Voice Feedback

Enable users to submit audio notes alongside their bug reports. Understand browser audio constraints, recording limits, and privacy settings.

Last updated:2026-07-11

TellBack allows users to record and submit voice notes directly from the feedback widget, providing verbal explanations of bugs to complement screenshots and text comments.

---

Recording Limits & Controls

When a user clicks the microphone button to start recording:

• 60-Second Cap: Recording is limited to a maximum duration of 60 seconds. A client timer automatically stops and saves the recording if the user reaches this limit.
• Audio Controls: Users can pause, resume, or discard their voice note prior to submission.
• Verification: An on-screen progress indicator displays the elapsed recording time.

---

Audio Mime-Types & Compatibility

Audio recording relies on the browser's native MediaRecorder APIs. To ensure compatibility across different operating systems and browsers, the widget checks for supported mime-types in the following priority order:

1. audio/webm (Standard standard container format)
2. audio/mp4 (Supported natively on Safari/iOS browsers)
3. audio/x-m4a (Alternative Apple container format)

If none of these specific mime-types are supported, the widget falls back to the browser's default constructor parameters.

---

Microphone Block & Permission Denied Fallback

To record voice notes, the browser must have microphone access permissions.

• Access Denied: If the user blocks microphone permissions or the browser fails to access the input device, the script catches the error, displays a client alert stating "Could not access microphone.", resets the recording state to idle, and performs cleanup.
• Empty Audio Guard: If the resulting audio file size is zero bytes (indicating a hardware encoding failure), TellBack discards the empty recording, resets the controls, and prompts the user to record again.

---

Voice Transcription & AI Tasks

AI Setting Configuration

Project AI settings control whether audio/transcripts, screenshots, and session context are included in AI task generation where supported by the current product configuration. If AI settings are disabled for the project, submitted voice notes will remain available for manual playback in the dashboard but will not be read during task generation.

---