Audacity Voice Recording Basic Editing

This guide covers how to create, post-process and export a voice recording quickly and without advanced skills, using Audacity 3.

Audio Setup

Everything outside of Audacity is too large of a topic to possibly cover it in brief. Due to different combinations of microphone technologies and connectors, interfaces, operating system settings, and other factors, there is no way to provide universal guidance there. You should be aware that if you're not satisfied with your recording quality even after following the guide in full, it will most likely be an issue with the rest of the setup.

In Audacity, before being able to record and edit, you want to set your input device (microphone) and output device (headphones or speakers). You can do so either in Edit → Preferences → Audio Settings, or by clicking the Audio Setup button in the main toolbar and selecting devices from there.

If you want to monitor your voice (hearing your own voice through headphones) while recording, it is highly recommended to do so through an interface instead of through Audacity/other means on your computer, as the latter will introduce a delay.

Recording

First, a recording has to be created. For this, create a new Track via Tracks → Add New → Mono Track (assuming a single microphone is used). The track will be created, and selected by default.

To start recording on the active track, either press the record button in the main toolbar, or use the hotkey (R by default). To eventually stop recording, press the stop button in the main toolbar, or the hotkey (spacebar by default). Whichever option you choose, try to make sure the activation or stop input does not get captured by your microphone.

To pause recording, either press the pause button in the main toolbar, or use the hotkey (P by default). You could alternatively use the stop function to pause as well, however that would result in multiple segments, which you would have to combine for easier editing.

Independent of the overall requirements of your recording, you want to make sure your recording has the following qualities if possible to make the post processing process as easy as possible:

• Maintain a constant distance to the microphone throughout the recording.
• Record an extended time (about twenty seconds to a minute) of room ambience (meaning you sitting in front of the microphone just like if you were recording, however without making any intentional sound) for the purposes of creating a noise profile later on.
• Make sure you leave at least a full second of silence before pausing or stopping of the recording to ensure that no syllables are cut off.
• Don't change the environment while recording, which means leaving objects where they are, don't change the position of electronic devices in your proximity, don't open any doors or windows connected to the room you're recording in, and things like that.

After you finish recording, but before working on post processing, you might want to save an unaltered version of your recording in a high-fidelity format such as wav; to do so, follow the steps described in the Export section below.

Post Processing

There are a number of individual methods to improve your recording, but not all of them are necessary. Listen to your recording, try and identify problems, and use the appropriate sections to try and mitigate those problems.

The following sections are listed in order. Maintaining this order isn't ultimately important, however you might get better results for some recordings by sticking to it.

In particular, since loudness normalization is the final step, you might have to turn up your volume to ridiculous levels in order to check whether the individual steps did what they should.

Basic Concepts

Keep in mind that all audio has noise - there is no such thing as a voice recording without noise, just one with a high signal-to-noise ratio. No matter what your recording quality is, the purpose of post processing is to reduce noise as much as possible without impacting the signal quality much.

Audio recording is rather quiet by nature in comparison to processed audio, so don't try and solve this by increasing the microphone's gain.

Furthermore, volume and loudness are not the same thing. Loudness is the "volume" of a signal within a medium, while volume is a setting on the device it's being played back on. You could increase or decrease your volume to make the recording sound good for you on your device, but you instead want to adjust the loudness so that it will presumingly play as expected (based on the device's volume settings) on every device.

Noise Reduction

With the room ambience in the beginning of your recording, you can subtract the ambient noise from the recording.

First, mark the room ambience area of the track. Then go to Effect → Noise Removal and Repair → Noise Reduction. A window that's separated in two steps will open. Ensure that your silence mark is still active, then click Get Noise Profile. After a very short processing, the window will close, and the noise profile will have been created in the background (there will be no further confirmation of any sort, which might be a bit confusing).

Second, mark the entire track that you want to apply noise reduction to, then navigate to the previous dialog once more. The default options should work, but you might want to increase the Noise Reduction (dB) value if your recording has a lot of noise. After confirming, the track will be processed and should now feature a lot less noise. In particular, the segment of room ambience in the beginning of your recording should be completely silent now.

Noise Gate

Noise gate is a method to reduce sounds below a certain volume, and as such can be used to get rid of background static, clicking noises, or breathing or handling noises. It is, however, a less sophisticated method than the previously mentioned noise reduction (in the sense that it might noticably distort the to be preserved parts of the recording to a degree to achieve its purpose), and should only be used in case noise reduction isn't sufficient.

To apply a noise gate, select a track or mark a track segment, then go to Effect → Noise Removal and Repair → Noise Gate, and set the desired values.

The settings heavily depend on the loudness of your recording as well as the nature of the issue you're trying to fix. You can however estimate the correct threshold for a section by marking and playing it, and then getting the maximum loudness from the playback meter in the toolbar.

Loudness Normalization

This step will ensure that your recording has loudness levels that are on par with other media to prevent jumps in perceived loudness between different recordings. If you work on multiple recordings, or together with others, you want to first decide on a unified loudness level that can then be applied to each recording. LUFS is a common unit to measure perceived loudness. For reference, -14 LUFS is a good starting point for music, while -16 to -19 works well for voice recordings.

To apply a certain level of perceived loudness, select the track to process, then go to Effect → Volume and Compression → Loudness Normalization. Make sure the normalization mode is set to perceived loudness, and set the value to your desired LUFS value. After confirming, your track will now be set to the specified perceived loudness, which depending on your source recording might make a massive difference.

Exporting

If you want to save your Audacity project file, for example because you recorded multiple tracks and edited cuts, you can do so by going to File → Save Project. That, however, will give you a project file that's meant to be opened by Audacity only, not an audio file that you can distribute for anyone to listen to. For the purposes of this guide, edits will be simple enough that doing them again might be easier than managing a collection of project files.

To export your track as an audio file, go to File → Export Audio. Here you can set a number of rather relevant options.

• The file format will heavily influence the quality and file size of your resulting audio file. Good default picks are wav for quality (for example if you want to make edits in the future), and opus for production export.
• Assuming you have a single microphone recording, make sure to select Mono channels, as Stereo channels would double your file size for no benefit whatsoever.
• Unless you have a specific reason, 48000Hz is the sample rate you probably want to pick.
• You can adjust the quality for the sake of reducing the file size, and depending on the file format you choose, you will have different options here.
• You can edit the file's metadata by clicking on Edit Metadata, however this is something you can always do after the fact as well.

Leave all other settings as they are unless you have a specific reason to change them, then export your file.