mastering shouldn't be seen as a 'magic bullet'. I only charge if a client is happy with the result (see Mastering Services) so you have nothing to lose by giving it a go.
- Jez Burns