- Set your microphone so speech peaks around minus 12 to minus 6 dBFS, leaving headroom, and record ten seconds of room silence at the start of every session for later noise cleanup.
- Use a microphone close to your mouth and a treated space, and avoid built-in laptop mics and Bluetooth headsets for narration, because both lower the quality you can recover later.
- Wear headphones whenever tab or system audio is captured, so the microphone doesn't re-record the playback as an echo.
Viewers forgive a soft picture far more readily than they forgive bad sound. A tutorial that’s slightly blurry but has clear, close narration feels professional. A sharp, beautifully zoomed recording with a hollow, echoing voice feels like something to close quickly. Recording audio quality is the part of a screen recording that people judge in the first ten seconds, often without noticing they’re judging it.
Most audio problems in screen recordings are not expensive to prevent and are very hard to fix afterward. A little hiss can be reduced. A room that rings cannot be removed cleanly. A voice that clipped at the peak has lost information that no editor can restore. This guide covers the chain that produces the sound you hear, the decisions that matter most at each stage, and the checks that take a few minutes before every session.
Follow the signal chain behind recording audio quality
Every sound in your recording passes through a chain, and each link can add a problem. Your voice reaches the microphone, which turns sound into a small electrical signal. The signal goes through your computer’s input settings, then into the browser’s capture of the microphone, then gets combined with any tab or system audio, and finally gets encoded into the file. A problem at any link is carried through to the end.
The useful way to think about the chain is to ask, at each step, what can go wrong and where it’s cheapest to fix. Distance and microphone choice fix problems at the source, before anything is recorded. Levels fix problems at the input, when the signal is being set. Room treatment fixes problems in the air, before the mic picks them up. Mixing and encoding are the last place to fix anything, and the least forgiving.
Set levels so you have room to work
Levels are where most beginners lose quality without realizing it. Your input level, the setting that controls how loud the microphone signal is before it’s recorded, has to sit in a narrow band. Too quiet, and you’ll have to boost the recording later, which raises the hiss and background noise along with your voice. Too loud, and the peaks clip. Clipping is digital distortion that happens when the signal exceeds the maximum the system can represent, and it can’t be undone.
Aim for speech peaks that sit around minus 12 to minus 6 dBFS, which is the scale digital audio uses, where zero is the absolute ceiling. That leaves headroom for the occasional louder word, a laugh or a sudden sound without clipping. Speak at your normal narration volume, watch the meter, and adjust until the loudest words land in that range. Most recorders show the meter, and most operating systems show it in the input settings.
Then check the level once more before you start, because input levels drift. Switching between a headset and a laptop, or plugging in a new device, can change the level without any warning. A thirty-second check at the start of every session catches it.
Pick a microphone for narration
The microphone decides more of the final sound than anything else in the chain. The main choices are the built-in laptop microphone, a headset, a USB microphone and a dedicated microphone with an audio interface. Each has a typical result, and each has a fix.
| Microphone | Typical result for narration | What to do about it |
|---|---|---|
| Built-in laptop mic | Thin voice, room sound, keyboard and fan noise | Move closer, or use a headset or USB mic |
| Wired headset mic | Consistent level, close to the mouth, but sometimes bright | Keep it at a steady distance and check the level |
| Bluetooth headset mic | Narrow, compressed voice that sounds distant | Switch to a wired or USB mic for recording |
| USB condenser or dynamic mic | Full, clear voice when placed close, good for most tutorials | Use a pop filter and keep a steady distance |
| Audio interface with a dedicated mic | Best control and the most headroom for later cleanup | Set gain carefully and monitor on headphones |
For most screen recordings, a USB microphone placed about fifteen to twenty centimeters from your mouth, slightly off-axis to avoid plosives, is a practical choice that doesn’t require any audio expertise. The distance matters more than the price. A modest microphone close to the mouth beats an expensive one across the desk.
The built-in laptop microphone is the most common source of thin, echoey narration. It sits far from your mouth, picks up the room, and often captures the fan and the keyboard. If you have nothing else, record with it close to your face and at a low level, and consider the headset option as the first upgrade.
Bluetooth headsets switch to a lower-quality mode: when a Bluetooth headset's microphone is active, the connection typically drops to a narrower, lower-fidelity voice profile, so your narration will sound distant and compressed even if the headset sounds fine when you listen. Record with a wired headset or a USB microphone instead.
Control the room
Rooms add sound. Hard surfaces, such as bare walls, glass, tile and a hard desk, reflect your voice back into the microphone a few milliseconds after the direct sound. The result is a hollow, ringing quality that sounds like you’re in a stairwell. You can’t remove that reflection later without damaging the voice.
Soft surfaces absorb reflections. A rug, curtains, a bookshelf full of books, upholstered furniture, and a closet full of coats are all useful, and the closet is a classic for a reason: the clothes absorb sound from every direction. You don’t need a studio. You need to reduce the number of hard surfaces the microphone can hear, and to keep the microphone close enough that the direct sound of your voice is much louder than the reflections.
Background noise is the other room problem. Fans, air conditioners, traffic and the neighbors are all easy to ignore while you’re speaking and hard to edit out later. Before each session, record ten seconds of silence with your mouth closed and the room as it sounds during the take. That room tone is the reference that noise reduction tools use to identify what to remove, and having it lets you clean the file without guessing.
Record ten seconds of room tone first: silence with the same room and the same equipment gives you a clean noise profile for later. It costs you a few seconds at the start of each session and saves you from guessing during cleanup.
Mix tab audio and narration without an echo
When your recording includes sound from the screen, such as a demo video or an app’s notifications, you have two sources to balance. The narration should sit clearly above any tab audio. Background music or a clip’s sound should be quieter than your voice, usually well below it, so the viewer can follow the narration without straining. Mixing too loud is the most common error, and it makes the narration hard to hear through the music.
The echo problem is more specific. If tab audio plays through your speakers while your microphone is open, the microphone captures the playback and records it again, slightly delayed. Headphones solve the problem completely, because the playback never reaches the microphone. If headphones aren’t an option, lower the speaker volume and keep the microphone close to your mouth, or record the narration separately and mix it in afterward.
Set loudness so viewers don’t reach for the volume
Loudness is the overall level of the finished file, and it’s measured in LUFS, which stands for loudness units relative to full scale. Platforms and players normalize audio to a target, and a file that’s far from the target will sound too quiet or too loud compared with everything else the viewer watches. Common targets for spoken content sit around minus 16 LUFS, and some platforms normalize around minus 14 LUFS. You don’t need to hit either exactly. The point is to aim for a consistent level across your videos, so a viewer doesn’t reach for the volume between two of them.
If your recorder or editor reports loudness, use it on the finished file and adjust until it sits near the target. If it doesn’t, listen on a laptop and a phone at normal volume, and match the level to your other videos. Consistency across a series matters more than the exact number.
Keep the file from degrading
The final stage is encoding, which compresses the audio so it fits the file. Encoding is usually handled by the recorder, and the settings you can change are the bitrate and the sample rate. A sample rate of 48 kilohertz is the common standard for video, and it’s usually the right choice. A very low bitrate compresses voices into a thin, watery quality, so keep the audio bitrate high enough that speech sounds natural.
Be careful with automatic noise suppression. A little suppression can remove a fan or a keyboard. Too much makes the voice sound processed and gives it a watery, artifact-filled quality, especially on sibilant sounds like “s” and “sh.” If you use suppression, apply it lightly, listen back at normal speed, and turn it down if the voice starts to sound strange.
When the audio is already recorded
Sometimes the recording is finished and the sound isn’t good enough. The narration clipped, the room rang, or the microphone sat across the desk. Editing can recover some of it, but it’s worth knowing what it can and can’t do before you start, so you don’t spend an hour making a file worse.
Noise reduction works best when it has a clean sample of the noise to learn from. That’s what the ten seconds of room tone is for. Without it, the tool has to guess what’s background and what’s voice, and the usual result is a voice that sounds underwater. Apply reduction in small steps, a few decibels at a time, and listen at normal speed between each step. Stop when the hiss is no longer distracting, not when it’s gone, because chasing the last trace of noise is where voices start to sound artificial.
Equalization can fix a voice that’s boomy or thin. A gentle cut in the low hundreds of hertz reduces the boxiness of a small room, and a small lift between roughly two and five kilohertz can help narration cut through a busy soundtrack. Make these changes in small amounts. Heavy equalization on a bad recording rarely produces a good one.
Clipping is the one thing you can’t fix. A clipped peak has lost the original waveform at that moment, and no tool can restore what’s missing. The practical response is to re-record the affected line, which is faster than any repair. If the clipping is brief and confined to a few words, a careful edit that replaces those words may be enough. If it runs through the whole take, re-record the section.
Keep the original file. Editing a copy lets you compare the result with the source, and it lets you start again when the first pass goes wrong, which it often does.
A short checklist before each session
Audio problems are cheap to prevent and expensive to fix, so a few checks before each session pay off across the whole series.
- Choose the microphone and place it. Wired headset or USB mic, about fifteen to twenty centimeters from your mouth, off-axis to reduce plosives.
- Set the input level. Speak at your normal narration volume and aim for peaks between minus 12 and minus 6 dBFS.
- Record ten seconds of room tone. Stay silent, let the room sound as it will during the take, and keep the clip for cleanup.
- Put on headphones if tab audio is captured. This prevents the echo loop completely.
- Play back a thirty-second test. Listen at normal volume on the device your viewers are most likely to use, and check for clipping, echo and background noise.
Once those checks are in place, the rest of the session is about your delivery, which is a separate skill. Clean audio gives the viewer no reason to notice the sound at all, and that’s the goal.
The short version
Clean screen recording audio comes from a short chain of good decisions. Choose a close microphone, ideally a USB or wired headset rather than a laptop mic or a Bluetooth headset. Set levels so speech peaks sit between minus 12 and minus 6 dBFS, and leave headroom. Treat the room with soft surfaces and record ten seconds of room tone. Use headphones whenever tab audio is captured, mix narration above background sound, and aim for a consistent loudness across the series.
Most of this happens before the recording starts, and it’s the part a browser recorder can’t do for you. RecordlyPro records in your browser with nothing to install, and adds auto-zoom, cursor effects, captions and chapters, which means the audio is the part of the process you control most. For the rest of the setup that shapes a clean recording, the screen recording tips for a clean result cover the visual side of the same job.
Keep reading
Async video: stop scheduling the meeting you could record
A hundred random links can do less than five relevant ones. Here's what makes a backlink pass real ranking value in 2026 — and what makes it worthless or worse.
Sofia Meyer
October 6, 20262 min readAuto-zoom in screen recordings: what it is and why your demos need it
A hundred random links can do less than five relevant ones. Here's what makes a backlink pass real ranking value in 2026 — and what makes it worthless or worse.
Sofia Meyer
October 13, 20262 min readHow to choose a screen recorder for product demos in 2026
A hundred random links can do less than five relevant ones. Here's what makes a backlink pass real ranking value in 2026 — and what makes it worthless or worse.
Sofia Meyer
October 16, 20262 min readDofollow vs nofollow in 2026: what still passes value
The link attributes changed, Google's treatment changed, and a lot of advice is out of date. Here's what dofollow, nofollow, sponsored and ugc actually do now.
The Recordly Team
October 10, 20262 min read