How to Sound Good (Part 2) - Plugin Basics and Gates

This is part 2 of my ongoing “How to Sound Good” series. If you missed part 1 , maybe start there first. Or don’t. Whatever. I’m not your mom.

In part 2, I’ll be going over the basics of working with digital audio (just enough to be dangerous), and the basics of noise gating.

Working with plugins

So far, we’ve only been talking about hardware (microphones, mic stands, etc). If you have the space and a whole lot of money, you can stay entirely in the world of hardware. Most commonly used audio software has at least some hardware device that will do something similar, and it tends to sound fantastic. That kind of audio processing is called “outside the box” (OTB) processing. If I could live in that world, I absolutely would. But I travel a lot, so I live almost entirely in the world of “in the box” (ITB) processing, where nearly everything I do happens on my computer.

What’s are plugins?

In the world of ITB processing, you generally use plugins to do the actual processing. A plugin is simply a small software module that takes one or more audio signals, maybe does something with it, and probably outputs the result. If that sounds vague, that’s because it is. A plugin is just software that within other software to add some functionality, so an audio plugin is audio software that runs within audio software to do audio things. They are usually formatted as VST (most common), AudioUnit (aka, AU. These are MacOS only), or AAX (if you use Pro Tools). For the most part, VST is your safest bet. It’s supported by almost all audio software and most plugins are available in that format.

The vast majority of common audio software is in the form of plugins, which modify the audio signal as it passes through, to do things like distort the sound, adjust its volume, etc. I know it might be kind of unclear if you’re new to the world of audio, but it’ll make sense soon.

Note: I’m overloading the word plugin here. Usually a plugin refers to something built by someone other than the host software that is running the plugin. Most audio software also includes some basic modules that work right alongside plugins and do similar things. For the sake of simplicity, I’m just calling all audio processing modules, regardless of where they came from, plugins.

Getting set up

The first thing you’ll need is some sort of plugin host. A plugin host is software that takes an audio signal (like your microphone’s audio output), runs that through a series of plugins, then outputs the processed result as a new audio signal.

There are a lot of really good options, and the one you choose really just comes down to what you prefer and how you want to use it.

If you have any plans for making music at all, then I’d use a digital audio workstation (DAW). A DAW gives you basically everything you need, from recording, to editing, to mixing, to mastering, that you’d need to produce a piece of music. You can also use them for things like podcasts, video, etc. Some popular options are Ableton Live, Logic Pro, Garage Band, Reason, Bitwig Studio (my personal favorite), Pro Tools, Cubase, Studio One, FL Studio, Reaper, Cakewalk, etc. If you want that kind of power, then just bite the bullet and get a DAW.

If you don’t need the full power and fury of a DAW, and truly just want to process your voice for remote meetings, then the magic term to search for is “plugin host” or “vst host” – a VST is a popular plugin format, and a lot of people use the terms “plugin” and “VST” interchangeably. As I said before, a plugin host simply routes audio through a series of plugins and outputs the result. They don’t provide the full set of mixing tools and stuff that you’d get in a DAW, which makes them cheaper and simpler to use. Before I graduated to a full DAW, I used Rogue Amoeba Audio Hijack, which was pretty great. You can also use things like Gig Performer, Element, VSTHost, etc. Their UIs can be a little dense, since they are focused almost entirely on routing audio signals, but they are very powerful.

Routing

…and audio input device, which would should set to your microphone, and an output device would you should set to a virtual interface. A virtual interface is a piece of software (notably, not a plugin), which pretends to be a real audio device. They usually have a couple of inputs and a couple of outputs, and anything you send into the inputs is immediately outputted. Blackhole and VB-Cable are both great free options. If you need something more powerful and use MacOS, then Rogue Amoeba Loopback is a fantastic choice. My Loopback review also goes into a little more detail about virtual interfaces. To use the virtual interface, you just install the software (and in the case of Loopback, add an interface), then set it as the output in your audio processing software. You then set the virtual interface as the input or microphone in your meeting software. Any sound that comes from your DAW or plugin host will be heard by everyone else in meetings!

You will usually need to do some basic setup in your audio software to connect your input and to the virtual interface. For example, with Audio Hijack, you have to create a project, then add “blocks” for the input and output devices, which it automatically connects. You then have to hit a “Run” button to make it actually start listening to your mic. In something like Gig Performer, you need to manually connect the nodes in the interface.

It’s a bit tricker with most DAWs, where you need to create an audio track, set its input to your device, and make sure the DAW is routing the track output to your master output. You’ll also need to make sure the master output is actually outputing to your virtual interface. And you need to make sure the track is “armed” so that it’s actually listening to your mic.

Adding plugins

From here, you’ll have some way to add plugins in between your input and output somewhere. In plugin hosts, that tends to be pretty visual and simple, where you drag and drop from a list somewhere or right-click and select the plugin from a menu. You might need to manually route the audio signal through the plugin.

DAWs are all a bit different. Rather than run through them all here, I’ll just cover Bitwig. In that, you’ll either see a gray box with a “+” button on a track (click that to add a plugin) or you can click on the track name, and press “D” on your keyboard to show the “devices” panel, where you can add devices (Bitwig’s term for plugin). You can also click the magnifying glass icon in the lower right, to bring up the “browser panel” to search through plugins in a nicely organized way.

Usually, the plugin’s UI will automatically appear, but you might need to click on it or on a nearby button to actually access the plugin’s controls.

Unwanted sounds - tackling our first audio problem

At this point, you’ve done a lot of work, and all you’ve accomplished is setting a complicated way to connect your mic to your meeting software, so what now? Well, the first and easiest problem to deal with is blocking unwanted quiet sounds.

The problem: there’s a TV in the next room. Or your kid is singing outside. You’re near a busy road. Maybe your fancy new mic has a bit of a hum to it, or you’re suddenly hearing the sound of your room, which you never noticed before you went out and dropped $200 on a new mic. You don’t want everyone else in your meeting to hear that, so let’s do something about it!

Gates

A gate is a simple idea: when your audio is below some threshold, the gate is closed and no sound passes through. When the audio is loud enough, the gate opens, letting it through.

If you’re using a DAW, it probably has a “gate” or “noise gate” plugin available. If not, there are a lot of good options out there. If you’re using MacOS and software that works with AU plugins, you even have one already included with your operating system! You’ve got the full range of options here, from free all the way up to something outrageously expensive like FabFilter Pro-G ($200 for a 14-year-old gate plugin absolutely blows my mind).

Here’s a quick rundown of some options:

Your DAW’s built in gate

This is a fairly typical DAW gate, with all of the basic controls. Perfectly capable, with a nice UI. I particularly like this style of UI, where it shows the signal over time, along with the gate opening and closing. When adjusting a gate during a meeting, this visual feedback is invaluable.

AUDynamicsProcessor

This one is included with MacOS, and is surprisingly decent. It’s not specifically a gate, but is instead, a generic dynamics processor. In audio, dynamics refers to the range of volume, and dynamics processors include gates, expanders, and compressors. Apple’s plugin is a nice little Swiss army knife of a dynamics processor, covering all of those. When I really like about it is how well it illustrates dynamics processing as a concept.

Front and center is the curve display, representing the transfer function. It shows what the processor will do to your signal at different levels. The X-axis is the input, and the Y-axis is its output. You can see that the output will stay silent until the input reaches about -30 dB (it’s negative because 0 dB is maximum volume). Input is -60? Output is -♾️. Input is -40? Output is roughly -♾️. Input is -25? Output is also -25! That’s a gate! (well, technically, it’s a wave shaper, based on that explanation… You need to add the attack and decay stuff at the bottom to be a gate)

I won’t go into it here, but you can also see how a compressor would work: rather than messing with the transfer function’s lower threshold that reduces volume below that threshold, you mess with the upper threshold, above which it lowers volume (thus, reducing or compressing the dynamic range of the signal). Similarly, if you adjust the slope of the line below the lower threshold, it doesn’t just cut off the signal below the threshold; instead, it will lower its volume, thus creating an expander. This makes the quiet parts quieter, expanding the dynamic range. Expanders are nice because they behave very similar to gates, but don’t sound quite as jarring.

Unfiltered Audio G8

G8 is one of the better premium gates I’ve used. It’s only $35, and regularly goes on sale for $20 or less, which is nice. The UI is a little ugly, but you get everything you need, with the display showing what’s happening over time. It also has some extra controls that aren’t in every gate (lookahead, hold time, hysteresis, reduction, mix, input analysis mode, and fairly capable sidechain filtering).

iZotope Neutron 3 Gate

This is really more of an expander than a gate, but it works really well and has a very intuitive UI. Unfortunately, I don’t think you can get the gate on its own; you have to get Neutron. Fortunately, I believe the gate is included in the Elements edition of Neutron, which frequently goes on sale and is often even available for free!

It doesn’t have quite as many features as G8, but for some reason, I’ve always found this gate to have one of the more intuitive UIs. It’s easy to quickly set your threshold and other parameters, and as an expander, it sounds more natural than most gates I’ve tried. Neutron 5 is out now, but as far as I can tell, its gate is basically unchanged since version 3.

Gate controls

Threshold

The main control in any gate is the threshold. Below that level, the gate closes and you hear silence (or in the case of expanders, it’s much quieter, if not silent). This is entirely dependent on your voice, your microphone, and your environment, so you can’t depend on any plugin presets for it. Fortunately, it’s easy to set yourself: talk for a second to see how loud your voice is, and then set the threshold somewhere between that and whatever the signal is when you’re quiet.

The lower you set the threshold, the more sensitive the gate will be, which may result in random pops as other sounds occasionally slip through. I tend to watch the gain reduction meter in my gate (this shows exactly what the gate is doing, live), and set the threshold juuuuust high enough for it to stay closed when I’m quiet. If I’m near a TV or something, I might go a little higher, just so the TV doesn’t bleed through as much.

Attack time

The next most important control on a gate is the attack time. This is how long it takes for the gate to open when the single goes above the threshold. If it opens suddenly, you get a bit of a pop sound, which sounds really unnatural, so raising the attack time to 30-50 ms makes it open quickly enough to still let your voice through nicely without the jarring explosion of sound you get when starting instantly from complete silence. Setting it too high will make it so bits of your voice can get cut off, if the gate opens super slowly.

Release time

The release time is similar: it determines how long it takes for the gate to close again after you stop talking. Again, if it’s too low, you’ll sounds really choppy and unnatural as your voice suddenly cuts out when you stop talking. You have quite a bit of wiggle room, though, since any background noise will fade away quickly even with fairly high settings. Set it too high, though, and you might as well not have a gate, since it will just seem to stay open all the time. 100-300 ms usually works well for me.

Depth

Not all gates have this parameter, but it can be handy when they do. This controls how much the gate reduces the signal when the gate is closed. If it’s -♾️, then you get total silence. You could also do things like set it to -12 dB, which would cut your volume by about 75% instead of introducing complete silence.

Hold

This is another somewhat uncommon parameter. It controls how long the signal needs to stay below the threshold before it starts to close. It can help with some of the unnatural sound that you get from a short release time.

Hysteresis

This one is even less common, and goes by different names (e.g., G8 calls it hysteresis, while Neutron calls it Open/Close). This separates the threshold for opening and closing the gate at different levels. So if you set the threshold at -30 dB and the hysteresis at -3 dB, the gate will open at -30 dB, but won’t close again until the signal drops below -33 dB. This helps keep the gate from closing inadvertently, by making it so you have to be even quieter for it to stop. It’s really more useful for things like reverb tails or guitar strums.

Mix

This isn’t specific to gates, by any means, but not all gate plugins include it. In plugins, the mix control refers to the mix of unprocessed (aka dry) and processed (aka wet) signals. At 0%, the gate is doing nothing at all, since it’s just giving you an entirely dry signal. At 100%, you only hear the processed sound. At 50%, you get half and half. When your processing is getting to be a bit too much, you can mix in some of the dry signal to bring some life back to it.

Sidechain

The sidechain on a plugin is an extra audio signal that affects what the plugin is doing, but you don’t directly hear that signal in the plugin output. In gates and other dynamics processors, that tends to be either an external sidechain signal (useful for cases where you want to do things like speak over music, and you want the music to automatically get quieter when you’re talking) or an internal sidechain, which is a modified version of your input signal. Generally, if there’s an internal sidechain open available, the plugin also provides some ways to filter it, allowing you to do things like control the gate with only high- or low-pitched sounds.

I tend to use the internal sidechain feature quite a bit with gates and expanders when I’m using them for remote work. Often, one of my biggest unwanted sounds is my daughter talking or singing nearby. She can be pretty loud, so I’ve found gating works better if I block the higher frequencies in the sidechain, so that only my lower voice will make the gate open.

What’s next

So you’ve got your nice mic all set up: running through a plugin host, then a gate or expander plugin, and finally to your meeting software via a virtual interface! What now?

Well, like always: listen to yourself.

You’ll notice that even though you’ve effectively blocked most unwanted sounds, when you’re actually talking, there’s a lot of noise. You voice probably sounds a little fuzzy and might even echo a little bit as it bounces around your room. This leads to our next stop on the path to sounding good: noise reduction.

Stay tuned for part 3!