How to Record Good Vocals

When recording vocals I have noticed many times that people are surprised at how they sound when they hear themselves played back over the monitors. This is because what we hear in our heads is not what a microphone will hear (pick up). People often think their voice is a tad bassier than it really is due to the proximity effect of their own voice. Proximity effect occurs when a sound source is very close to a mic mic thereby increasing the sources bass greatly. Think radio voices.

There are really only 3 things to consider when recording vocals: The volume, the tone and gain structure.

Regarding volume we are dealing with two things:

1. Is the person very soft? If so you will have to turn the mic up to compensate for this. And likely EQ for intelligibility of the words. This often means using EQ and cutting frequencies lower than 80hz and maybe a dip in the middle range around 300hz and maybe adding a little peak EQ between 3 to 7kHz.

2. Is the person very soft and then very loud? This will often mean applying the soft voice needs above as well as using a limiter or compressor to keep the loud parts from being too loud. A compressor will restrict sounds from getting too loud by a dialing in a certain loudness ‘threshold’. The problem is that if you compress to much the person will suddenly sound like they are in a box if they get too loud and overcompress the vocals. Finding a happy medium here can be tricky and each engineer has a way they like to do it. In the end the most common approach seems to be the engineer ‘riding’ the level knob with the vocals. As the singer gets louder the engineer turns them down a touch. This means the engineer has to have good, actually great, timing and feel and be very familiar with the vocals. For a person to do this alone is not possible. In this case the best solution is to simply record two separate tracks; one loud, one soft. This whole explanation of course takes into account that the singer should have practiced ‘working’ the mic. That is the singer should know how to come off a mic as they get louder and move in a bit as they get closer.

Regarding tone we are looking for the over all tone of the singer.

1. Are they bright as sibilant or are they thick and bassy sounding. The idea here is that you would place a less bright mic on a bright singer. Bright singer plus bright mic will end up more than bright, it may end up shrill. Typically dynamic mics are less ‘bright’ sounding than condensers. Consider Michael Jackson: although he could afford any mic in the world (actually he could of bought ALL the mics in the world and just told everyone to “shut up”) he actually used a $400 Shure SM7 microphone. This is because his producer Bruce Swedien wanted to have control over the bright side of his voice during mixing and didn’t want it accentuated during the tracking stage. Michael also has a somewhat brighter and mildly raspy voice and a condenser would of only unnecessarily added more of the same.

2. Is the singer bassy or mid heavy? (This is my voice.) In the case of an unbright voice you would typically add a brighter mic most likely a condenser. You would also likely cut some low EQ to make the bottom more clear in the recording. Most vocals benefit from an EQ cut of -6db at 120hz. Not much information in the human voice below 200 hertz.

Lastly: gain structure.

Gain structure is setting the levels of the mic pre, the compressor, the input of you mixer or analog to digital interface and the recoding software itself. This is a very important step and one that must be done correctly. If any of the points in the ‘signal chain’ are to loud there will be distortion. If any points are too soft there will be lack of clarity and or noise as the piece following the quiet piece of equipment will need to be turned up by the following link in the chain and will likely add some noise as it turns up a ‘too soft’ signal.

One last point is that to record vocals you need some isolation. This can be achieved by hanging some blankets in a V in front of you and behind the mic. The point is that you do not want too much room sound getting into the mic. This allows you to add the ambience later. If there is already a ‘roomy’ sound in the recording it cannot be removed.