Loading…
Loading…
The foundation of Visylix. A proprietary streaming engine built from scratch with native H.264/H.265 support. Not a wrapper around FFmpeg or an open-source streaming server. Handles thousands of concurrent streams per node (internal testing), with bounded per-stream memory and multi-cluster federation designed toward city-scale estates. Multi-protocol support including GB28181.
Two ingestion modes to accommodate every camera, encoder, and deployment scenario. Visylix adapts to your infrastructure, not the other way around.
Cameras and encoders push streams directly to Visylix via RTMP, SRT, or WebRTC. Ideal for dynamic environments where devices initiate connections, such as mobile units, drones, and body-worn cameras.
Visylix pulls streams from cameras and NVRs over RTSP or other protocols. Best for fixed installations where the VMS manages the connection lifecycle and camera inventory.
Seven core protocols shown here, out of 13+ in the engine. Each protocol is purpose-built for specific latency, reliability, and compatibility requirements.
Industry-standard protocol for IP camera integration. Supports TCP and UDP transport with ONVIF compatibility.
Mature push-based protocol for encoders and broadcast equipment. Persistent TCP connections with low overhead.
Sub-second WebRTC live view with DTLS-SRTP encryption, WHEP playback and WHIP ingest. Ideal for real-time monitoring and two-way communication.
Secure, reliable transport over unpredictable networks. AES-128/192/256 encryption with ARQ retransmission for WAN delivery.
Apple Low-Latency HLS for web and mobile clients. Partial segments and preload hints minimize glass-to-glass delay.
Standard HTTP Live Streaming for maximum compatibility across browsers, smart TVs, and set-top boxes at scale.
China national standard for public safety video surveillance. Full SIP signaling with RTP media transport for government and municipal deployments.
Capture what matters, when it matters. From continuous 24/7 archival to intelligent event-triggered recording.
24/7 recording with configurable retention policies. Frames are written directly from the ingestion pipeline with zero re-encoding overhead.
Define recording windows by day of week, time range, and camera group. Supports recurring and one-time schedules with timezone awareness.
Recording starts and stops based on AI detection events, external API triggers, or sensor input. Each rule-triggered recording runs for a set duration from 5 seconds to an hour.
Users or API consumers trigger recording manually for specific streams. Supports duration-limited and indefinite capture with real-time status updates.
One universally playable container that survives abrupt shutdowns and opens instantly in any browser.
Standard MPEG-4 recording, fsync-durable and finalised faststart (moov first) for instant browser playback and seek. Fragmented MP4 and MPEG-TS are used for LL-HLS, DASH and HTTP-TS live output.
Long-term archival, export, and evidence management
Frames flow from ingestion to output without unnecessary memory copies, achieving maximum throughput with minimal CPU overhead.
Frames arrive via RTSP, RTMP, SRT, or WebRTC and are placed into shared memory buffers. No copy occurs during protocol demuxing.
The stream router maps each frame reference to subscribed consumers: recording, AI pipelines, and output protocols. Only pointers are forwarded.
AI models read frames directly from shared memory. Recording writes from the same buffer. No intermediate copies at any stage.
Output protocols (WebRTC, HLS, LL-HLS) read the original frame data for packaging and delivery. The frame is released only when all consumers finish.
A proprietary engine purpose-built for surveillance workloads. Every component is designed for maximum throughput, minimal latency, and predictable resource consumption.
The engine scales horizontally across nodes, and multi-cluster federation is designed toward very large, city-scale estates. Exact per-node capacity depends on resolution, codec, frame rate, and analytics load, so we validate it on your own hardware during the proof of concept.
Live monitoring via WebRTC delivers sub-second glass-to-glass live view, enabling real-time decision-making for security operations. Contrast that with the multi-second HLS-based viewing typical of cloud VSaaS.
H.264, H.265 (HEVC), AAC, G.711 (A-law/u-law) and Opus are handled in-engine. No FFmpeg anywhere in the media path.
The engine is built for dense deployment on commodity hardware. Exact stream density per server depends on resolution, codec, frame rate and analytics load, so we measure it on your own hardware during the proof of concept.
Intelligently drops non-reference frames for slow subscribers while preserving keyframes. Viewers on constrained connections still see smooth video.
New viewers receive the latest Group of Pictures immediately, so streams start on the cached keyframe instead of waiting for the next one.
Detects camera codec capabilities and performs real-time H.265 to H.264 transcoding for browsers without HEVC support. Zero manual configuration.
G.711 and AAC audio are transcoded to Opus for WebRTC delivery with automatic sample rate conversion. Crystal-clear audio on every client.
Monitors CPU load in real-time and gracefully degrades at 70%, 85%, and 95% thresholds. It sheds load in stages instead of falling over when a spike hits.
Per-stream GOP caches, write buffers, and connection buffers are all bounded to prevent out-of-memory conditions. Predictable resource consumption at scale.
Complete operational visibility from a single pane of glass. Monitor cameras, streams, AI detections, storage, and system health in real-time.
See online, offline, and connecting cameras at a glance. Instant visibility into your entire camera fleet from a single pane of glass.
Real-time viewer count per stream and total active connections across the platform. Know exactly who is watching what, when.
Live recording status with duration tracking and per-stream storage usage. Never lose track of what is being captured.
Real-time detection counts by AI model type. Monitor face recognitions, license plates, safety gear violations, and more as they happen.
CPU, memory, disk, and network throughput displayed in real-time. Full visibility into infrastructure health without leaving the dashboard.
Total used space and available capacity, with configurable low-disk alerts. Plan storage expansions before you run out of space.
Detection trends, viewer counts, and system metrics visualized over time. Spot patterns and anomalies in your surveillance data.
24-hour, 7-day and 30-day trends for detections, viewers and system load, ready for stakeholder reviews.
Drill into any individual camera for detailed analytics: uptime history, detection events, viewer patterns, and storage consumption.
Every camera's stream, recording and connectivity health in one in-product view, so problems surface before footage is lost.
Acknowledge, assign and close alarms from one operator console, with a full history of every action taken.
Live camera status and alarm badges on your own floor plans, plus a geographic map view on your own map tiles.
Deploy, configure, monitor, and update your entire surveillance platform without leaving the browser. Built for operators who value simplicity.
Deploy the entire Visylix platform with a single curl command. From bare metal to fully operational in minutes, not hours.
The installer detects your hardware and automatically tunes database pools, worker threads, buffers, and connection limits for optimal performance.
The dashboard flags every new release, and the signed one-command updater backs up first and rolls back automatically if anything fails.
Monitor the health of all internal services from a single view. Immediate alerts when any component requires attention.
Browse, filter, and download logs directly from the dashboard. No SSH access or terminal required for troubleshooting.
One-click download of all system logs with sensitive data automatically redacted. Share with support without security concerns.
The system detects available GPUs and configures AI inference pipelines automatically. Plug in a GPU and the engine uses it.
Database pools and worker counts are sized from the host's CPU cores and RAM at install and start-up. Resources are allocated where they are needed most.
Configurable low-disk alerts with automatic notification. Receive warnings at custom thresholds before storage runs out.
Learn how AI analytics, deployment options, and integration APIs build on top of the VMS Core Engine.
The Visylix VMS Core Engine supports thousands of concurrent video streams per node using an optimized architecture, with multi-protocol ingestion including RTSP, RTMP, WebRTC, SRT, and HLS. Multi-cluster federation is designed toward very large, city-scale estates, and exact per-node capacity is validated on customer hardware during the proof of concept.
Visylix supports RTSP, RTMP, WebRTC, SRT, HLS, and LL-HLS protocols with both push and pull ingestion modes, enabling seamless integration with virtually any IP camera or video source.
The Visylix streaming engine uses a multi-layered protection system: adaptive backpressure drops non-reference frames for slow subscribers while preserving keyframes, a circuit breaker monitors CPU load and gracefully degrades at 70%, 85%, and 95% thresholds, and all per-stream buffers are memory-bounded to prevent out-of-memory conditions. Combined with GOP cache for instant playback and automatic codec negotiation, the engine maintains stability even under unexpected load spikes.
The Visylix dashboard provides complete operational visibility including camera status overview (online, offline, connecting), per-stream viewer counts, recording status with storage usage, AI detection counts by model type, system metrics (CPU, memory, disk, network), storage usage analytics, 24-hour, 7-day and 30-day trend charts, a per-camera detail view, a device-health console and an alarm and incident console.
Visylix can be deployed with a single curl command. The installer auto-detects hardware and tunes database pools, worker threads, buffers, and connection limits. Ongoing administration is handled from the browser dashboard: new-release notifications (applied by a signed one-command updater with automatic rollback), service health monitoring, log viewing and filtering, one-click diagnostic bundle download with sensitive data redacted, automatic GPU detection for AI inference, and configurable low-disk storage alerts.