Compression in Mastering

• Compression is not about removing dynamics but managing them.

• It is about finding where the song is loud enough, tense enough but also have enough dynamic contrast to bring emotions.

• Get a balanced EQ first then apply dynamic processing.

• Can use multiband compressors
  • Bass and mid band won’t affect the high

• Adding compression will change how the EQ sound

• Cycle: Adjust level of the song – adjust EQ – adjust dynamic – adjust EQ

Examples of Mastering Compressors

Plugins/Hardware versions:

SSL G-COMP

SHADOW HILLS

API-2500

Limiter in Mastering

• Limiters are excellent at bring up the levels in the mastering process without allowing the peaks to clip.
  • Especially the modern plugins are extremely precise in catching peaks and won’t allow signals to pass through the ceiling

• Excellent at controlling the transients, short sharp spiky signals in the beginning of the notes like the initial attack on a drum or finger clicking noise on guitar.

Examples of Mastering Limiters

L3 MULTIMAXIMIZER

L2 ULTRAMAXIMIZER

Breif History of Compression

Before compression effects were introduced, engineers would have to control the dynamics by ‘riding the faders’, that is manually reducing the volume during loud sections and pushing it up during loud sections. This was laborious, error prone and only really worked over longer periods of time, such as a whole verse / chorus rather than individual words on a vocal for example.

Early compressors were valve-based and tended to create a ‘warm’, if not transparent sound. Not many would be available to the engineer. It should be noted that tape-saturation could be used by the engineer to create a reduction in dynamic range. As multi-track technology developed through the 1970s and 1980s, more compressors were required to control the dynamics of more sources. It would not be uncommon to have perhaps 10 compressors on a mix over vocals, bass, drums and busses. Solid-state (transistor technology) allowed for the compressors to become smaller, cheaper and more reliable which made them more suitable for this setting. This led to very clear, precise and punchy mixes.

Early compressors simply had an input and an output but later ones also included a side-chain connection to the sensing circuit. This was useful for broadcasting to duck sound / music when the commentator or DJ speaks. However, creative uses were found by producers and the sound of a synth pulsing up and down, side- chained to a kick-drum is a classic feature of EDM. Unusually, compression didn’t really become popular as a digital hardware effect (unlike time-based effects).

However, in the 1990s, especially with the launch of Cubase VST, compression could be used directly in the DAW as a software plugin. Multiple instances and be loaded in to the software without having to buy more physical units. The settings of the compressor are also saved within the DAW project file so recall of settings is not required. As the computing power increased, software compression could start to emulate / model the sound of analogue compressors including valve characteristics and modelling of some classic units such as the UREI 1176. Some engineers realised that more compression could be applied without creating audio artefacts by splitting the frequency spectrum.

Early attempts at multi-band dynamics would involve using 2 compressors and a cross-over to split the audio in to bass and treble. Later, manufacturing companies designed single box solutions including increasing number of frequency bands. As the technology was ported over to software even more bands could be incorporated. This can smooth out and peaks in frequency but can also lose the defining character of a sound so must be used with caution. Parallel-compression is a means of blending a compressed signal with an uncompressed signal to maintain a sense of the original dynamics and transients, but increasing the overall level / excitement. This was a contrived process of sending a channel to a bus, compressing the bus then turning up the bus return to the desired level.

However, many new software compressors have a parallel or wet/dry control to achieve this with no routing necessary. Compression was used by mastering engineers as a means to get a good level on to a vinyl disk and not overmodulate the groove to create errors or unplayable disks. As the mastering became a software process, huge amounts of gain reduction could be applied in order to achieve a loud master.

A process of one-upmanship led to the loudness wars. This was where each new record seemingly needed to have a higher average level to sound louder than the last one. This was to the detriment of the listening experience and the listener would simply turn the volume down and get audio that lacked dynamics and punch.