Processing something “in parallel” is a common technique today in the mixing world that goes back over 30 years. The technique first started popping up in the late 70′s and became more popular throughout the 80′s, especially in New York-based studios when adding compression to a mix (hence the technique is often referred to simply as, “New York Compression“). It is a very cool alternative to simply tossing a compressor on an insert because it gives an additional way to control how much of the effect to include in the mix.
So, what is it?
Traditionally when processing a specific channel (or bus), the effect is placed in-line with the channel via an insert. This routes the entire channel’s signal path through the piece of gear, and is often used with an outboard EQ, compressor, gate, etc.The result is that the effect is heard on the entire signal: stick an EQ across a vocal channel and that EQ will affect the sound of the whole vocal.
By comparison, parallel processing means only some of the signal ends up being processed, and this then gets mixed back in with the unprocessed.
Imagine it this way: you are mixing a song where the lead vocal just isn’t cutting through the rest of the mix, and you decide it needs to be a little brighter. You then bring in the lead singer’s identical twin and have them record the exact same vocal take on a new track, giving you two identical tracks. You crank the EQ, but only on the second vocal (3 kHz 10 dB for a start). Lastly bring down that EQ’d track’s fader and just add in enough until the vocal has a little more edge to it and cuts through. That is parallel processing: split the signal in two and process only one of them, mixing them back together at the end.
To process in parallel in the studio, the particular channel (or bus) needs to be split in some way, sometimes by “multing” at the patchbay (creating two channels each fed by the same source), or by using a send on the console. One of these signals is left unprocessed, the other is routed through the outboard processor, and finally both signals are mixed back together to taste. The result is still a processed signal, but because the original unprocessed signal is present along with the processed version, heavy processing can be performed while still retaining the original’s character.
If you’re new to the idea of parallel compression and all this routing sounds tricky, but you’re interested in giving it a try, then I have some good news for you: Auria has a very easy way to apply parallel-style compression. It is done with the Subgroups, specifically on the PSP BussPressor found on every group. Cue the tutorial!
Setting up the BussPressor
A great use of parallel compression is to process multiple channels at once. In this example I’ve created a drum bus by bussing the kick and snare tracks over to an empty subgroup; to route a channel to a subgroup simply tap the “Subgroup” box on that channel (it’s right above Aux 1) and select one of the eight available groups (one cool tip is that if you name your Subgroup channels those names will appear in your drop-down list, making it easier to manage your subgroups).
Here is that Kick and Snare subgroup soloed with no bus processing turned on:
Note that in the above sample there is already some individual track compression happening on both parts, but we haven’t added anything to the bus.
Now I’ll go about adding some compression. I enable the BussPressor (tap the “Bypass” button so it’s NOT yellow, and tap the “CMP” so it is yellow), and dial in some pretty heavy compression: a high ratio of 10:1, a quick attack time of around 5 ms, and I adjusted the release time so it mainly compresses only the front-edge of the kick and snare punch while leaving the drum’s sustain in place. I also used the Makeup Gain knob to raise the compressed signal so it was closer in volume to the original – tap the CMP button to cycle the BussPressor on and off to make level matching by ear easier.
Here is how I have the BussPressor setup so far (click to enlarge):
Take a look at the meter in the above screenshot, you should see that there’s about 12 dB of gain reduction happening. So, yeah, fairly heavy, and much more than you’d usually want on a drum bus. Here’s what this amount of compression sounds like:
As you can hear the attacks are really squashed, but because of the release time the “back end” of each drum is higher in volume relative to the front. The drums are fatter, but at the expense of any sort of edge (like a snare’s crack) that would cut through a mix.
Blending the Two
Now here is the meat of the tutorial. There’s a knob on the BussPressor labeled, “Mix”. The Mix knob adjusts how much of the uncompressed signal should be passed through vs the compressed version. At 0% you would hear only the original, unprocessed signal (Sound clip #1 above), and at 100% you would hear only the compressed signal (Sound clip #2 above).
The last step is to start adjusting that Mix knob. What I want to find is a setting which is still mostly unprocessed signal, enough so that the kick and snare really have that bite, but with some of that overly-compressed signal blended in to fatten up the sustain. I ended up setting the Mix to around 30% (i.e. 70% uncompressed).
And here is the final result:
The final version retains that initial up-front attack, but by blending in some of the compressed signal in parallel there’s a little bit of thickening that happens, too. This would be tricky to do by compressing in-line, and though by no means impossible I think this method is much quicker.