As you may already know, audio signals can be processed using a plethora of tools called effect units. Producers and engineers often insert them directly into the signal chain so that the unit captures and affects 100% of the sound. Just as often, they combine the unaffected signal (dry) and the affected signal (wet). This technique is called parallel processing. It preserves the presence, character, and location of the source while still coloring it and allowing you to control the balance.
If you’re familiar with audio compression, you’ve probably heard of parallel compression. In that case, you know that compressors bring loud levels down and low levels up – thus reducing the dynamic range. However, since our ears are way more sensitive to the high levels being limited than the low levels being boosted, a new technique was born to highlight upward compression. Therefore, it is perhaps the most utilized implementation of parallel processing. In addition, it lets you make sounds bigger while retaining their dynamics and makes the artifacts of compression less evident (since the compression is being done on low levels).
This means you can drive the compressor harder than you could in serial without damaging the sound. It also allows you to adjust the amount of compression in the mix with faders instead of parameters. What is more, there are a number of ways to achieve parallel compression in both analog and digital setups. An analog setup requires some re-routing of signal and of course, auxiliary tracks (since you can’t simply duplicate/create tracks like in a DAW). In this section of the blog I will demonstrate a quick way to set this up digitally in Lumit.
First, drag an instance of Lumit’s Compressor/Expander onto the track you wish to process. This unit will address the downward compression aspect by gently reducing peaks above the threshold. Set it’s parameters to something like this.
Create a return track and call it something like comp. Insert another instance of Lumit’s Compressor/Expander onto it. This will address upward compression.
Set its parameters to something like this…
Bring up the send fader for comp on the instrument track.
Experiment with the balance between the track’s output, it’s send output, and the return track’s output until you hear something you like. Listen to the drastic difference between serial compression and parallel compression. Notice how parallel compression preserves the details and presence of the kick drum better than serial.
Like in parallel compression, parallel distortion gives you control over the amount of distortion in the ultimate output of a source. Since you can distort at low levels in this configuration, it allows you to drive the effect a bit harder than you could in serial distortion without it sounding bad. This way, the source comes through clean but not TOO clean. Here’s how you set up parallel distortion in Lumit using a bit crusher.
Create a return track and drop an instance of Lumit’s bit crusher onto it. Name it something like crush.
Set its parameters to something like this….
Remember, dont be afraid to really drive parallel effect units. A unit this damaging may destroy a nice recording in serial but can be balanced when used in parallel.
Bring up the send fader for the track you wish to parallel distort.
Listen to the drastic difference between serial distortion and parallel distortion on this synth sound.
I would argue that parallel reverb is critical to a professional sounding mix. This is why just about every software reverb (and some hardware units) feature a dry/wet parameter. It expedites the process by performing the duplication and merging for you. The drawback of dry/wet knobs is that you can’t further process the wet signal independently like you can using return tracks. Parallel reverb simulates standing in front of a performance in a room because you get both the direct signal and the reflected signal. Furthermore, serial reverb (100% wet) can sometimes make the source sound far away in space, with an ambiguous location. Check out this processing method in Lumit that involves both serial and parallel reverb. I used the technique to separate early and late reflections such that the early reflections are in serial and the late reflections are in parallel. This provides a realistic sense of space.
Additionally, drag and instance of Lumit’s reverb unit onto the instrument you wish to affect (in serial) and set its parameters to something like below. Make sure than the wet balance is at 100% and the ER/LR balance is set fully to ER.
Furthermore, create a return track like in the previous examples and drag another instance of Lumit’s reverb unit onto it. Name it something like reverb.
Then, bring up the send fader on the instrument you’re affecting.
I would advise you to experiment with the balance between the instrument’s level, it’s send level, and the return track’s level until you’re happy with it.
In the end, listen to the drastic difference in spacial orientation between the serially processed sample and the sample processed using this technique. Moreover, notice the reverb tail and sense of space is more realistic in the second example.