When mixing vocal harmonies I like to start with a gate to attenuate background noise and unwanted breaths. Then I’ll attenuate the double and harmony’s fundamental with an EQ, introduce de-essing to balance the frequency response and fix timing issues, and then tune as needed.
The chapters in this video are in order and can be used as a chain, but feel free to change things up when using these tips in your session.
For this session, I’ll have my fully processed lead vocal and then introduce processing to a double, high harmony, and its double, as well as a low harmony and its double. Also, all vocals were recorded using a $20 condenser microphone with no outboard processing so you should get similar if not better results using this chain.
I liked to start affecting my BGVs or harmonies with a gate - I won’t fully attenuate the signal when it falls below the threshold, but use a range to only attenuate it by 6-10dB. I’ll use lookahead, and a soft knee to smoothly attenuate and tighten the vocals.
Let’s take a listen to how this creates a more cohesive sound.
For my double, and optionally for my harmonies, I’ll attenuate the low frequencies using a high-pass filter with a 6dB per octave slope using this Shade EQ - this gives the lead vocal more room and keeps the BGVs or harmonies from sounding too prominent. Optionally, I’ll attenuate sibilance frequencies to clean things up.
This also keeps the BGVs from clashing with the lead. Let’s take a listen.
Next, we can edit the esses of the BGVs and harmonies with clip gain or use heavy de-essing to attenuate the sibilance significantly. Since sibilance is similar to transients, this will help keep the BGVs sounding more in time by reducing noticeable out-of-time aspects.
Additionally, it keeps the overall frequency response of the vocals balanced, since excessive sibilance is almost always unpleasant. Let’s take a listen.
You may not need to tune your BGVs and harmonies, but most engineers will do this, especially when working on a contemporary mix. Since harmonies will blend in with the lead, using heavy or aggressive tuning isn’t too problematic and will help the lead sound more in tune.
Let’s listen to our BGVs with aggressive, somewhat stylistic tuning added.
Similar to tuning and de-essing, compression can be introduced aggressively to control dynamics and create a cohesive sound. This again helps blend the BGVs with the lead, as well as reduce noticeable timing differences between any of the performances or between the BGVs and the lead vocal.
I found the PSP impresser worked well for this since it offered lookahead and automatic make-up gain.
Let’s take a listen.
With my BGVs and Harmonies balanced dynamically and spectrally, as well as tuned, I’m going to start affecting their imaging. I like using a sample delay for this, and delaying one of the channels by a few milliseconds to reorient the stereo placement, using different timings for each track.
Let’s listen to how this creates room for the lead in the middle and makes the background vocals more impressive and expansive.
If you want your BGVs and harmonies to stick out, use unique temporal processing, but if you want them to blend in with the lead use the same temporal effects that you used for the lead. That said, I'll send them to the same delay and reverb.
In this example, I sent the BGVs to these processors without the lead being included, meaning the effects will process slightly differently than how they would if the lead was included.
This is a small distinction to make but thought I’d mention it nonetheless. Let’s take a listen to how this blends the BGVs in with the lead.
Up next, I’m going to adjust the levels of harmonies to prioritize some over others - since their individual amplitudes won’t be affected by more processing, I can set the levels knowing they won’t change in relation to one another. Then I’ll pan to affect their imaging.
2 chapters back, sample delay affected their image within the 180-degree field, but panning will adjust that already affected signal within the 90-degree stereo field.
Let’s take a listen to these changes.
Next, I want to send my BGVs and harmonies to a single bus where they can be collectively processed. Additionally, I’ll change the output of my temporal processing sends to this bus as well so that everything will be processed collectively, but this is just my personal preference.
On this collective bus,I’ll introduce an M/S EQ and boost the low mids and highs on the side image , while dipping the stereo image, or both mid and side, around 2kHz-5kHz.
This makes the BGVs more impressive due to the side image boosts, but keeps them from overpowering the lead with the roughly 2kHz-5kHz cut.
Let’s take a listen.
I want to control the BGVs and harmonies dynamics and timbre a little more, so I’ll use Arturia’s Tube Culture plugin to saturate everything collectively. In the advanced section, I cut out some of the lows from the input, as well as subtly compressed the incoming signal.
With the output EQ section, I used the tilt filter, combined with the Air function to subtly amplify the bus’s high frequencies. Then I reduced the output to compensate for gain changes.
Let’s take a listen and notice how this gives the vocal a uniform timbre and creates collective timing via the subtle compression.
Lastly, I wanted to ensure that my BGVs and harmonies didn’t have aggressive and unpleasant resonances, as well as help, distinguish them from the lead. Using Soothe 2, I side-chained the lead vocal and introduced the plugin’s dynamic attenuation to roughly the full frequency range.
Let’s listen to how this helps the lead vocal sit on top while simultaneously balancing the entire BGV and harmony frequency response.