VMA Video Analyser with AI support

VMA Video Analyser now includes an optional AI-assisted image assessment to support broadcast operators during live monitoring.

The AI does not replace scopes, alarms, or engineering judgement. It provides a short, human-readable verdict that complements the existing technical measurements already available in VMA.

What the AI Provides

  • A clear broadcast safety verdict
  • The main visible image issues
  • Practical adjustment suggestions using standard controls (black level, gamma, saturation, knee)
  • A brief explanation of why the issue matters on air

The output is intentionally concise and action-oriented, designed to be read and acted upon within seconds.

Local AI — No Cloud

All AI processing runs on a local AI engine (for example LM Studio / LM Tools), either on the same machine or the local network.
No images or data are sent to external cloud services.

This ensures:

  • Data privacy
  • Offline operation
  • Predictable latency
  • Full control over the AI model in use

Operator Always in Control

The AI provides a second opinion only.
Scopes and alarms remain the primary tools, and the final decision always stays with the operator.

The goal is not automation, but assisted decision-making in real broadcast workflows.

Requirements

This feature requires a local AI engine such as LM Studio and a compatible vision-language model.
Both the AI engine and suitable models can be downloaded and used free of charge.