Clean a Loom Recording: Sharper Voice for Async Video

Loom audio problems start at capture

Loom recordings are short, informal, and watched asynchronously, which makes their audio surprisingly consequential: a viewer who is confused by muffled sound does not ask a question, they just watch something else. The good news is that most Loom audio problems are capture problems, and Loom's own support documentation organizes them the same way: missing audio, low volume, distorted or intermittent sound, and audio that alternates between clear and muffled.

The causes the documentation names are practical rather than exotic: microphone permissions in the browser or operating system, the wrong default input device, an input level set near zero, system-audio capture failures, and on Windows, an operating system enhancement that interferes with the microphone. Almost all of these are fixed before the recording button ever gets pressed. Loom's categories also tell you what each symptom usually means: missing audio is permissions or a system-audio toggle, low audio is an input-level setting, distorted audio points at conflicts or resources, and audio that alternates between clear and muffled often follows the microphone itself rather than the app.

That ordering matters. A cleanup pass can improve a recording that carries room noise, but it cannot add a microphone that was never selected or restore a signal that clipped at capture. The checklist below addresses the capture side; the cleanup path below that handles the files that already exist. None of it is unique to Loom; it is the standard web-capture stack, which is also why the fixes transfer cleanly between tools.

The capture checklist

Run this once per recording setup, and most Loom audio problems never occur:

1. Select the microphone explicitly in Loom's recording settings instead of trusting the system default, which may be a webcam or a disconnected headset. Permission problems look identical from inside Loom; the difference shows up in system settings.
2. Test the exact input with Loom's camera and microphone test page before recording anything worth keeping.
3. Check the input level in your operating system's sound settings and raise it if the meter barely moves; low volume is usually a capture-side setting, not a Loom bug. Write down the level that worked; it is the setting most likely to be reset by an update.
4. On Windows, check Sound settings for Voice Clarity or Voice Focus under the microphone's audio enhancements and disable them; Loom documents these OS features interfering with capture.
5. Decide whether system audio should be captured, and enable it deliberately for walkthroughs that include video or app sound.
6. Close or pause other apps that may claim the microphone; a conferencing app in the background can hold the device and quietly degrade the capture.

The Windows enhancement conflict

The Voice Clarity and Voice Focus settings deserve their own note because they are easy to miss and hard to diagnose. A Windows update can enable them silently, and from then on the operating system applies its own processing between the microphone and every application, including Loom. The sound can become thin, gated, or oddly processed, and the cause lives one layer below the recording tool.

The fix is a two-minute visit to Windows sound settings: select the active microphone, open its audio enhancements section, and disable the features. If the voice suddenly regains body in the next recording, the conflict was the whole story. Mac users with the same symptom have equivalent places to look in System Settings under Sound and Privacy, where input levels and app permissions behave much the same way.

Recording modes and their audio

Loom records in two environments, and the audio behaves differently in each. The desktop app runs with full system access: it can capture system audio directly, use the fallback recorder when GPU processing misbehaves, and reach microphones without browser mediation. The Chrome extension lives inside the browser's permission model, which is why its troubleshooting starts at the browser's settings and tab-audio checkboxes.

The choice matters for troubleshooting, not for quality: both capture the same microphone when configured the same way, and both produce MP4s that can be cleaned identically. When an extension recording gives trouble that a permission reset does not fix, Loom's own documentation suggests switching to the desktop app for that recording. Beyond that, the two modes converge.

Cleaning a Loom recording you already made

For an existing recording with a steady noise bed, the file-based path is short:

1. Download the Loom video. Keep the original file untouched and keep the Loom link as a backup reference; the link is the version of record until a cleaned file replaces it.
2. Upload one file to SimpleClean. Up to 500 MB and 15 minutes; the MP4 is supported directly, and the cleaned video keeps its picture and timing intact.
3. Compare Original with Cleaned across a pause, a sentence, and a loud passage, listening for both the noise floor and the voice.
4. Keep the better version for sharing or re-uploading; the steady bed of room tone, fan noise, or hiss is the target, not the keyboard between phrases.

When to re-record instead

Async videos are short, and that changes the economics of repair. A ten-minute walkthrough recorded with the wrong microphone takes longer to salvage than to re-record with the checklist done properly, and the second take usually lands better anyway because the presenter has rehearsed once. A second take also tends to be shorter and clearer, which pays off for the viewer.

The judgment line is simple: if the fix is settings, re-record; if the fix is a steady bed under otherwise good audio, clean it. A video that is already shared can still be replaced with a cleaned or re-recorded version while the same link keeps working. One more signal worth trusting: if watching your own recording makes you wince, your audience will too, and the wince is a reliable quality gate.

What good Loom audio sounds like

A useful bar to aim for: the viewer forgets the audio exists. Speech stays intelligible at low volume, the room does not announce itself between sentences, and nothing clicks, gates, or pumps. That bar is reachable with a headset and a quiet corner, and it is the standard that both the capture checklist and the cleanup pass serve.

Why audio quality decides completion

Async video has no presenter to smooth over a bad moment; the recording either holds the viewer or it does not. Audio carries most of that load, because comprehension and trust both degrade faster from poor sound than from poor picture. A screen recording with a clear voice and a plain image reads as professional; the same content with a hollow, hissy voice reads as unfinished, and viewers quietly stop watching.

That is the real reason to spend two minutes on the capture checklist and five on a cleanup comparison. The improvements are small in production terms and large in how the work is received, which is exactly the trade that makes them worth doing. It is also the cheapest kind of production value: no lighting, no set, no editing, just a microphone that is set up properly and a file that sounds like someone cared.

Related guides

The most common intruder in screen recordings has its own guide in removing keyboard noise from audio, and the meeting-recording workflow in cleaning remote interview and Zoom recordings applies to any captured call. For the general video round trip, see removing background noise from a video recording. The capture checklist above is worth pinning somewhere visible for the next recording; the cost of skipping it shows up in every file that follows.