How to improve how your microphone sounds on Windows, for free, without “virtual cables” or lag

Fabio Tamagno
Incredible Machine
Published in
5 min readJan 22, 2021
we don’t have that much money for this

I like sounding a bit more like a radio presenter, and have that “HD feeling” to the voice recorded in my streams — audio filters help with this. Unfortunately, it hasn’t been too straightforward to apply the effects I desired across all the applications I use without caveats.

I dived a bit too deep into what for me is the barely-known (I am a computer engineer, not an audio engineer) and put together the most reasonable solution that could give me peace of mind, for free.

What do we want to achieve

To improve the output of your microphone you need to be able to enable VST filters without too CPU overhead or lag. For my current solution I am using the following filters:

  • noise cancellation — RNNoise, based on neural networks and somewhat lightweight
  • compressor — TDR Kotelnikov is powerful and free
  • equalizer — Marvel GEQ is easy to use with some sane presets

There are more filters you may want to add, like limiter, expander or gain, many of which are full-featured and free.

Solutions that didn’t work

I tried for days with setups based on the following:

  • using NDI audio streaming (too much lag and unstable)
  • using OBS audio filters with virtual cables (too much lag, too much CPU usage)

Lag is a huge issue. You wouldn’t want your video to be ahead of your voice. If you try to reduce lag by reducing the buffering, what you will get is choppy, crackling sound which completely defeat the purpose of the solution.

Solve the problem with Equalizer APO

Equalizer APO has sorted out all my needs spectacularly well.

It’s not the most user friendly software you might find, but here’s the best part of it: once you set it up, you won’t need to open it again, unless you want to tune your microphone again. The application intervenes directly against the Windows API which means you don’t need Equalizer APO running in your taskbar.

First configuration

You can install Equalizer APO from the following link: https://sourceforge.net/projects/equalizerapo/

As soon as it starts the first time, it will ask whether you want to open the configurator (Yes).

In the configurator, head to “Capture Devices” and select the microphone you want to apply the filters to

Then click Close.

You will now find yourself in the main interface with a new configuration open:

Press on + and head to “Select Device”

Then select your Microphone:

Which brings us to this status:

At this point we can install some filters.

Install VST filters

You can choose your preferred filters. For the purpose of this guide, we’ll install the ones I’m accustomed to. Install the Windows versions of the following:

Apply VST filters

For each VST filter plugin, click + again in Equalizer APO and reach for “VST Plugins”

You will not be presented with an interface to load your plugin:

Then click on the folder icon to find where the VST filters are installed. In my case, they are installed at the following paths:

  • C:\Program Files\Common Files\VST2\librnnoise_vst.dll
  • C:\Program Files\Steinberg\VSTPlugins\TDR Kotelnikov.dll
  • C:\Program Files\Common Files\VST2\Voxengo\Marvel GEQ.dll

Once you add them all, you should find yourself with this configuration:

Listening to the difference

On Windows it’s easy to listen to your own voice. Make sure you are using headphones to avoid audio loops, and keep the headphone volume low in case you make any configuration mistake not to send your ears into the shadow realm.

Right click on your volume icon, and reach for Advanced set-up

Then reach into the Recording tab and double click on your microphone:

On the Listen tab, select “Listen to this device” and press apply.

You should now be able to hear yourself. To stop hearing yourself, click on “Listen to this device” again to disable it and press apply.

Configuring the filters

In Equalizer APO click on “Open panel” for Marvel GEQ. Here’s my configuration that’s based on one of the presets. It makes my voice sound a bit deeper.

Remember to enable “Apply Automatically” to listen to changes in real time.

Again, in Equalizer APO, click on “Open panel” for TDR Kotelnikov. Here’s my current configuration, which might not work for you depending on microphone and environment.

Again, remember to have “Apply Automatically” enabled to quickly hear the changes.

Finally, RNNoise doesn’t have any setup —it just works like magic.

There’s no next step, ever again

The absolute best part of this setup is that you can forget about it. Use your microphone in any application on Windows, and the filters will be applied with no perceptible lag.

One note, make very sure your videocall software doesn’t add more filters. In example, for Zoom reach into the audio options and make sure you enable “Original Audio”, otherwise you’ll be applying noise cancellation twice.

If anything is unclear about this guide, reply in the comments below and I’ll clear it up.

Credits and resources

--

--

Fabio Tamagno
Incredible Machine

Talk about React.js, Node.js, PHP, Magento 2 and other terrible, terrible things.