Overview

The HyperFlow Media Transform service provides comprehensive image and video processing capabilities through both simple UI controls and advanced JSON DSL scripting. Built on PIL (Pillow) for images and FFmpeg for videos, it supports sophisticated workflows including batch processing, scene analysis, and Key-Value Store integration for persistent media workflows.

See the HyperFlow Media Transform Service - Users Guide for detailed specs and the Media Transform Operations Guide for details and examples of the operations supported by this service.

Operation Modes

Simple Mode (UI-Driven)

Perfect for single operations with visual parameter configuration. Select an operation type in the HyperFlow interface to expose relevant UI controls (sliders, dropdowns, text inputs). The system automatically generates the underlying DSL based on your selections.

Advanced Mode (JSON DSL)

For complex multi-step transformations, write custom JSON DSL scripts. This provides full control over the media processing pipeline with support for batch processing, Key-Value Store integration, and sophisticated operation chaining.

Media Input System

Input Sources

• Primary Content (media1): Main media from Transform Content node input
• Secondary Media (media2): Optional additional media input
• Additional Media (media3): Third media input for complex operations
• Key-Value Store: Load previously saved ImportSet references

Automatic Batch Processing

When ImportSets contain multiple media items, they are automatically loaded as:

• Single item: media1
• Multiple items: media1_0, media1_1, media1_2, etc.
• Item sets: Grouped for multi-input operations like video concatenation

JSON DSL Structure