LuxASR dictation interface

Alternative interface for the speech-to-text transcription service LuxASR by the University of Luxembourg.

About this dictation user interface

This alternative interface is only available in english, uses the official API of LuxASR and is designed specifically for continuous dictation workflows.

Features and limitations

ℹ️ What this page does:

Continuous dictation is better suited for long documents. Closer to the dictation feature offered on smartphones.
Transcriptions insert at your cursor position.
Automatically transcribes after periods of silence. Duration and volume treshhold can be configured.

⚠️ What this page doesn’t support:

Multiple input languages: Only Luxembourgish is supported.
File upload: Only live microphone recording (use the official web interface to transcribe audio files)
Advanced output formats: Only plain text output (use the official web interface for SRT subtitles, TextGrid, or JSON with timestamps)
BAS TextGrid integration: Not available on this page

Ready to start recording

Recording Options

Speaker Diarization: Separate text by speaker

Output Format: Format of transcription

Silence Threshold (0-100): Audio level below this is silence

Auto-transcribe after (seconds): Silence duration before auto-transcribe (0 to disable)

Audio Level 0%

0.0 seconds

Press Esc to cancel

How to use:

Start Recording: Begin capturing audio from your microphone
Pause: Temporarily stop recording without transcribing (press Start to resume)
Transcribe & Continue: Transcribe audio recorded so far, insert at cursor position, and continue recording seamlessly
Finish: Stop recording completely and transcribe all remaining audio
Auto-transcribe: When enabled (duration > 0), recording will automatically finish and transcribe after the specified silence duration

Keyboard shortcuts:

Tab: Navigate between controls
Space/Enter: Activate buttons
Escape: Cancel auto-transcribe timer