In the first article in this series we said it is essential to take into account the different stages of music production and their variables in order to reach the mastering stage with the best chance of tackling it successfully. In this way, mastering becomes the final polish on a great deal of prior work.
Global goals of mastering
- Make the content properly audible. Regardless of the playback system, the master should have appropriate loudness, intelligibility and sonic transparency.
- Achieve appropriate loudness levels. Although the aim is often to make them as high as possible—something widely done today from a commercial standpoint, though completely unmusical—we will need an appropriate program density, all without reaching saturation or degrading the mix, with a maximum ceiling of –0.3 dB.
- Identify and create the data that accompanies the audio: PQ points, ISRC codes, UPC/EAN and pre-emphasis
Stereo or stem mastering
Normally, mastering the tracks of a stereo program is the most common approach. However, in recent years, and given the homemade character that many mixes and masters have taken on thanks to cheaper technology, mixing by stems, or the parts that make up the whole (for example, drums, guitars, vocals, keyboards, loops), is left in the hands of the person responsible for mastering. Thanks to greater experience, that person can take a small step back (toward the mix) to correct things that are almost impossible to fix in a stereo master. However, this in turn means that the roles of the mix engineer and the mastering engineer overlap, and on low-budget projects they may even be the same person.
Mastering on the computer or with outboard gear
With today's fabulous advances in computing, what can be done exclusively on a computer is getting ever closer to a master based on a combination of analog and digital equipment.
Analog gear allows greater dynamic range, but in many cases at the cost of introducing background noise. In any case, processing with expensive analog equipment is very hard for the digital realm to surpass or replace. Systems such as the TC Electronic 6000 or Sadie are good examples. Digital setups that rely exclusively on internal plug-ins, and therefore share their calculations with the computer's main processor, leave the master without depth or body.
That is why major brands such as UAD, SSL Duende, TC Electronic Powercore or Focusrite Liquid Mix offer us specialized DSP-based (Digital Signal Processing) equipment, sometimes external (PCIe, or external units based on FireWire or Ethernet), so that the heavy calculations needed to achieve their amazing quality are done outside the computer.
The importance of the listening space
In home or semi-pro settings, a common mistake is not paying enough attention to the acoustics of the space where we do our listening. The same quality equipment sounds different in different rooms. Only a neutral listening space can help us make a sound judgment about the master we are working on. Parallel walls, monitors close to walls or corners, the top surface of a large mixing desk, poor listener positioning and inappropriate volume can all color and interpret, on their own, the listening we need to do of the master. A room determines its resonant modes—the frequencies that will stand out over others—based on its dimensions. If the dimensions are large, the first resonant modes will be at low frequencies and will become denser as the mode numbers studied increase. This is therefore particularly important in small rooms, since the first modes fall at audible frequencies that are spaced apart from each other.
Only frequencies up to 300 Hz are responsible for a room's resonant modes, given their wavelength.
- Axial: has to do with the reflections between two parallel walls, or between ceiling and floor. These are usually the most important.
- Tangential: the second most important. They involve two pairs of parallel surfaces: the four walls, or two walls plus the ceiling and floor of a room.
- Oblique: six reflecting surfaces come into play: four walls, floor and ceiling.
A room EQ can in no way substitute for good acoustics in the listening space. Since problems with hearing frequencies are tied to our position in that space, a small shift will make the EQ of little use, and it can even have a negative effect. Early reflections, flutter echo, modal resonances or inappropriate reverberation times can never be tamed with an EQ. Our search here should aim for balanced acoustics, a mix of "dryness" in the space and a reverberant ambience.
Monitoring: speakers
Speakers are, in terms of how they work and what they deliver, the most primitive element in the audio chain. Even expensive speakers will tend to color the signal because of how they work. So there is little we can do about it, other than get to know them thoroughly.
Active or passive? By default, a passive speaker system will be more expensive than an active one, since we need the power amp and the corresponding quality cabling. Passive speakers receive the signal from the amp and, by means of a passive filter, split the signal among their drivers. In an active speaker, the signal is split first and feeds two small amplifiers, each of which then drives the woofer or the tweeter. In quality speakers, this can result in greater listening fidelity.
Speaker placement
Correct speaker placement is essential in a mixing or mastering studio. With stereo monitoring, the L-R speakers should form a 60° angle with each other and the listening position. The height should match the level of our ears with the speaker's tweeter, around 1.20 to 1.50 m (roughly 4 to 5 feet). In turn, the stands they sit on should be as solid as possible and have minimal contact with other elements. They usually come with spikes, thus minimizing their contact with the floor. Placing them on top of the mixing desk's meter bridge is not recommended. Nor should they be close to the back wall or corners, since this will affect the low frequencies with a boost of up to 18 dB.
Headphones
Headphones should not replace speakers but serve as an additional tool for finding certain errors or details that may go unnoticed when listening on speakers. With speakers, the L signal reaches our L ear and, with a minimal delay, our R ear, and the R signal reaches our R ear and, with a minimal delay, our L ear. This doesn't happen with headphones, since they deliver sound directly to our ears without taking the surrounding space into account.
Metering
Phase Scope
In the previous article we discussed peak and RMS metering and their uses. Another very interesting measurement tool is one that lets us measure the phase of a stereo program (Phase Scope in Steinberg's WaveLab). With it we can see the phase and amplitude between the two stereo channels, as well as the spread of the signal between L and R. A rounded display of the signal indicates good balance. If the figure "leans" toward one side, the frequency balance, its level and the indicated side need to be corrected. Viewing the effect of, say, a reverb on a vocal is also particularly useful.
Phase correlation
The phase correlation display is also an interesting tool. It shows us, in real time, the phase relationship between the two sides of the stereo program. It is correct if it sits in the positive zone, whereas if it falls into the negative zone we must correct the phase of that side. High-end mixing desks also have this feature. Personally, I have sometimes seen an audio interface output a signal that is out of phase between its two channels.
Spectrum meter
A classic. It shows us, also in real time, the presence of frequencies across the entire audible range by means of up to 60 bands. Viewing the spectrum of music similar to what we need to master will give us a very close idea of what we should be aiming for at each moment in the different bands and the relationship between them.
FFT spectrum analyzer
It should not be confused with the previous one. This meter is especially useful if we want to copy a song's spectrum onto another song by means of equalization. To do so, we would simply work on the song being retouched and get closer to the result of the original track, shown by a continuous line along the top representing its maximum levels.
Bit metering
The job of this meter is to tell us the bit depth, or audio resolution, we are working at at any given moment, or how a plug-in is responding. At first we might think this is a fixed value tied to the initial resolution of our file, but it isn't. As soon as any real-time process is applied, the audio data is handled at a higher resolution (32-bit floating point) to achieve maximum quality. One example of its use would be the possible need to apply dithering. If we intend to play back or save the file at 16 bits and the Bit Meter tells us higher values are being used, that process would be necessary. It can also happen that, although a file is at 24 bits, only 16 bits are actually being used (or a 32-bit file where only 24 bits are used).
ESSENTIAL EQUIPMENT
Ears: the main tool in the studio is our ears. Always watch your listening level, rest often, listen selectively, switch monitors (speakers 1 to speakers 2), and take into account the style and era of the recording. Use a sound pressure level meter so you always know your listening level, which should not exceed 85 dB, since that is the most suitable level for perceiving lows, mids and highs in balance.
