The framework for building
professional media systems.
A modular C++ framework for professional video and audio pipelines. Pick the modules you need, compose them into a system of any complexity, and run the same code on Windows, Linux, and macOS — in C++, C#, or Python.
// Define a pipeline in JSON — // xMedia SDK builds the node tree { "pipeline": { "ingest": { "type": "SourceStreamsAdd", "url": "srt://cam1:9000" }, "process": [ { "type": "Transcode", "codec": "h264", "hw": "cuda" }, { "type": "ReplayBuffer", "depth_sec": 300 } ], "output": [ { "type": "Publish", "url": "srt://cdn:4200" }, { "type": "Record", "path": "/archive/{date}.ts" } ] } } // Same config. Windows, Linux, macOS.
A framework, not a library.
xMedia SDK is built around a tree of processing nodes. Each node is a module — a decoder, mixer, recorder, streamer. You connect modules into a pipeline. The framework handles data flow, threading, synchronization, and inter-node messaging automatically.
Modular
Pick only the modules you need — Base, Playlist, Mixer, Replay, CG, Chroma Key, Writer, xNetwork, Subtitles. Each is independent. Add capabilities without touching existing code.
Cross-platform
One codebase. Windows, Linux, macOS. The same API, the same module interfaces, the same behavior — whether you deploy to a cloud VM, on-prem appliance, or edge device.
Multi-language
C++, C#, Python — all share the same module API and interfaces. Control any pipeline remotely via WebSocket from any language or platform, including browsers.
Built for speed
Codec integration, hardware acceleration, SRT/NDI/SDI support, sync, buffering — already built. You compose pipelines from ready modules, not implement infrastructure from scratch.
Try it now. Full pipeline in minutes.
The core pipeline modules are production-ready. Download the sample from GitHub — runs on Windows, Linux, and macOS out of the box. No configuration required to get started.
Download sample
Open-source sample on GitHub. Clone, build, run. Full ingest-process-output pipeline ready in minutes.
Get on GitHub →Web UI included
Open browser, connect to your running xMedia instance. Add sources, monitor streams, control playback — no extra setup needed.
Runs locallyWebSocket API
Control everything programmatically. The same API the Web UI uses — automate from Python, JS, C#, or any language.
Port 8765 · JSON · RPCIngest
- SRT, RTMP, RTSP, UDP, HLS
- NDI sources on the network
- Blackmagic SDI / HDMI
- File: TS, MP4, MXF
- Remote xMedia peer streams
Process
- Decode, scale, FPS conversion
- HW transcode: CUDA, QSV
- Mixer — multiple sources
- Graphic / logo overlay
- Video filtergraph support
Replay Buffer
- Frame-accurate seek & scrub
- Slow motion, reverse playback
- RAM or disk, any depth
- Export: MP4, TS, PNG frame
- Publish delayed output
Output
- Push SRT, RTMP, UDP, RTSP
- NDI output — name and publish
- Blackmagic DeckLink SDI/HDMI
- 24/7 segmented recording
- Browser preview via WebRTC
Works on all major platforms — today
Same sample, same API, same behavior across operating systems.
One framework. Many systems.
xMedia SDK is the infrastructure layer inside broadcast tools, OTT platforms, sports replay systems, MAM solutions, and security software.
Live production & channel automation
Ingest live feeds and file assets into a single pipeline. Mix sources, overlay CG graphics, encode to broadcast specs, push to SDI, NDI, and CDN simultaneously. Add the Playlist module for schedule-driven channel automation — one codebase runs the whole channel.
Instant replay & slow-motion
Buffer every ISO camera feed from the moment broadcast starts. The moment the play ends — seek to any frame, scrub at any speed, export the clip. All frame-accurate, controlled via API from any operator interface you build on top.
Scalable ingest and delivery platform
Build the ingest and processing core of an OTT platform or media server. Multi-bitrate transcoding, DRM-ready packaging, delivery to unlimited endpoints — all in one pipeline, one codebase, deployable on any OS.
Recording, browse & clip workflows
Continuous 24/7 recording segmented into files. Read from a file that's currently being written. Frame-accurate clip access for browse and export workflows. Subtitle decode for archive search and indexing.
Distributed camera processing at scale
Ingest RTSP/ONVIF streams from IP cameras across your network. Process, record, and distribute video to monitoring stations, archive, and remote viewers. Lightweight per-node footprint — runs headless on Linux appliances, managed remotely via WebSocket API.
Modular by design. Add capabilities without rewriting.
Each module is a self-contained unit with a clean API. Add it to your pipeline when you need it — the rest of your system doesn't change.
Base module
Node framework, actor model, module factory, JSON pipeline config, inter-module messaging
WebSockets module
Inter-module communication across servers — enables fully distributed systems
xNetwork module
SRT (with bonding), WebRTC with SFU/MCU, low-latency delivery over IP
Playlist module
Seamless multi-format playout, live + file mixing, schedule-driven broadcast automation
Writer module
Simplified encoding API — choose codec and output, the module handles the rest
Sync module
NTP/PTP clock synchronization across pipeline nodes, <10ms accuracy
Mixer module
Combine multiple video/audio sources with varying formats into a unified output stream
CG (HTML5) module
Overlay dynamic HTML5 graphics as on-screen CG visuals — data-driven, animated
Chroma Key / Video Effects
Real-time background removal and replacement — live broadcast and virtual studio ready
Replay module
Frame-accurate buffer (RAM/disk), seek, slow/fast motion, reverse, clip export as MP4/TS
Subtitle modules
Decode and encode DVB subs, DVB teletext, SRT, bitmap and text-based subtitles
Up and running in minutes.
Download, run, send your first command. All samples are ready to copy and launch — no boilerplate needed.
Run the instance
Launch xMedia on your machine. WebSocket server starts automatically on port 8765.
./xmedia-server --port 8765Copy a sample and go
Pick any sample from GitHub. Copy, adjust your source URL, run.
github.com/Medialooks-LLC/xmedia_api →API-first by design
Every capability is equally available through the API — with the same performance and latency guarantees. Pick the interface that matches your stack.
WebSocket API
Primary control interface. Full bidirectional RPC with JSON payloads. Subscribe to stream events, issue commands, get real-time status. Works from any language, framework, or browser.
C / C++ Native API
Deep integration for building products on top of xMedia SDK. Zero-copy access to stream data, direct frame callbacks, minimal overhead — full control over the node tree.
gRPC API
High-throughput service-to-service automation. Strongly typed Protobuf schema, bi-directional streaming, lower overhead for high-frequency control loops.
Ready to build?
Get API access and start building your media pipeline today. Our team works directly with integrators — no ticketing queue, no generic onboarding.
Request a call back
Schedule a call with our engineering team. We'll discuss your use case and recommend the right integration path.
Request a call backOrder a demo
See xMedia SDK working live against your specific pipeline — ingest, replay, or output scenarios.
Order a demoAsk a question
Have a specific technical question? Our team responds fast — usually within one business day.
Ask a question
1. Types of information we collect
1.1. Information you provide to us
When you sign up for our services, download our Products from the Website, use our Products, place an order for any of our Products or services, communicate with our technical or sales teams, or communicate with us in any way (including offline), you are deliberately giving us information that we collect. This information may include your name, email address, address, phone number, etc. You may provide us with audio & video materials that may be necessary to help with a support claim.
We refer to this voluntarily-submitted information as "Client Information". By giving us this information, you consent to this information being collected, used and processed as described in this Policy.
1.2. Information we collect automatically
When you visit our Website, we automatically collect certain information about your device, including information about your web browser, operating system, IP address, language, time zone, etc. As you browse the Website, we collect information about the individual web pages that you view, what websites or search terms referred you to the Website, and information about how you interact with the Website. We may collect this information through the use of cookies or other tracking technologies such as pixel tags, local storage, or server logs.
When you voluntarily submit a support or error report from within our Product, we may optionally collect additional data such as software version, settings, system specifications (such as make and model of computer components), streaming or recording logs and error messages or reports contained within the event logs of your operating system.
We refer to this automatically-collected information as "Device Information".
2. How we use your information
2.1. Client Information
We may use Client Information (most notably, your name and email address) to fulfill orders or requests that you may place with us, including, but not limited to:
- provide you with a way to sign up for or download our Products;
- process your order: organize and provide presales and sales activities;
- provide you with ways to use or access our Products (such as by delivering a license key or login credentials);
- provide support services for the Products you use;
- inform you about activation instructions and subscription expiration deadlines;
- inform you about new Products, new versions of the Products, our activities as a company and new information available on our Website or other resources;
- in general, communicate with you about our Products and services.
2.2. Device Information
We may use your IP address to:
- protect from fraud on our Website (such as to prevent repeated software downloads);
- establish connections between Product instances.
We may use information that our Products automatically collect (such as Product version, device hardware specifications and network performance data) to improve the performance of our Products.
We use Google Analytics to help us understand how our customers use the Website and our Products. We do not use Google Analytics in our Direct Take application.
We use Google AdWords remarketing services to advertise to you on third-party websites (including Google) after visiting our Website.
3. Storage
3.1. Information required to fulfill your orders is stored by us on our servers in Germany.
3.2. We use third-party providers to support our sales (CRM) and customer service (ticketing system) activities. Storage is therefore defined by the policies of such providers.
4. Control
4.1. All our newsletters (with the exception of transactional emails) contain an unsubscribe link; you always can use this link to opt-out from our newsletter.
4.2. To access, modify, export or delete your Information, please submit your request via email to sales@medialooks.com.
4.3. You can always manage, enable or disable cookies in your browser. To modify the ways you allow your browser to use cookies, please see the documentation of your browser.
Fill out the form and we'll be in touch as soon as possible.