Product / Audio enhancement

Make your point.
Less noise in the way.

When a noisy microphone competes with your explanation, people miss the detail. Clean up the audio on your Mac after recording, so teammates can focus on what you are showing.

See how it works
01 / 03

They came to hear your point.

A fan, a busy room or an everyday microphone can get between your explanation and the person watching it.

Keep the message in the foreground.
02 / 03

Make a noisy room less of the story.

On-device enhancement reduces background noise as part of the native recording workflow. Choose the cleanup that suits your microphone and voice.

Less noise competing for attention.
03 / 03

Let your ears decide.

Compare the same sample before and after the recorder’s actual enhancement pipeline. Listen to the pauses and the character of the voice.

Real processing. An audible comparison.
YOUR EXPLANATIONOn-device enhancement
“Here’s the issue.
Let me show you.”
Room noiseVoice, brought forward
LESS DISTRACTION

Keep attention
on the explanation.

Hear the actual before-and-after sample

Illustrative waveform · real audio comparison follows
Scroll to follow the storyRecreated product scenes · synthetic data
Listen to the difference

Same explanation.
Less in its way.

A synthetic voice with added fan-like noise, processed through Thrume’s actual on-device enhancement pipeline.

A 13-SECOND BUG REPORT
“The button stops, but the invitation never arrives. I have marked this moment for the developer.”

Try headphones. Listen to the room noise during pauses, then to the clarity and character of the voice.

DeepFilterNet3 · Full cleanup
Same source and playback gain
A

Original

Synthetic speech + added noise

B

Enhanced on device

Processed by the native recorder’s engine

Press play on either version to compare.

Transcript & how this sample was made

“Here is the issue. I open the team page, enter an email address, and send the invitation. The button stops, but the invitation never arrives. I have marked this moment for the developer.”

Created locally with macOS synthetic speech and deterministic artificial noise. Both versions use the same input gain and MP3 encoding; the enhanced version has no separate loudness boost. Cleanup uses the production AudioEnhancement.process method, DeepFilterNet3 and Full strength. This is not a Voice+ demonstration.

Results vary by microphone, voice and environment. Listen for changes to voice character as well as noise reduction. View sample provenance.

Designed around the work

Get your voice ready to share.

Read the setup guide
01

Keep attention on what you are saying.

Enable voice enhancement to clean up microphone audio after recording. Processing runs on your Mac before upload, with a cleanup strength you choose for your voice and room.

02

Find the right cleanup for your setup.

Try the available strengths and listen for both background noise and voice character. Voice+ adds another on-device option through a model download in recorder settings. A short sound check helps you choose the result you prefer.

03

Know which microphone you are recording.

Set your microphone priority or pin a specific input, then use the sound check before an important take. Keep the original audio available to retry enhancement or use the unprocessed version.

04

Share a clearer take. Choose AI separately.

Enhanced recordings upload to Thrume for sharing. Optional cloud transcription and summaries have their own workspace settings; they are separate from the voice cleanup that runs on your Mac.

Public beta · Approval required

Make the explanation
easier to hear.

Request beta access

Native Mac recording.
Shared understanding, everywhere.