Understanding how powerful a tool parallel processing is can help unlock the key to the mixes of your dreams.
Parallel processing is probably more common than you realise, and even if you don’t understand it completely, you might be doing it already! Mix knob on your compressor? Your reverb? Delay? That’s parallel processing!
Time for a quick English lesson; parallel refers to something moving or running side by side without ever crossing over or touching, like a train track. Parallel processing in audio is two tracks playing simultaneously, one with different processing to the other. While the advent of parallel compression is somewhat new, patchbays have had ‘mults’ for decades, allowing single sources to be patched, recorded and processed separately.
Read up on all the latest features and columns here.
Let’s assume these two tracks are set to an equal level, 0dBFS for example. If you compress one super hard, it’ll be crushed, squashed and super controlled, but you’ll probably lose a lot of dynamic. The uncompressed track, however, is still playing back at full volume, dynamics and all. The summed audio retains the dynamic of the uncompressed track, while the compression adds energy, life and punch to the source. Adjusting the faders and blend changes this relationship, offering you near limitless opportunities for fine tuning the final sound. Let’s continue this example with more parallel compression.
Parallel compression
Parallel compression is also referred to as ‘New York compression’ because of how quickly it caught on in New York studios. Originating in Dolby A noise reduction circuits (from the 60s!) these units featured two parallel buses with compression on one of them, and articles written by engineer Mike Beville described the use of parallel buses to achieve heavy compression without too many artefacts or overly squashing and degrading the source.
Parallel compression allows you to heavily control the dynamic of drums, keeping them stable and rock-solid, and blending it in with the original, uncompressed signal. This technique works great for elements of a mix that need to remain steadfast like rhythm elements like drums and bass, but equally well for a source like a lead vocal or guitar solo that again, needs to retain dynamic and nuance while requiring control or taming in an arrangement.
Fast compressors and slow compressors alike will work for parallel compression, depending on exactly what you’re after in your mix. Faster compressors will clamp down on transients faster, allowing the unprocessed track to shine, while a slower compressor will allow those transients to poke through, summing and combining with the raw track! Experiment with EQing your parallel track before or after a compressor, and you can seamlessly tuck the effect in. Speaking of EQ, you can also use EQ in parallel.
Parallel EQ
Parallel EQ has a variety of uses, and differs slightly to simply EQing a source. EQing adjust the frequency response of a source, while parallel EQ changes a few things, 1) what your mix bus is hearing and 2) offers a more distinct frequency change than the subtle bell shape that an EQ will impart. Both have their case uses, but today we’re discussing it in parallel processing!
Parallel EQ also allows you to process the parallel track differently. For example, rather than simply boosting the high end of a vocal, why don’t you high pass it to say, 9kHz and saturate it heavily? The result will be a more balanced, bitey and aggressive sound than simply boosting the main track above 10kHz.
At the opposite end of the spectrum, is there a bass solo during your song? You could run a super-EQ’d and subby bass track below your static track to help give the bass a push when it’s the time to shine. You can even have this track muted for all the song except for the solo, much like automation. This applies to mids for solo guitar, horns or synth as well, a heavily band-passed track can appear in parallel to give lead instruments a heavy-handed push forward at different parts of a song!
The first example of saturating a vocal can also serve as its own technique: parallel distortion.
Parallel distortion
Parallel distortion, saturation or overdrive can be a great way to do… well, almost anything. Heavily high-passed distortion can add smack and attack to tracks, while band-passing (high and low passing) can help to focus the saturation in whatever area you might want. Low-passing can help to add girth, weight and heft—though be careful of adding mud!
While tools like FabFilter Saturn, Ozone Exciter or Wavesfactory Spectre can do multi-band saturation with ease, running it in parallel and EQing these tracks yourself allows you to use additional effects to really refine your overall effect and sound.
We mentioned using saturation earlier to bring out the detail of a vocal, but this works equally well for adding attack to drums or bright string noise to bass guitar! You can also experiment by saturating an entire mix and dialling back the parallel track. Total obliteration can often be too much, but tucking in a little saturation can serve as glue and a little extra connection and cohesion between all the elements.
Auxiliary sends
Finally we’ll discuss auxiliary sends, something we often forget are uses of parallel processing. Time-based effects like delay, reverb and modulation are often used via sends (originally from a console etc.) and returned onto faders. This process hails from days before plugins, where a studio might have a single reverb chamber, plate or rack unit for an entire mix. With an auxiliary send on a console (usually used for headphone mixes), multiple tracks could be sent, during mixdown, allowing an entire drum kit, or just shells etc. could be pre-mixed to a single reverb.
This goes for delays and modulation as well, from the days of tape delays and chorus units, before the advent of simply clicking open an insert onto your track! This use of auxiliary sends does a few things. Firstly, it offers a more cohesive sound than multiple reverbs, delays and effects. Putting everything into the same ‘space’ or reverb sound can add a cohesive feel to the whole arrangement, albeit with you able to control how much of each signal is sent to the reverb.
Secondly, the return of the reverb is obviously open to more effects like EQ, compression, gating (if you’re a Clearmountain fan!) and more. Even when using reverbs as inserts, and even if they have a Mix control, shaping the reverb doesn’t really work quite as well as when it’s an auxiliary. The integrity of your original sound remains intact, and all you’re doing is augment it.
Our friends at iZotope have some more info about the pitfalls and perils of parallel compression. Keep reading here.
