Pro vocal production tips and techniques for mixing and mastering vocals. Compression, auto tune and vocal processing.
Struggling to make your vocals sound great?
Follow my tips and techniques for producing amazing vocals, every time
Vocals are often the centrepiece of a track and so it is important to give them the attention they deserve. Afterall, you’ve invested time into writing meaningful lyrics and crafting the perfect hook melody, so it would be a shame to undermine that with mediocre production. So, whether you are just starting out with recording and mixing vocals or you are an experienced producer looking to brush up on your production skills I hope you will find something useful here that will help take your tracks to the next level.
I would like to start by sharing some of the basic techniques I use when recording vocals to make later processes much easier. Most of these are simply good practice, which you may already be applying, but they are all fundamental in capturing the best vocal recordings possible with the equipment you are using.
Firstly, lets briefly cover pop shields. You will, more than likely, already be using a pop shield to control the plosives (that popping sound when a rush of air hits the diaphragm of your microphone) on your voice recordings. However, a surprising number of people position them incorrectly, usually far too close to the microphone. This renders the pop shield ineffective as there is not enough distance to disperse the air flow before it reaches the mic. Correctly positioning a pop shield largely depends on the style of vocal you are capturing; for softer recordings place it no less than 2” from the mic, and for louder performances somewhere around 6”. Your singer can then position themselves 2-3” from the pop shield, giving you a controlled vocal with minimal room ambience.
As a general rule, I always try to capture all of the lead vocals in one recording session, without the vocalist moving out of position. That way, when it comes to comping together takes, they match more accurately. Afterall, even a slight change of position in front of the microphone can have a significant impact on the tone because of the proximity effect; the closer to the microphone the singer performs, the more low-end is captured. This is fixable with EQ but getting things right at the source is always preferential over needless processing. It’s also worth noting that a vocalist’s delivery and tone may change over long periods of time too and this is something which is much harder to fix in post-production.
When I am recording with a vocalist, I always capture 3 to 4 takes of each part too. This way I have a choice of versions I can edit together to create the perfect performance. I do record more versions if I am planning to stack vocals on, for example, the chorus. But it is important to not overdo this and end up with unnecessary amounts of takes and a vocalist whose voice is getting tired.
Click image to enlarge
Avoid over-compressed dynamics, pre-treat your vocal with manual volume automation.
The best things in life are made by hand
One of the most common mistakes people make when processing a voice recording is to over compress it in an attempt to control unruly dynamics. This results in a squashed and lifeless sounding vocal. Instead, try using automation to manually level the volume of your vocal takes. You can then apply softer compression later in the mixing process to balance out any peaks, whilst maintaining the character of a vocal performance. This process is also really useful for balancing all the sibilant sounds in your recordings so that you don’t have to push your de-esser too hard in later processing.
In the following example you can hear an untreated vocal recording, followed by a levelled version. Notice how each syllable sounds balanced without sounding overly squashed, and remember this is before any compression has been applied.
This process can take a while if you have a lot of vocal parts to edit but don’t get complacent, it will be worth it in the end. There are plugins available which can automatically level in this way, for example Waves Vocal Rider, but I have always found I get better results doing this by hand.
There's nothing 'Auto' about it!
Once I have comped and levelled my vocals, I then move on to pitch correction. People often mistakenly refer to pitch correction as Autotune, the sound first heard on Cher’s ‘Do You Believe in Life After Love’ and later championed by T-Pain. Autotune, in fact, is the plugin that is most commonly used to achieve this vocal effect. The key to pitch correction is actually to be as transparent as possible, unless you are using it as a purposeful effect, so that the performance sounds natural and unaltered. If you’ve done it correctly then people shouldn’t notice it’s been done at all – and there is nothing ‘Auto’ about this kind of tuning!
I always pitch correct a vocal, regardless of how “pitch perfect” this singer is, to tidy up any fractionally sharp or flat notes. Most DAW’s have some sort of stock pitch correction but for the best results it’s best to use dedicated plugins such as Melodyne. The precision and depth of this plugin is unrivalled and, once you have learned it’s intricacies, you can achieve stunning results.
Melodyne is a whole tutorial in itself, however, so watch this space for that coming in the not too distant future
Visually time up your vocal stacks as well as using your ears to make judgements. Seeing the waveform will make a big difference to your editing decisions.
Increase energy and lift
Stacking vocals is a great way to thicken a part. This can be as simple as re-singing the same phrase again and mixing it with your lead vocal, a technique championed in the 60’s by The Beatles and The Beach Boys. For me, I like the contrast of switching from a single vocal part on a verse to a stacked one for the chorus. This helps increase the energy and lift the song. My most common approach for a chorus is to have the lead, backed by two further takes, one panned out to the left and one to the right. This, not only thickens the part, but also adds some stereo width. I also use this technique on any vocal harmonies too as demonstrated in the following clip.
Stacking vocals can get messy very quickly however, especially as you add more layers, so here are a few points to bare in mind. Firstly, it’s the subtle differences in pitch and time create that thicker sound, so don’t make the mistake of heavily pitch correcting your doubles. If they too accurately replicate your lead they will start to phase and this can take energy away. You only need to loosely tune the layers for it to be effective.
Secondly, don’t forget to conscientiously time up any additional layers. If you have a lead and its associated left and right doubles, plus a harmony with left and right doubles, that is 6 vocal parts in total. And if they don’t quite synchronise it can sound incredibly chaotic, especially on any sibilant words. As with the pitch correction we don’t need them to be mathematically precise, but getting the start and end of each word to reasonably match will go a long way to creating a solid vocal stack. I tend to use Melodyne to adjust timings but most DAW’s have the capability to flex and warp audio. Some even have dedicated functions for this, for example ‘Audio Alignment’ in Cubase.
A final note on vocal stacking – try cutting the breaths out of all but your lead vocal. This way your vocal part will still sound natural without sounding like
Together we are stronger
As you can see from the image at the top of this page, I like to use groups (sometimes referred to as buses) a lot. I do this for multiple reasons; not only because I like to keep a tidy project so that I know where everything is, but also for processing options, which I will come to shortly. I tend to group my vocal parts into sections - verses, chorus lead, chorus doubles, bridge and harmonies, which I then send to a master vocal bus. Within these groups I have my audio tracks, with the breaths separated on to a different channel to give me control of their volume throughout the mixing process, and the associated reverb and parallel compression sends. Having individual sends for individual parts means I can process them differently without having to mess about with lots of automation, although this is an option if your computer processor cannot cope.
To contextualise this, I will use an example from a track I completed for a recent project; I wanted to put the parallel compression louder on just the harmony and after doing so needed to cut some of high frequency content from this parallel compression. Fine, just put an EQ straight after the compressor and its done, without affecting the parallel compression on my verses and chorus – the benefits of grouping in practice.
Waves H-Comp is my go-to plugin for parallel compression - Click to enlarge image
Soundtoys Microshift is great for adding width to a chorus or harmony - Click to enlarge image
Waves SSL Comp is the final part of the vocal chain
A chain is only as strong as its weakest link
Vocal processing is dependent on a number of different factors; the recording quality, the vocal performance, the style of song, the list goes on. So, what works for one track is unlikely to work for another. I do tend to have a standard vocal chain, however, which is my usual starting point when it comes to processing.
Saturation - This is always first in the chain for me. Adding some harmonic saturation is great for rounding off a thin sounding vocal, filling out frequency content and warming up your sound. Soundtoys Decapitator and Scheps Omni-Channel are my go-to plugins for this.
EQ - This is usually second in my chain. I high pass between 80hz and 160hz, depending on the voice, to clean up the low end and leave room for bass instruments. Be careful to not take too much low off however as this is will leave your vocal sounding thin. I also do any surgical EQ at this point, reducing resonant or harsh frequencies. Fabfilter’s Pro Q2 is ideal for this.
Compression - I like to use a compressor with a bit of character here; and there are tonnes of amazing emulations of real analogue hardware out there. Because I manually levelled my vocal earlier, I do not need to compress too heavily, so a ratio of between 2:1 and 4:1 with a medium attack and slower release should suffice.
De-Essing - I tend to do this after I have compressed, in case the compressor has pulled up the volume of any sibilant sounds. Fabfilter’s DS is great for this but I also use the one within Scheps Omni-Channel. The latter is particularly useful, not just for taming sibilance but also, as a dynamic EQ for any disruptive frequencies.
Parallel Compression - Blending in a heavily compressed version of your vocal is a great way to add more character and punch to your parts. By doing this on a send effect channel you also have the option to process that compression further, using an EQ to soften higher frequencies or saturating the signal. Waves H-Comp is a regular in my toolbox for this job.
Reverb - Getting your reverb right can be really tricky and is entirely dependent on the style of track. Often people will mix it too loud, making the vocal too distant, or too quiet, leaving the vocal sounding flat. Take your time adjusting the reverb parameters until it compliments your mix and pay attention to the decay time, you don’t want the tail of your reverbs spilling into the next phrase too heavily. Don’t forget to utilise the pre-delay, if necessary, to stop your vocals becoming too distant. And remember to EQ your reverb too; you will be amazed how high or low passing your reverb sends impact the clarity of your mix. My favourite reverbs for vocals are Altiverb and Valhalla Vintage Verb.
Stereo Width - I will often add Soundtoys Microshift as a send effect to add more stereo width to a chorus or harmony. Occasionally I will low or high pass this if the frequencies are interfering with the original part.
Spot FX - sends also tend to be where I add any spot effects, such as delays or reverb tails, that I wish to turn on and off.
Master Vocal Bus
EQ - By the time the signal reaches the master vocal bus, it should sound pretty much finished. Any EQ adjustments here are simply to help the vocal sit better in the mix and I generally do this with an analogue emulating EQ. Don’t forget to do these tweaks with the music playing too in order to carve out those frequency ranges.
Bus Compression - This is the last thing in my chain and is used to glue the various sections of vocal together. You don’t need to slam it too hard, just tame any rogue peaks. I love the Waves SSL Comp for this job.
And finally, there are infinite variations around this chain so this is by no means the only way to mix a great vocal. Following these tips, however, should help you get in the ballpark. Happy mixing!