Skip to main content

Vylux

Vylux is a standalone media processing service that separates real-time image transformation and asynchronous video processing into deployable infrastructure capabilities. It does not own your business model. It accepts source objects, validation data, and processing parameters, then produces cacheable, traceable, playable media assets.

Current core capabilities include:

  • real-time image transformation through /img
  • signed original and thumbnail delivery through /original and /thumb
  • async jobs: image:thumbnail, video:cover, video:preview, video:transcode, and video:full
  • HLS CMAF output with AV1 and H.264 ladders
  • encrypted playback with Bearer-token key delivery through /api/key/{hash}
  • PostgreSQL job state, Redis queues, Prometheus metrics, and OpenTelemetry tracing
Start from the page that matches your job

Most confusion comes from reading an API page before you have chosen your deployment shape or trust boundary. Pick the path below that matches what you are actually trying to do.

Runtime shapes​

all​

  • runs the HTTP server and the worker in one process
  • best for local development, staging, and smaller single-node deployments

server​

  • runs only the HTTP server, image delivery endpoints, playback routes, and main metrics surface
  • pair this with a separate worker when you want cleaner scaling and failure isolation

worker​

  • runs only the queue consumer and the worker metrics listener
  • use this when FFmpeg, libvips, and packaging workloads should scale independently from HTTP traffic

This lets you use the same image for local development, Docker Compose, single-node deployments, and split server/worker layouts on Kubernetes.

Verified capabilities today​

  • /img real-time image resize, format conversion, and caching
  • animated video:preview output in webp and gif
  • video:transcode HLS CMAF output
  • dual-codec ladders with AV1 and H.264
  • portrait and non-16:9 video output sizing
  • raw-key CBCS / SAMPLE-AES protected streaming
  • /api/key/{hash} Bearer-token key delivery
  • PostgreSQL job state, Redis queues, Prometheus metrics, and OpenTelemetry tracing

First run​

  1. Start with Getting Started
  2. Continue with Configuration and Deployment
  3. Finish with Observability once the service is reachable

Application integration​

  1. Read Integration Guide
  2. Continue with Jobs API and Image Delivery API
  3. If you need streaming or DRM-style flows, continue with Playback API and Encrypted Streaming

Operations​

  1. Read Deployment
  2. Continue with Configuration
  3. Use System Endpoints and Observability for probes, metrics, and troubleshooting

Docs scope​

This docs site focuses on:

  • service structure and core data flow
  • media processing pipelines
  • HTTP endpoints, auth models, and curl examples
  • deployment, configuration, observability, and testing

Docs principles​

  • current code and tests take priority over old design notes
  • operational usefulness takes priority over abstract architecture prose
  • knowledge that used to live in temporary root-level files or local helper folders is being folded into these docs so the published site remains self-contained