What does the mastering engineer do?
In a previous blog I explained what mastering was and why it is necessary. I would now like to look at the actual work that the mastering engineer does. The task can be divided in two parts; premastering and mastering. Premastering, the more artistic task of the two, is today’s topic. Mastering is more technical and will be covered next time.
Premastering
The first task, premastering, optimizes the audio and is an artistic task. It is during premastering that the engineers can express their talent and really add value to the final sound. The audio to be mastered needs to be transferred to the engineer’s digital audio workstation (DAW). Popular DAWs for mastering include Pyramix, SADIE, Sequoia and Sonic Solutions. The mastering engineer then starts by listening to the mixes in the context of an album and concentrates on various aspects of the sound, listening for spectral (tone, EQ…), dynamic (compression…) and stereo width discrepancies or imbalances from one track to the next. Making the final product somewhat homogenous is a major task. For example, the bass content from song to song must be similar so that the bass does not sound louder in one song compared to another. If the project is a compilation of songs from different sources or different artists, creating a homogenous product can be a challenge: The songs may have been recorded in different eras, in different studios and by different engineers. The mastering engineer provides the neutral ear and the trusted accurate listening environment that the task requires.
Although mastering engineers generally have degrees in audio engineering, acoustics, or both, the engineer’s experience is also key in knowing what the standards for the different music genres are and what is expected of the final product in the industry. The engineer’s acute listening skills allow him to notice inconsistencies and incoherent elements in the sound that had until then remained unnoticed. The experienced engineer will know how to intervene to correct them.
However many tools (hardware or software) may be available to help him, the engineer’s ear remains the best guide. These tools include frequency spectrum analyzers, phase monitors, level indicators (PPM or VU for example) and stereo width monitors. In a worse case scenario, when no other solution is available, the mastering engineer may require that the audio be remixed before he can work on it.
Signal processing tools are also used: EQ (from Weiss notably), compression, maybe delay and reverb. Multiband compressors are typical in the mastering studio as they allow different frequency bands of the same signal to be compressed differently. Noise reducing is often necessary to get rid of background noises, hiss or the other odd buzz. Cedar and Sonic Solutions No Noise are examples of noise reduction software.
Analog signal processors (EQ, compressors, etc) tend to be preferred although digital processors are also used. If analog gear is used, the signal or tracks may need to be converted to the analog domain first. With digital signal processors different tricks are available once the incoming signal has been converted to a M/S (Mid/Side) signal. Using an M/S signal allows manipulation of only the stereo part of the sound or, the opposite, modifying just the mono elements of the sound. It is difficult to say whether analog is better than digital or the opposite because digital signal processing is highly dependent on the algorithms used the same way analog signal processing is highly dependent on the quality and type of components used.
The engineer will listen at various levels, between 70 and 85dBSPL, sometimes louder depending on the type of intended playback environment. The ear’s frequency response is not linear and varies with the sound pressure level. In order to hear the softer parts correctly, the mastering studio must be a sufficiently quiet listening environment. Listening in headphones or in a cheap sound system are also good ideas to test how the mix translates.
Over time, the loudness war has caused producers, artists, radio and record label personnel to push for louder records from mastering engineers. Without entering the loudness war debate, suffice it to say that a louder record can only be obtained by compromising its dynamic range. The mastering engineer becomes the referee between the conflicting demands of a loud record and one that breathes with sufficient dynamic range. Some engineers believe their role is to improve how the program will translate from one system to the next, other engineers believe their role is to make as big a sonic impact as possible.
Once all spectral, dynamic and possibly ambient/reverb modifications are made, the transitions between songs need to be addressed: How much silence between tracks? When to start fading out a track? How much overlap between crossfading tracks is necessary? These are all questions that need to be answered depending on the project. Input or precise instructions from the client are more than necessary for this task.
Mastering for vinyl requires that the mastering engineer check that there are no phase issues between the left and right channels in the bass frequencies, otherwise pressing a vinyl becomes impossible. In addition, the high frequencies dynamics (change of level from low to high) must not require the stylus to accelerate faster than it is physically capable. It is also important to know that the louder the record, the shorter the possible running time per side.
Questions or comments ? k.blondy@musitechnic.net
