Six years ago, I recorded my first podcast episode in an old warehouse in West Oakland. My co-host and I couldn't afford mics then, so we planned to pull our audio directly from the camera we were using. We set up a wobbly table and wooden chairs beside a gorgeous south-facing window and pressed record.

Before I listened to the raw audio for the first time, I thought back to all the laughter and wisdom we shared. I was sure the episode would be incredible.

Then I pressed play. It was unlistenable. I could hear every car passing and every beat of the music playing outside that big, beautiful window. Whenever our bodies shifted in the old wooden chairs or the table wobbled back and forth against the cement floor, the microphone picked it up.

But believe it or not, none of those proved to be the kiss of death for our would-be podcast. What killed us? Our audio leveling. I had no idea what audio leveling was, or why it mattered. So like a lot of creators, I learned the hard way. Hopefully by reading this, you can avoid that painful process and nail your levels from the start.

What is audio leveling?

My lesson from that day: to create a respectable-sounding podcast you need high-quality audio, and getting your audio levels consistent is a big part of that.

Audio leveling is the part of your podcast audio production workflow that ensures that all of the sounds in your recording are balanced and bearable to the human ear.

Audio that’s too loud, too quiet, or — worst of all — too loud at some points and too quiet at others, will probably frustrate your audience — and may even lead them to dive for the stop button. It’s your job to ensure that your podcast audio levels are balanced. It’s a skill that takes some time to master, but with a few podcast recording tips, the right editing tools, and a little know-how, you can make your audio sound more professional.    

The ideal audio level for a podcast

The good news? Creating balanced audio is far easier than many other kinds of balance, like riding a unicycle, or holding down a job while you try to break into the zydeco music scene.

The standard of measurement of the loudness of a piece of audio is LUFS, or “loudness units relative to full scale.” Although there isn’t an industry standard for loudness, the ideal audio level for a podcast should fall between -20 and -16 LUFS.

Monitoring audio levels  

Most audio interfaces and recording devices allow you to set an input level or monitor your loudness levels while you're recording. When possible, it's best to have someone on your production team listen to the audio through headphones, so they can catch any spikes or drops in audio levels.

Along with listening, it’s also a good idea to watch your audio. That is, monitor your audio levels by watching the peaks and valleys on your waveform — the  visual depiction of how loud or soft your recording is at any given moment. Your goal is a steady and even waveform, avoiding periods where the audio spikes (too loud) or sinks (too quiet).

No matter how careful you are during recording, you’ll inevitably bump into some hiccups with your audio levels. And even if you don’t identify as an audio engineer, with the help of software, you can channel their skills to right the ship in post-production.

Manipulating audio levels  

Here are two post-production tricks to improve audio quality and make your podcast sound a lot better.

Equalizing (EQ):  Equalizing boosts or reduces various audio frequencies to improve the clarity of your audio and reduce unwanted background noise. With Descript’s podcast EQ setting, you can overlay the effect on your entire composition, individual tracks, or multitrack sequences.  Learn more about using Descript's Equalizer tool here.

Normalizing: If you find that your audio is too loud, you may want to opt for a normalizer. Volume normalization is a feature you can use to adjust the overall volume level of an audio file. This allows you to ensure your audio is consistent across the entire project.

In Descript, we have two types of normalization:

• Peak, which measures the volume of the loudest part of your audio file. It then uniformly boosts the volume of the entire file until that point reaches 0dBFS (the maximum allowed volume for a digital file).
• Loudness (-24 LUFS), which normalizes according to the human-perceived loudness of the overall track. This method uses the EBU R 128 standard.

Learn more about using Descript's Normalizer tool here.