[XB] Voice Messages: Audio Recording, Voice Notes & AI Transcription 1.0.0

Record, upload, transcribe, and share short voice messages directly inside XenForo content.

This add-on provides native voice messaging for posts, direct messages, and profile posts with recording controls, audio uploads, waveform playback, storage quotas, optional AI transcription, and detailed administrator analytics.

---

Features:​

• In-Browser Voice Recording: Record voice messages directly from supported content creation forms.
• Audio File Uploads: Allow selected user groups to upload existing audio files as voice messages through a separate permission.
• Pause & Resume Recording: Pause and continue the same recording without counting paused time toward the duration limit.
• Submit-Time Upload: Keep recordings local until content is submitted, with an optional Add button for uploading earlier. Failed uploads remain available locally for retry until the page is reloaded.
• Recording Preview: Listen to or remove a recording before submitting content.
• Integrated Audio Player: Play voice messages using a compact player with playback controls, duration, and waveform preview.
• Single Active Player: Starting another voice message automatically pauses the currently playing recording or local preview.
• Remembered Playback Speed: Preserve the preferred playback speed across voice messages, page reloads, and other open tabs.
• Waveform Preview: Display a visual waveform for recorded and uploaded voice messages.
• Seekable Playback: Seek through voice messages during playback.
• Optional AI Transcription: Allow authors to request transcripts or automatically transcribe newly published recordings using OpenAI or an OpenAI-compatible self-hosted Whisper service.
• Automatic Transcription: Optionally transcribe recordings automatically when they become visible, including content published after moderation approval.
• Granular Permissions: Control recording, audio uploads, transcription requests, deletion, duration, count, storage, and file-size limits.
• Recording Duration Limits: Configure maximum voice message duration globally or per user group.
• Message Count Limits: Limit the number of voice messages that can be attached to a single message.
• File Size Limits: Configure global and permission-based maximum audio file sizes.
• Storage Quotas: Set per-user-group storage limits and display remaining storage directly in the recorder interface.
• Per-Content Controls: Enable or disable voice messages independently for supported content types.
• Automatic Cleanup: Remove abandoned recordings and clean up detached audio files and transcripts after permanent content deletion.
• Optional FFprobe Validation: Enable enhanced server-side validation for reliable audio duration checking.
• Extensible Content Support: Designed to support additional XenForo content types through integrations.
• ACP Analytics: Review upload activity, storage usage, playback totals, content-type distribution, and the most active voice message authors.

---

Supported Content Types:​

• Forum Posts
• Direct Message Replies
• Profile Posts

---

Permissions:​

• Record Voice Messages
• Upload Audio as Voice Messages
• Request AI Transcription
• Maximum Voice Message Duration
• Maximum Voice Messages per Message
• Maximum Storage Space
• Maximum Voice Message File Size
• Delete Own Voice Messages
• Delete Any Voice Message

---

Requirements:​

• HTTPS: Required for browser microphone access.
• Microphone Access: Members need an available microphone and browser permission to use it.
• Modern Browser: Requires support for browser-based audio recording.
• FFprobe - Optional: Recommended when reliable server-side enforcement of recording duration limits is required.
• AI Transcription - Optional: Requires OpenAI or an OpenAI-compatible Whisper transcription service when transcription is enabled.

---

FAQ:​

• Can someone access a voice message without permission to view its post or direct message?
  No. Audio is served through a permission-checked route rather than exposed as a public file. A listener must be allowed to view both voice messages and the XenForo content containing that recording.
• Is audio uploaded as soon as a member starts recording?
  No. A new recording remains in the member's browser until the content is submitted, unless the administrator enables the optional Add button and the member uses it. An uploaded but unpublished recording is private to its owner and is removed after the configured cleanup period.
• Can voice messages be added while editing existing content?
  Yes. Members can add recordings while editing a forum post, direct message, or profile post when XenForo allows them to edit that content. Existing recordings remain attached unless the member removes one and has the required deletion permission.
• Who can request and read an AI transcript?
  Only the recording's author can request transcription, and only with the transcription permission. Once a transcript is available, it follows the same visibility rules as the voice message, so it does not make private or restricted content public.
• Will enabling automatic transcription also process private direct messages?
  Not by default. Direct-message transcription has a separate administrator option. When enabled, the audio is sent to the configured OpenAI or OpenAI-compatible provider only after the message is published and visible.
• What happens if a user reaches or exceeds their storage quota?
  Existing voice messages remain available, but the member cannot upload another recording until enough quota is available. Deleting one of their recordings releases its storage usage, and administrators can rebuild all quota totals from the ACP if files or records have been migrated.
• What happens to audio when its XenForo content is permanently deleted?
  The recording is detached and queued for cleanup together with its transcript. The scheduled cleanup removes the stored file and updates the author's quota; temporary uploads that were never published are also pruned after the configured retention period.
• What changes when FFprobe validation is enabled?
  FFprobe verifies that an upload contains audio, matches its declared container, and stays within the duration limit using server-inspected metadata. If this option is enabled but the ffprobe executable is unavailable, uploads are rejected until FFprobe is installed or the option is disabled.
• Where are voice-message files stored, and what must be included in backups?
  Files are kept in XenForo's internal data storage, while message metadata, associations, quotas, play counts, and transcripts are stored in the database. A complete backup or migration must therefore include both the database and XenForo's internal data directory or configured internal-data storage.

---

Useful for:​

• Community Discussions: Let members respond with short spoken messages instead of typing.
• Mobile Communities: Make quick replies easier on mobile devices.
• Voice-Focused Communities: Add native voice interaction without relying on external chat platforms.
• Alternative Communication: Give members another way to participate in discussions.
• Private Conversations: Exchange voice messages through direct messages.
• Searchable Spoken Content: Add transcripts to voice messages when AI transcription is enabled.

---

Screenshots:​

Recording and Playback​

Administration and Analytics​

Configuration​