Parallel Compression explained (Part 1)

So, in my last post I gave you guys a sure fire way on how to set up parallel compression on any track you like in Pro Tools.  The how to do it is basic enough, send one stereo bus to a new auxillary track and throw a compressor on it and begin tweaking knobs, but the why is a bit more complicated.  In order to fully understand exactly why we use parallel compression, or NY compression as it is commonly called, you need to know that there are two major types of compression used in audio, upward compression and downward compression.

Downward compression is the most ubiquitous kind of compression out there.  All audio and mastering engineers know this kind of compression inside and out.  The main characteristic of downward compression is the function of making the louder portions of a signal quieter.  This is usually done by setting a level known as the threshold, but on some machines and plug ins this value is determined by a mix of a set of input and output values, for now I'll focus on the threshold control since it's most common.  The threshold is a level that is set by the user that tells the compressor at what amplitude, or volume level, the compressor should begin to compress, or become engaged.  Any signal that goes above this threshold amount is compressed at a ratio usually set by the user.  This ratio determines how much of that input signals level is attenuated above the threshold amount.  A 1:1 ratio produces a linear output, thus what goes in is what comes out.  A 2:1 ratio will compress, or attenuate, the portion of the input signal that is ABOVE the threshold by a ratio of 2:1, but only the amount of dB ABOVE the threshold will be attenuated.  Therefore, if an input signal goes above the threshold by 2dB and the ratio is set at 2:1, the output of that signal above the threshold will be only 1dB, essentially halved.  A 4:1 ratio would output the signal at .5 dB and so on.

An input/output plot showing the action of a compressor with a threshold set at -20dB.  Each color above the threshold represents the output of a different ratio.

 This seems simple enough right, a downward compressor is shrinking the dynamic range of a signal, or instrument, by a set ratio by making the louder sounds quieter and the softer sounds louder relative to the overall output of the signal.  Now, you might think to yourself, "but I thought compressors made things louder."  Very simply they don't.  Any form of downward compression is not really making anything louder, it is only squashing the dynamic range.  But, since the louder parts of the signal are attenuated, you can use make-up gain on the output of the compressor to either restore, and even boost, the loudest parts of the signal.  Therefore, compressors don't themselves make anything louder, but by shrinking the dynamic range, they can give you the illusion of things sounding louder, sometimes even punchier.  This is sometimes, but definitely not always, a very useful effect.

Graph of the same signal from above, but with a 7.5dB make up gain set.

In my next post I'll explain the functions of upward compression and how these two kinds of dynamic processing relate to the usefulness of parallel compression.  Happy mixing.