Vocals are the most important part of your mix. The vocals sell the music. Whether they’re a seasoned pro or your tone deaf aunt, everyone listens to the vocals.
When you focus your mixing energy on the vocals, you are focusing on the one part of the mix that every listener is guaranteed to hear.
Of course, good vocal production starts well before the mixing phase. It even starts before the recording phase, with good preparation and pre-production.
It would be fair to say that at least 80% of good vocal production comes down to the quality of the performance and the recording. But it’s your job as a mixer is to take that already great-sounding vocal and mold it to perfection.
Unfortunately, everything can all-too-easily go wrong in that final 20%. And I see new mixers make the same mistakes over and over again.
Today, we’ll look at 13 of the biggest vocal mixing mistakes that I regularly see my students make (and how to avoid them) so that you can produce vocals that sound professional, expensive and impressive.
Mistake 1: Not Sufficiently Controlling Dynamics
By nature, vocals have a large dynamic range. A vocalist can go from a whisper to a shout in a matter of seconds within a single song.
Because of this, vocals need a lot of dynamic processing. This is often done in several stages, and with a variety of tools. It’s not uncommon to see three or more stages of compression, volume automation by the word or syllable, and even a touch of brickwall limiting.
If you fail to sufficiently control the dynamics of the vocals, they will never quite sit right in the mix. If you find that the vocals suddenly go from sounding too low in the mix to too loud, then you need to apply more dynamic processing. Vocals with wildly uncontrolled dynamics are a hallmark sign of an amateur mix engineer.
Lighter genres that make a point of maintaining dynamics, like jazz and acoustic styles, are an exception to this rule. But even there, a gentle touch of dynamic range control is often warranted.
Mistake 2: Using Too-Fast Attack Times
By using too fast of an attack time when compressing vocals, you are cutting off the transients and losing the attack and articulation of the words. This has the effect of placing the vocal further back in the mix.
With lead vocals, this is bad. Generally, you want the lead vocal to be right in the listener’s face.
On the other hand, using a fast attack time on backing vocals is a great way to set them further back in the mix and stop them from stealing attention from the lead vocals.
I rarely use an attack time below 5ms for most of my gain reduction when working with lead vocals, and some would argue that even that is too fast for many performances.
The exception to this rule is if you are working with a particularly aggressive vocalist, and you want to make them sound smoother or if you have a particularly dynamic vocalist.
In the latter case, fast attack times can help tame loud transients, and shouldn’t cause too much trouble as long as you aren’t overly aggressive aim for around 2dB of gain reduction or so at the most.
In this context, I recommend using a fast compressor to tame the transients followed by a slower compressor to add more control.
Mistake 3: Using Too Much Reverb for the Genre
The current trend in modern mainstream music has been for vocals that are fairly dry, and it has been this way since the 90’s.
Adding reverb to a sound source makes it sound further away from the listener. But in many styles of music, we want the vocals to be up at the forefront of the mix, right in the listener’s face.
Use delays to create a space around the vocals without putting them further back in the mix. Try a stereo slapback with a different delay time on the left and right, between 30 and 200ms delay on either side. Or, try adding a mono delay with a bit more feedback (around 15-20%) timed to the beat of the music.
For modern mainstream styles, you might use an extremely short stereo reverb on the vocals, as long as there is no noticeable reverb tail. This helps to add width and shimmer to the vocal without drawing attention to itself or having it take up too much space. Make sure it’s barely audible—you should only miss it when it’s gone.
You can also try adding a gate sidechained to the vocal to cut off the reverb as soon as the vocals stop. Lastly, if you want a taste of reverb but don’t want it to sound too apparent, try adding a subtle reverb to the delay aux to make it sound a bit smoother.
Mistake 4: Applying Overly Aggressive EQ
You hear people talking every day. You know what a voice should sound like. As soon as you start applying drastic EQ curves to a vocal, it starts to sound odd and unnatural.
Unless you’re going for a special effect, keep your EQ moves subtle. A good guideline is to avoid boosting or cutting by more than 5dB. Instead, use microphone choice and placement to alter the tone of the vocal.
Exceptions to this rule are high pass filters, which can be set to make steep cuts to remove low end noise, and high shelf boosts, it’s common for many mix engineers to boost the top end quite aggressively.
Some engineers will make deeper, narrow cuts to remove room resonances, but if you’re EQ settings are starting to remind you of swiss cheese, it may be a sign that you’re pushing things further than they need to go if it’s a natural sounding vocal you’re after.
Mistake 5: Making One Compressor Do All the Work
Compressing in series is an easy way to control dynamics without making your vocals sound overcompressed. In the words of Dave Pensado, “Don’t try and make one thing do too much of the work”.
For most genres, I try to add 2-3dB of gain reduction each at three different points in the signal chain: During tracking (with an outboard compressor), before my main, tone-shaping EQ and after the EQ. (I’ll also use parallel compression on most mixes)
Everyone has their own method, but the outcome is agreed upon: Applying compression in several small chunks often sounds much more musical.
Mistake 6: Reducing Sibilance With EQ
Sibilance is a common issue when mixing vocals. Ideally, it would be addressed and reduced in the recording phase when possible. But oftentimes, we’ll need to help control it in the mix.
Attempting to reduce sibilance with EQ however, is a mistake, and one that many novices make.
Any tonal issues that aren’t static in nature should be addressed with multiband compression, not with static EQ. By cutting 5-7kHz to reduce sibilance, you are affecting the tone of the whole vocal. Why do that when you can just gently tap down on that area only when it becomes a problem?