Recording & Transcription

Everything about how AsPeach records your voice and turns it into text.

Starting & Stopping a Recording

There are two ways to start recording:

Keyboard shortcut — Press your configured shortcut (default: Alt+Shift+Space). Press it again to stop. This works from anywhere, even when AsPeach is in the background.
Click the peach button — Click the big circular button in the AsPeach window. Click again to stop.

How Do I Know It's Recording?

The peach button turns pink and starts pulsing with a glow effect.
A timer appears showing minutes and seconds.
The status text changes to "Recording..."
If sound effects are enabled, you'll hear a short tone when recording starts.

Recording Duration Limits

Free tier: Maximum 60 seconds per recording.
Pro tier: Maximum 30 minutes per recording.

When you hit the limit, recording stops automatically and transcription begins.

Pause & Resume

While recording, you can press the Pause button to temporarily stop without ending the recording session. The timer pauses too. Press Resume to continue.

When is this useful? If someone interrupts you, if you need a moment to collect your thoughts, or if there's a loud noise you don't want captured.

Voice Activity Detection (VAD) — Auto-Stop

VAD makes AsPeach automatically stop recording when you stop talking. You set a silence timeout — for example, 3 seconds — and if AsPeach hears silence for that long, it stops recording and starts transcribing.

How to Set It Up

Go to Settings → General.
Find the VAD Timeout slider.
Set it to a number between 1 and 10 seconds. This is how long AsPeach waits in silence before auto-stopping.
Set it to 0 to disable auto-stop (you'll always need to manually press the shortcut to stop).

Visual countdown: When VAD is about to trigger, you'll see a countdown indicator on screen, so you know auto-stop is coming and can keep talking if you want to cancel it.

Audio Quality Indicators

AsPeach gives you real-time feedback about your audio quality:

Audio Level Dot

A small colored dot visible during recording:

Gray — No audio detected. Your mic may be muted or not working.
Green — Good audio level. This is what you want.
Yellow — Audio is getting loud. You might be too close to the mic.
Red — Audio is clipping (too loud). Move back from the mic or lower your volume.

Confidence Badge

After transcription completes, a badge may appear showing an estimated quality percentage:

High (green) — Clear audio, transcription should be accurate.
Medium (yellow) — Decent audio, may have some inaccuracies.
Low (red) — Noisy or quiet audio, transcription may have errors.

This is based on audio characteristics (volume, clipping, duration) — not a guarantee of transcription accuracy, but a helpful indicator.

Noise Gate

The noise gate is an optional setting that shows a visual indicator when your audio is below a certain threshold. It does not cut off audio — it just tells you that the sound is too quiet. Set it in Settings → General → Noise Gate (value 5–50, or 0 to disable).

Dead Mic Detection

If AsPeach starts recording but detects zero audio input for 1.5 seconds, it shows a warning that your microphone might not be working. This catches situations where:

You selected the wrong microphone.
Your mic is muted in system settings.
Your mic got disconnected.

Chunk-Based Processing

For longer recordings, AsPeach splits your audio into 45-second chunks and processes them one at a time. This means:

Progressive results: You start seeing partial transcriptions while you're still talking. You don't have to wait until the very end.
Better performance: Processing smaller chunks is faster and uses less memory.
No lost audio: All chunks are stitched together at the end into one complete transcription.

What Happens After Recording Stops

The status changes to "Transcribing..." and the button shows a spinner.
Whisper AI processes your audio locally on your device. Nothing is sent to the internet.
Text processing runs: filler word removal, grammar fixes, dictionary corrections, snippet expansion (if enabled).
If any AI transform was detected in your speech (like "make it formal"), that runs next.
The final text is either auto-pasted into your active app, or shown in the preview panel (depending on your settings).

Sound Effects

By default, AsPeach plays short audio tones when recording starts and stops. This gives you audio confirmation without needing to look at the screen — helpful when you're focused on another app.

Toggle this in Settings → General → Sound Effects.

Waveform Visualization

During recording, a live waveform animation shows your audio input in real-time. This is purely visual — it helps you confirm that audio is being captured and gives a sense of how loud you're speaking.