Audio Tips for Content Creators

For the past couple years, I’ve spent ~~too much~~ time recording and editing dozens of videos, both for YouTube and courses. I’ve even done a couple livestreams. And one of the most frustrating parts is managing and fixing audio.

I’ve ruined my audio multiple times. Too much noise. Too loud, or too quiet. Even messed it up with my USB hub (more on that later).

So, for my sanity, I poured over dozens of tutorials and guides about audio quality. All to fix my audio. Here’s a collection of audio tips from my findings, from preparing to record to fixing what’s necessary in post.

The tl:dr — don’t overthink it. With even some of these basics handled, your audio is probably tolerable for most.

Most audio tips online are overly hardcore. They’re for noise-proofing rooms to record crisp song vocals and musical instruments for a hit album. Content creators are just narrating. You don’t need the hardcore stuff. Just the basics.

That’s what this post is meant to be. It’s on the longer side, as a reference guide, so strap in.

Pre-Recording

Choosing a Room

Stand where you will be recording. Turn on the microphone so you can see its levels. Listen for background noise. Clap. Talk loudly. Listen to how the room responds. Is there an echo? A ringing resonance that builds up with louder sounds? Does the microphone pick it up?

Try different rooms out, see which work well for you and your microphone. But remember: you don’t need silence. Just keep the volume of background noise and the volume of your voice as far away from each other as possible.

Be louder than the room. If the room is loud, you have to be louder. If the room is quiet, you don’t have to be as loud.

If you are recording audio but not video, you don’t need a visually nice setup, or even be in front of your computer. Many USB microphones can plug into a phone or tablet. I’ve recorded audio with an external mic + my iPad before.

So, where should you record? In short, anywhere with soft or foamy surfaces work. Even in suboptimal places, you can edit it to sound a lot better. Do what you have to do. If Markiplier can record multiple videos in a bathtub, then so can you.

Some common options:

Bedroom: mattresses, curtains, and blankets can dampen echo and resonance
Rooms with carpet, fabric couches, or any soft/foamy objects
Closet can work, but it can sound stifling or muffled in some cases

Avoid:

hard floors and walls with no decorations
skinny rooms
windows where it’s especially noisy outside

Reducing Noise

But when your chosen spot still has noise, how can you reduce it?

Turn off anything noisy, within reason. Fans, washer and dryer, other appliances. Watch your microphone levels as you turn off each one. See which makes a difference. Depending on the mic, you may be able to leave some on. So long as there’s a wide gap between the noise and your voice, you’re good.
- Some devices or wall outlets may hum. Try plugging them into a different wall outlet or port. These hums are often a very specific frequency, so you can remove it in editing. But the less you need to edit, the better.
- If your mic can pick up your PC’s fans, readjust the position and angle of your microphone or PC so they don’t point at each other. You can also slightly throttle the CPU so the fan doesn’t run full speed. On Windows, search and click “Edit Power Plan,” click “Change advanced Power Settings,” look up “Processor Power Management” => “Maximum Processor State.” During recordings, I throttle my CPU to 90-95% normal usage. Enough to keep fans below max speed, without affecting PC performance much. You can even save it as a separate power plan for easy swapping.
Place sound-dampening objects to reduce reflection and echo.
- Acoustic panels are expensive but great for more professional recordings. I got a couple made by a local woodworker who happened to know audio production. Those alone mitigated a lot of echo and resonance. But find some popular, good quality brands online, if you want to go the luxury route.
- For a cheaper but surprisingly decent solution: any soft, absorbent, or foamy objects work well. I used a moving blanket from “The Home Depot” when I was in an apartment. Carpet, mattresses, soft couches, even thick towels — anything like that will absorb some sound. That’s why closets famously work (though it can sound too stifling).
- But where to place them? A common rule of thumb I’ve read is “don’t have two hard walls parallel to each other.” Carpet often takes care of the floor and ceiling facing each other. With blankets or panels, hang them on the walls, about the same level as your microphone.
Talk quieter. No really. For more echo-y rooms, you can mitigate it just by talking quieter. Quieter voice, quieter echo.

Microphone Placement

Some people just hold the mic. If you’re good with that, or constantly moving anyway, that’s fine.

I initially placed a short microphone stand on my desk. If you’re just narrating yourself, and not working while you’re talking, that’ll be perfect.

However, the mic picked up noise from my typing on the keyboard as deep thuds. So for my first add-on course, I used a C-stand placed on the floor. Don’t overspend; they can be relatively cheap, <$50. Local filmmakers may have a spare one. If the floor underneath it shakes, you’ll hear it in the audio. But so long as my feet don’t brush up against the stand, I could type freely.

After the course, I bought a cheap boom arm and attached it to a separate desk. Mainly because my desk does not have enough lip to support a boom arm. It keeps my typing quiet and frees my feet. And it worked with my Podmic, saving me money from having to buy an “official” one.

Every microphone is different. From where you speak into the mic, to how far away you need to be. To prevent strong “S” or “T” sounds, it’s best to not speak directly into the microphone sensor, just across from it.

A good rule is: point the mic at your mouth, but don’t point your mouth at the mic.

Test different distances from the microphone. Ensure it can hear you, without making your voice too boom-y or breathy.

Buying a Microphone

Most microphones will work perfectly fine. Really, I’m a firm believer of that. I’ve seen creators record with Apple earbud mics, and it sounded fine. Or a mic off a Logitech webcam. Or a Blue Snowball. So long as your room isn’t as loud as you, and you’re not louder than the mic, you’ll be fine and can edit out most extra noise.

But as you record more consistently, you will quickly learn your microphone’s limits. Every microphone has a volume limit, where louder sounds are clipped. Adjust gain and speak quiet enough to avoid clipping.

As you see these limits, you’ll know when you need an upgrade. And when you need an upgrade, or don’t have a microphone to begin with, these are common criteria to decide from:

Condenser or dynamic. This is referring to how the microphone records its audio.
- Condenser mics are most common. A budget choice that works perfectly fine for YouTube. They are more sensitive to noise and have a smaller range (meaning, the range between silence and its clipping volume).
- Dynamic mics are more expensive, often what you see on bigger budget podcasts. They have a wide range of sound, meaning more resilient to background noise, and more directional in which sounds they pick up.
Speaking of, directionality.
- Directional mics are great if your head is not moving much, and you’ll have something loud, like a computer, nearby. Just make sure you’re pointing the right direction.
- Omnidirectional mics pick up sound in all directions. This lets you move your head around freely. But it will also pick up other noises, like your PC, more easily. Best for more on-the-go usage or in setups where people move around.
Check what output types the microphone supports. Does it have a USB port? USB-C? XLR? Does it support wireless? Depending on the ports available, you may need to buy adapters. USBs are more plug-and-play, whereas XLR has better audio quality and flexibility. Consider which device will be recording and what ports it has.

USB Bandwidth

In a couple videos last year, I got feedback that my audio suddenly became really muddy. I couldn’t figure out for the life of me why, until I checked my USB setup.

I use a USB hub, with extra ports for my mouse, keyboard, and gaming controller. And only for a few of my videos, I made the mistake of also plugging in my mic.

To explain why this is an issue, imagine each USB device sends information to the PC through this hub:

Mouse    Keyboard    Microphone
  *          *            *
  |          |            |
  \          |            /
   \_______USB HUB_______/
             |
             |
             PC

However, from the hub to the PC, all four devices share the same USB cable. For intermittent input devices like keyboards and controllers, they’re not affected much. But a microphone is constantly streaming data. So if data from the other three devices overwhelm the USB cable, it muddies the microphone data. It adds extra noise not in the original recorded audio. It can also increase latency.

To fix this, plug your microphone directly into your PC whenever possible:

Mouse    Keyboard    Microphone
  *          *            *
  |          |            |
  \          |            /
   \_USB HUB_/           /
        *               /
        |              /
        PC_____USB____/

If your PC doesn’t have enough ports, record audio on a separate device and sync it later.

Recording

Alright, the setup is ready. How do I actually record?

Prepare your mouth

Drink water for the couple hours leading up to recording. This will reduce spit in your mouth.
Breathing and singing exercises, like mouth buzzes, get your mouth loose. Think the first scene of A King’s Speech with how the announcer prepares, as reference (also a great movie intro).
Practice how you speak. For me, I speak like I’m singing: project with my diaphragm, and limit mouth movement. Moving your mouth a lot can lead to your lips smacking and “spit sounds.” Sure, you can manually mute those in post, but do you really want to? I’ve done it a few times. Avoid it if you can.

In short, anything that benefits a singer will also benefit you.

Rehearse

Rehearse, rehearse, rehearse. Some socialites may be able to get away with not doing it, but it vastly improves recordings.

Why? It reduces rambling, keeps your narration focused and clear. And practicing the entire video, as a dry run, often helps me find problems. I find so many issues in my own add-ons or tutorials just from rehearsing first.

Be okay with knowing it may take multiple tries. Ducky3D, a fellow Blender tutorial creator, has said in an interview that he rehearses three times. I often take more than four.

The number is not the magic part. The magic is being comfortable with the material and what you’re saying.

Rehearsing lets you relax because you know the material. You can be yourself. Making audio and videos is not about sharing information. It’s sharing you. Your personality. How you teach. How you inform and entertain.

Consistent Volume

Some people are very dynamic in their volume. They yell and shout, then go down to near-whisper. Some people speak dead-pan. The former is good for content. The latter is, frankly, much easier to edit. But you can edit so even the former sounds good.

Once you decide the distance between your mouth and the microphone, hold that distance. Little deviations are fine. Just know that, shifting your head a few inches while speaking loudly to a microphone farther away, is much more forgiving than speaking quietly to a mic only a few inches from your face.

Fix it in Post

Ideally, the previous sections mitigate editing audio to a bare minimum. But there will always be moments of, “Dang it, the mic picked up the dog barking outside.” Being comfortable with fixing up your own audio is important.

The good news is most things can be fixed in post. But my hope is that you won’t want to, except for the bare minimum of edits and filters, which I’ll share here.

Being Comfortable with Your Voice

In the editing chair, hearing your own voice feels different. I get why many creators hire someone else to edit themselves. For the first years, it is really hard to listen to myself. It took longer to get used to seeing my face on camera.

Some tips below, like EQ, will make your voice sound more natural to you. Honestly, the best solution is exposure therapy. Keep recording and editing. Resist the urge upon hearing your voice to throw the whole project in the trash. Just let yourself sit with it for a while, and post it. Slowly, but surely, you’ll get used to it.

I promise. That’s how it was for me.

P.S. if you have an accent, don’t worry about it. Just going a little slower and annunciating will be enough for most. And you’d be surprised by the number of viewers who will enjoy your narrations because of your accent.

Effects

Now let’s get into audio filters and effects. When editing audio, test long segments of you talking various sounds (s’s, ch’s, vowels, etc.) as well as a segment of expected “silence.”

The best part is you can save all your settings for future reference! Davinci Fairlight has preset options for tracks or buses. Super useful when making lots of separate projects recorded in the same spot.

Noise Gate > Denoising

Sometimes denoising is necessary, depending on your recording setup. I personally avoid it. Aggressive denoising can make edit your voice in a way that sounds strange. That’s why reducing noise before you record is always better.

Instead, I use a noise gate. This is best for quieter noise you just can’t get rid of. Noise will still be heard while you’re speaking, or as other loud noises come through. But it enables your audio to have actual silence between words. And most noise while you’re speaking won’t be nearly as discernible.

Equalizer

The equalizer (EQ) is your friend! Use it to tweak the sound of your voice, or reduce annoying frequencies. A lot of this comes from personal experimentation based on your voice and room, so take time to tweak and look up online guides for further reference. Here’s what I often do:

Cut out lowest frequencies (called a “high pass”, as in letting higher frequencies pass through). This often reduces many external noises. For me, it’s <100hz, but others may be able to go higher, if your voice is higher-pitched. Just avoid cutting out your own voice.
Cut out high frequencies (a “low pass”), for me anything >18-20k. Most super-high frequencies come from electronics or musical instruments. For just your voice, you won’t need it.
Raise the near-highs (somewhere in 10k-20k) slightly if you want a lighter sound to your voice. Davinci Resolve’s default EQ presets for voices are often a good place to start, but always tweak to your voice and room.

While adjusting EQ, you may notice a few resonant frequencies or hums that get through. This is how I catch them:

Add a really really sharp peak (set the Q factor to like 6 or something), set its volume really high
Slowly move it across the EQ spectrum during samples of your voice. Find spots that get suddenly harsh, loud, and weird.
Lower that spot to negative decibels so it gets quieter. Just -2 to -4 decibels is enough. Adjust Q factor if needed, but keeping it sharp ensures the rest of your voice stays the same.

De-Esser

This reduces harsh “S” or “T” sounds when you speak. Technically, it’s just a very specific equalizer. You can limit most harsh sounds like this by not talking directly into your mic. But I sometimes tilt my head just enough to speak into the mic by accident. This will help mitigate that.

Compressor

This squishes the volume of your audio to ensure the quiet and loud parts of your narration are less quiet and loud, keeping the volume more uniform.

Add a compressor to the bus. Adjust the threshold so it catches only the louder parts of your audio. Add gain to compensate for the quieter audio. Now your voice is in a compressed range of volume, and louder.

Limiter

This mostly ensures you don’t blow out your viewers’ eardrums by accident with any sudden loud noises or when speaking too loud. I usually set the max volume to around -1 or -2.

Export

Near the end of my editing, I go over the entire video, raising or lowering volume on individual clips especially loud or quiet. Mostly because I am guilty of muttering and trailing off (like I said, project your voice!).

Once mixed, adjust the bus volume or use the Davinci’s audio export settings to normalize the audio, so it’s fit for the platform it’ll be on. YouTube preset recommends -14 LUFS, but for narration I often go with -16 or -17. LUFS is a measurement of loudness over time, used measure loudness more perceptually.

And that’s it! If any experts reading this note anything I missed, message me on the Fediverse (or wherever you’ve seen me online) and let me know. I’ll consider this a living document and update it as needed.

Beyond the Horizon