VALT 6.7.0-1

August 14th, 2026]

New Application Features

• Speech-to-Text Transcription: added the ability to automatically transcribe recordings using an on-premise, GPU-accelerated transcription service, in addition to the existing AWS option
  • Added an "Auto Transcribe" group right; recordings by users in an enabled group are queued for transcription automatically when a recording finishes
  • Transcripts generate caption (VTT), plain-text, and word-searchable files, and can also be run on demand from the Review page
• Added new report types
  • Duration Marker report: lists duration-marker comments with their measured durations and a summary graph
  • Camera Assignment report (admin): visualizes the flow of Media Server to Camera to Rooms/Views
  • Cronbach's Alpha report: computes per-field reliability scores across evaluations
• Added freehand drawing comments: users can draw annotations directly onto a recording during review
  • Drawings scale with and stay attached to the correct position on the video
  • Includes freehand, line, and shape tools with a dedicated drawing toolbar
  • Drawings can be edited after saving and created alongside a file attachment
• Added a Multipoint question field to evaluation templates, with configurable rows, columns, and score values
• Added 12-cell (4x3 and 3x4) and 16-cell (4x4) multiview grid layouts on the Observe page and in recording playback
• Added per-camera motion detection, allowing motion detection to be enabled or disabled per camera within a room
• Added the ability to choose between Compass (wheel) and Joystick PTZ control modes as a per-user preference
• Added support for BEAM devices in saved Views on the Observe page
• Added bulk recording management: multi-select on the Review page to share or change the author of multiple recordings at once
• Raised the maximum number of devices per room to 16
• Redesigned the Home page Reports widget with an updated layout, support for comment and recording-time reports, an edit menu, and automatic refresh after edits
• Added sharing-link controls for guest commenting on the Observe page
• Added the ability to download multiple camera views as a single zip file
• Added a configurable camera offline/online alert delay
• Added a default schedule duration
• Increased character limits on template text fields

Application

• Fixed an issue where the Stop/Edit recording buttons could disappear on the Observe page
• Fixed guest access via shared links so guests can add comments and markers without receiving an access error
• Resolved multiple evaluation and information template issues, including multipoint field creation and validation, default template selection, field reordering, and score-field input restrictions
• Corrected scheduling behavior, including overlap warnings, occurrence consumption on failed starts, blocking room changes on in-progress schedules, and audit logging of schedule edits
• Fixed playlist behavior, including author and delete permissions, share notifications, and autoplay, mute, and zoom state carrying between recordings
• Addressed ROAM upload playback and visibility, including playback on macOS and correct status updates for parallel uploads
• Numerous iPad and tablet fixes for streaming, playback, and control sizing

Security

• Fixed an issue where comments and related data on a shared recording could be viewed or added by an anonymous guest when the sharing link did not permit it
• Enforced evaluation edit permissions so only the authorized author can modify an evaluation
• Hid evaluation scores from users without the view-scores permission

Accessibility

• Completed a WCAG accessibility audit across the application, adding keyboard operability for custom interactive controls, accessible labels and corrected ARIA attributes on form controls, discernible text for icon-only buttons and links, and valid list semantics for menus and lists

Technical

• Added a local Speech-to-Text service container and a serialized heavy-job queue that processes GPU transcription jobs one at a time
• Added Wowza GPU detection to report transcription capability per media server
• Set the default stream type to WebRTC for new servers
• Added new deployment container images (Speech-to-Text, Redis init, RabbitMQ init) to the build pipeline
• Added a container to manage Axis HTTPS