Whether you’re recording a complete drum kit or have a sub-optimal recording you need to fix, noise is always a factor we need to consider. In this article, I will cover noise gate and noise reduction in-depth.
These might seem like intimidating subjects; I am here to demystify any misconceptions you might have about the two and teach you how to discern whether you need to use noise gating or noise reduction in your day-to-day workflow. So let’s get started! What are the differences between noise gating and noise reduction, and when should you be using them?
A noise gate aims to remove unwanted noise/s in a recording, leaving purposefully recorded audio completely intact. More specifically, targeting only areas when the desired sound in your recording is not playing. To put it plainly: A noise gate will block all unwanted sounds from being heard.
For example, a snare drum mic recording some of the cymbals, toms, and kick drum. Let us assume the snare drum is the loudest part of the recording. We would set the threshold of our gating plugin or hardware unit to allow the snare through, and when it isn’t playing, the track is silent. You could also use a noise gate while recording to prevent these unwanted noises from even being recorded. This technique, however, is not recommended as you might end up gating out audio that you want.
Noise gating is a powerful tool that, when used correctly, will improve the quality of your audio immensely.
In the above diagram, the Output Level is what you will hear.
Noise reduction is distinctly different from noise gating as you would most often use it on a sub-optimal recording with a constant or ‘static’ noise present. Typically, this process would target the entire recording. As an example, a podcast with white noise or buzz throughout. A noise reduction program or plugin is needed to reduce the noise to the point where it is inaudible relative to the loudness of the spoken words. We do this by telling the program what it should consider noise and then telling the program to reduce that sound only.
One might also consider closing a window or turning a fan off as noise reduction. For the sake of this article, noise reduction will refer to the process of removing unwanted sound from a recording, specifically when the desired sound is present.
Note how this process reduces the noise, and the desired sound is essentially untouched.
So how exactly do we go about working with a noise gate or a noise reduction program? What do all the controls on a noise gate plugin or hardware unit do? How will you be able to tell if a recording is ‘fixable’ or not? I will be answering these questions and more in the following article.
How does a noise gate work?
A noise gate works by setting its parameters to remove unwanted sections in a recording and allowing the desired parts to play. If the gate is closed, the recording is inaudible. Any noise gate worth using will have four basic parameters you can tweak: threshold, attack, release, and ‘hold’ or sustain.
Arguably the most important control is the threshold. How much of the signal we will hear is determined by it. The audio won’t come through if it is set too high, and if it is set too low, no gating will occur. Finding the correct level to use is entirely dependent on the loudness of the audio signal. You’ll need to experiment a bit to find the sweet spot.
The attack parameter will determine the speed at which the gate opens. Sounds with natural transients will typically use a faster setting, while smoother sounds will often use a more liberal one.
The release of a noise gating plugin controls how fast or slow the gate closes. A slower release time will allow the sound to fade out naturally, while a faster release time will shorten the tail of the sound. What to set the release to will depend entirely on the audio and what effect you want.
The hold function determines how long the gate will stay open before the release function takes effect. While this function isn’t as crucial as the others, it is essential to note that setting the hold to a higher value will allow audio below the set thresholds loudness through. ‘Hold’ could work well for a recording that has a high dynamic range.
When should I use a noise gate?
Ideally, you want to use a noise gate on any recording with unwanted noises in it. Whether it’s sound spilling from the headphones of a vocalist, or your neighbors speaking loudly, a noise gate used correctly will drastically improve the quality of a recording.
You’ll see the best results when using a noise gate on a ‘dry’ signal. i.e., Before you place effects like reverb or delay in your signal chain. There are cases when you might use a gate after ‘wetting’ effects like reverb to control the reverbs tail; or if the recording has reverb embedded in it. You could also use a gating plugin to achieve a ‘stuttering’ effect. In this case, the gate will ignore the signal and act according to parameters set by the user. Listen to the intro of the following video for a practical example of this effect.
Stuttering effect in use:
How does noise reduction work?
Noise reduction works by attenuating unwanted noise/s or sounds that are audible simultaneously as the desired audio. It can reduce high-frequency hissing, component buzz or hum, or even the sound of a computer fan (as a few examples) to the point where the signal-to-noise ratio is acceptable.
For this process to be effective, a passage in the recording where only the noise is present is needed so that the program can correctly identify what it should consider noise. The program will then attempt to attenuate only this sound for the entire recording, leading to a better signal-to-noise ratio.
Note: This process can be quite taxing on the quality of your audio, as it might attenuate similar sounds that share the same frequency range as the noise. For example, High-frequency mouth sounds such as F’s, H’s, and S’s will share some of the frequencies with a hissing sound. So be careful! Always be conservative with this process, and it will more often than not serve you well in day-to-day applications.
What is a signal-to-noise ratio? Simply put, the signal-to-noise ratio is the difference between the level of the wanted and unwanted sounds in a recording or environment. The higher it is, the better. The decibel (dB) is the typical unit used to express this ratio, and you calculate it by subtracting the noise level from the signal level.
In real-world scenarios, a perfect ratio is virtually unattainable as there will always be some degree of noise in a recording.
When should I use noise reduction?
If there is audible noise in a recording while your desired audio is present, it’s time to consider using noise reduction. Let’s say we recorded some killer vocals, but we recorded the ceiling fan in the process. If the fan is too loud, to the point where it’s audible while the vocalist is singing, the signal-to-noise ratio is unacceptable.
Our options at this point are to either re-record or use noise reduction.
‘Constant’ noises (such as buzz or hum) aren’t the only sounds that noise reduction can target; It can also reduce mouth ‘clicks’ or even guitar ‘squeaks’ (sliding your fingers along the strings rapidly). Check out the following video for a quick tutorial on removing guitar squeaks.
How to tell if a recording is fixable?
‘Fixing’ a recording is never ideal, as obviously, sound quality is paramount. However, if you’re working with a podcast, mobile phone recording, or any sub-par audio, a lower signal-to-noise ratio is more often than not something you’ll need to address.
As a solution, we could use noise reduction to improve the sound quality to the point where the signal-to-noise ratio is acceptable. Suppose the signal-to-noise ratio of a recording is less than 1. In that case, the desired sounds are relatively softer than the noise, and the recording is approaching the point of being unfixable.
When it comes down to it, you need to ask yourself one question: What is the recording’s final destination? Is it a record label? Perhaps it’s a podcasting platform? Or is it simply for personal use? The answer will help you determine if you can get away with less than flawless sound quality.
What program/s should I use for noise reduction?
The marketplace for this type of plugin or program is brimming with options for you to consider. I’ll briefly be going over a few paid and free options to get you started on the road to better sound quality.
Acon Digital Extract: Dialogue
Designed exclusively for noise reduction in vocal recordings, Dialogue can effectively reduce any noise in interviews, podcasts, or even singing. This plugin is ideal if you don’t want to do too much tinkering, at it does most of the work for you.
Sonnox Oxford De-noiser
The Oxford De-noiser has a much more hands-on approach. Everything will need to be adjusted manually to achieve your desired outcome; having said that, you can get quite surgical with targeting specific types of noise.
iZotope RX 8
iZotopes RX8 is a plugin bundle dedicated to audio repair. While simple noise reduction isn’t this bundle’s only function (far from it), It is hard to find anything that comes close to Izotopes RX series’ ability to remove unwanted noise from a recording. If you need a little more convincing, check out this video:
Audacity is a separate program that functions independently from your DAW. While not as powerful as some paid options, it does the job regarding noise reduction. As far as free programs go, Audacity is a reasonable choice if your budget is a little tight.
Check out this tutorial on how to remove noise in Audacity:
Bertom Audio De-noiser 2
Denoiser 2 is a plugin designed specifically for podcasts, post-production, and music. It has an easy-to-learn interface, with intuitive green vertical bars to show you the occurring noise reduction. If you’re looking for a quick and straightforward answer, Denoiser 2 is the plugin for you.
ReaFIR is an all-in-one noise suppressor, equalizer, and compressor. This plugin requires you to create a noise profile for it to work as intended, and in that way, it is similar to Audacity. It also consumes a relatively low amount of CPU, so it’s excellent for lower-end PCs.
Whichever program you decide to go with, you’ll be well on your way to producing better audio.
What could cause unwanted noise in a recording?
There are many different types of unwanted noise you could encounter in your recordings; however, we can divide them into two main categories: External and internal. These two categories are separated by whether a noise occurs inside or outside your electronic setup.
- External noise
Noise appearing in the audio that the microphone has physically recorded. Whether it’s Bass rumble or wind sounds, these external sounds all have one thing in common: they all occur outside your electronic equipment.
- Internal noise
Noise appearing in the audio that the microphone has not physically recorded. From grounding loops to faulty computer components, all these types of noise occur inside your electronic setup.
For a more detailed list of what could cause unwanted noise, check out this article.
Tip: If you’re having trouble telling if noise/s in a recording occur/s internally or externally, listen for reverb. Internal noise will be ‘dry’ sounding, while external noise will more often have a roomy or spacious character.
So to summarize: Noise gates target noise/s occurring when the desired audio is not playing, and noise reduction targets noise happening when the preferred audio is playing.
Hopefully, by now, you know the difference between noise gating and noise reduction and when to use each. In a perfect world, noise gating and reduction would never be necessary. All we can do is use the tools available to us to produce the best audio quality possible.
Cameron is a practicing sound engineer and music producer with over a decade of experience. Based in South Africa, he is also one-half of the production duo 2wice Shye. When in between writing songs,
he loves to create sound effects and soundscapes and tinker with production technique ideas.