Tech Tip:Parallel Effects in a DAW World

By Craig Anderton



Back when hardware ruled, creating parallel effects was pretty easy. You'd send an output into a Y-cable, split it to two different effects, and there you had it: instantly parallel processing, where one signal could take two different paths. One obvious use was creating stereo effects out of a mono source; for example, the parallel processing could consist of two chorus devices, or delays set to different delay times, each panned to opposite sides of the stereo field.


Digital Audio Workstation (DAW) software is a different story. In almost all cases, the program will assume you want to put the effects in series, one right after another. There are a couple of exceptions; Mackie's Tracktion lets you insert complex combinations of series and parallel effects with a track, and Ableton Live 6 lets you create instrument racks with parallel effects. But for most programs, you'll need to get a little creative.



Do You Copy?


One way to achieve parallel effects is to copy (clone) the track to which you want to apply the parallel effects, resulting in several parallel audio tracks. You then apply effects to these tracks as needed. For example, suppose you want to add a parallel effect to a piano track, where a noise gate lets through only the peaks; furthermore, this goes to a reverb that's panned far left. Meanwhile, a second noise gate sends a different set of peaks through a short delay, to a different reverb that's panned far right. You could do this with aux sends, but there's an alternative.



For this example, we need three parallel tracks:


  • Straight piano only
  • Straight piano + noise gate + reverb1 (panned left)
  • Straight piano + noise gate + delay + reverb2 (panned right)


Copy the straight piano track two times for a total of three piano tracks. The first track is the "straight," unprocessed track. In the second track, insert the noise gate and reverb, then pan the track toward the left. For the third track, insert the noise gate, delay, and second reverb, then pan that track toward the right. (Of course, you could also slide the third track behind a bit in time to create the delay, but sometimes it's a lot more convenient to just dial in a delay, particularly if you need to sync to tempo.)


Because tracks in today's DAWs are aligned with sample accuracy (and assuming the effects paths have delay compensation), you won't hear any flamming, comb filtering, or other undesirable effects when you combine the tracks.



"Virtual Mics" with Parallel EQ


Here's a real-world example of using parallel effects to create a wider stereo image.


In some ways pianos are fun to record, because they generate sound over a wide area. Stick a couple mics in the right places, and you'll end up with some great stereo imaging. But other instruments, such as classical guitar, accordion, percussion, etc. don't have a wide stereo image if you hear them from more than a few feet away—although up close, it can be a different story. If you're facing a guitarist, your right ear picks up on some of the finger squeaks and string noise from the guitarist's fretting hand. Your left ear picks up some of the body's "bass boom;" although not as directional as the high-frequency finger noise, it still shifts the lower spectra somewhat to the left. Meanwhile, the main guitar sound fills the room, providing the acoustic equivalent of a "center channel."


This all became very clear to me when recording a guitar/keyboard duo, where the keyboard had a nice spread but the guitar kept getting shoved to the center of the image. What to do? I tried using two mics on the guitar, but the phasing issues were unacceptable. Then I thought about what made the sound "wider" as you got closer, and a solution suggested itself. I've also used the following technique to stretch a piano and organ's image beyond what I could obtain simply by using two mics; in fact, this basic principle works for most sound sources where the bass doesn't need to be in the middle of the stereo image.


The first step in simulating the effect of being close to the guitar was to copy the original guitar track to two more tracks. The first clone provided the "squeak" component by including a highpass filter that cut off the low end starting around 3kHz. This was panned toward the right. The second clone for the "boom" channel used a lowpass filter with a sharp cutoff from 430Hz on up. This was panned a little closer to center than the squeak channel.


I also needed to remove some low end from the "straight" channel to make some sonic space for the boom channel. Adding these two tracks to the main track pulled out some of the "finger squeaks" and "boom" components that were in the original sound, and positioned them in a more realistic stereo location. This also stretched the stereo image somewhat. And because these signals were extracted from one mic, there were none of the phasing problems associated with multiple mics.


As to mixing these three elements, the drastic amounts of high and lowpass filtering on the cloned channels brought their overall levels way down, even without touching the channel fader. If you isolate these tracks, it seems as if their impact would be non-existent due to the low level and restricted frequency range. But if you mix them in with the main channel, the entire sound comes to life...which is definitely a Good Thing.