There are a lot of little noises that can sneak in between phrases. These extra noises are not noticed while the vocalist is singing but you don’t want to hear bleed from headphones, rustling paper, or any other distractions in between. It is quite common in the home recording studio world to put a noise gate on vocal tracks to eliminate these types of problems. Tuning a noise gate to correct this can be a time consuming process which is prone to creating one or two unnatural sounding words here and there. I prefer a more natural sounding approach.
Noise gate on vocals
The problem with putting a noise gate on your vocals is the ease with which the vocals can end up sounding gated. If you want them to sound gated this is fine. Most often it is the case that we don’t want them to sound gated but do want them to sound clean, polished, and professional. It is possible to configure a noise gate to produce natural attack on vocals. Often the noise gate settings that sound great on one section of the song don’t work on another sections. This leads to much automation of the noise gate settings to change throughout the song. It is also not uncommon to think you have the noise gate set up perfectly for the vocals, but on the hundredth listen while mixing, you discover there is one word right in the middle that sounds gated and unnatural. Sure, you can fix this one occurrence, but a few hours later (on the 150th listen perhaps?) you will hear another word that has started to sound gated. Working in the home recording studio world of modern DAW software, we can forget the noise gate altogether and quickly clean up the vocal tracks ourselves.
Manually cleaning up vocals
Instead of using a noise gate we will manually trim each region of vocals to eliminate silence. This gives us complete control over where the vocals come in and out without the pains of trying to balance a noise gate’s threshold, attack, release, and hysteresis controls. Take a look at this bit of a raw vocal track I recently recorded in my home studio:
Raw, untrimmed vocal track
The sound clip pictured represents a single verse with four vocal phrases. Notice the circled low level noise between each phrase? This is what we are trying to eliminate. In order to use this technique you are going to need to know how to use your home recording software enough to split sound regions like this one and resize them (usually by dragging their edges). If your studio software has any kind of snap to grid feature, be sure it is off. The first step is to split the sound region right in the middle of each section of silence to be removed.
Split in the middle of each noise section
You can see my sound file has been split basically around the middle of each of the previously circled sections of noise. The next step is to grab the edges of individual regions and roughly bring them in around the beginning and end of each phrase. You will end up with something like the following picture.
Sound regions have been roughly resized
Once you learn your way around your home recording software you can get up to this point in about twenty seconds. Now we are to the slightly tougher part. We are going to finely tune the beginning and end of each vocal phrase. Zoom in on the beginning of the first phrase.
|What we see here is the vocalist taking a breath in preparation for singing the first line, followed by the attack of the first vocal syllable. Since this is the first vocal phrase and is preceeded by a portion of musical silence, I have decided to leave this initial breath in to create a moment of anticipation of the vocal grand entrance. Rather than trim out this human element, I’m going to create a short fade to enhance the excitement.|
|I have chosen to add a logarithmic fade (blue line) that starts at the beginning of the breath and ends just as the actual vocals are about to kick in (white line). I have chosen to keep the view zoomed in which doesn’t show you the whole fade but better illustrates the successful removal of the noise at the beginning of the file. Compare the view to the right with the view above. Both are at the same zoom level but now the breath is not preceeded by extra noise.|
With the beginning of the first vocal phrase trimmed, we are ready to move on to the end of the first phrase.
There are three things easily noticed at the end of this phrase. A very pronounced ‘t’ sound at the end of the word beat. This is followed by a sharp click which was probably caused by plastic adjusting on the headphones as the vocalist grooved while singing. Then you see the onset of the bassline bleeding from the headphones and being picked up in the vocal mic. We want to keep the ‘t’ sound and remove the click and headphone bleed.
|I pulled the end of the sound file to just after the ‘t’ sound and again added a short fade (blue line) to smoothly bring the vocal out before the click comes in. At first you’ll have to do quite a bit of auditioning to get the timing of the fades just right, but as you familiarize yourself with your home recording tools the process gets much faster. The ends of phrases are where vocalists tend to divert their attention and rustle lyric sheets, adjust headphones, smack lips, or make other undesirable noises. Pay special attention to these areas.|
Follow these same techniques for the beginning and end of the other three vocal phrases and you may end up with something like this:
I have modified the zoom levels on this track to show all the fades and edit points. Pay special attention to the custom fades at the beginning and end of each vocal phrase. This is the type of organic, changing treatment per phrase that you will lose with a noise gate. The breaths we save at the beginnings of some of vocal phrases are especially troublesome for a noise gate. A breath will often not be much louder than the noise or sometimes it will even be quieter. This causes a problem for a noise gate which might open momentarily only at the loudest part of the breath. Noise gates can also harm the initial transients of the first word in a phrase which is part of what gives the vocal that gated sound that is so artificial.
This technique is not difficult to learn and is easy to master quickly. I can do this type of editing on a song in about ten or fifteen minutes. You would easily spend that much time (likely much more) setting up a noise gate and you won’t have as natural a sound. Drop those noise gates and get to cuttin!
Your home studio is your playground, have fun!