Vision Platform / VisionEDGE
01.1 / Vision Platform Production Protects the signal

VisionEDGE

Station-level monitoring, where the broadcast chain begins.

A comprehensive, GPU-accelerated platform for complete monitoring of your entire channels and streams ecosystem. VisionEDGE watches every signal path across every supported protocol — video, audio, transport and metadata — detecting quality events, logging compliance and isolating anomalies through a single operator interface.

Bring us a feed See what it monitors
VisionEDGE — Multiviewer29 channels live
VisionEDGE multiviewer showing four live channels with per-signal video statistics, captions and alarm states
↳Live multiviewer — per-tile video statistics, caption status and signal/black/static/CC608 alarm states, monitored in real time.
Who it's for

Built for the people who keep the signal on air.

Chief Engineer

Signal integrity across the whole plant.

Provable fault detection on every path, with far fewer false alarms — so plant-wide signal integrity is something you can demonstrate, not just assert.

Master Control Operator

A real-time multiviewer at the desk.

Live monitoring with adaptive alerts that fit the tools you already trust — faults isolated and queued the moment they happen, where you're already watching.

Director of Engineering

One standard across the group.

Consolidate monitoring across heterogeneous facilities without ripping out what works — a single observability standard the whole group runs on.

What it does

Real-time observability at the station edge.

VisionEDGE consolidates real-time observability of every signal traversing master control onto a single appliance. Video, audio, transport and metadata are monitored together, and the moment a fault appears it's surfaced to the operator — in the multiviewer they already watch, with alerting that adapts to the station's existing escalation rules rather than imposing new ones.

It ships as a configurable on-prem GPU appliance in three hardware tiers, built to fit master-control workflows, not displace them — with zero cloud dependency for core monitoring. A protocol-abstraction layer normalizes all 26+ input protocols into one unified monitoring framework, so every stream — OTA antenna, SDI, IP multicast or adaptive streaming — enters the same analysis pipeline.

◆On-prem GPU applianceThree hardware tiers · 5 to 70+ concurrent streams
◆Zero cloud dependencyCore monitoring never leaves the plant
◆Protocol normalization26+ protocols into one pipeline
◆Deploys in hoursOperational within hours of hardware delivery
VisionEDGE — DashboardLive · 29 active
VisionEDGE command-center dashboard: statistics overview, alarm-status rings and live alarm feed
↳Command Center — statistics overview, per-group alarm-status rings, a live alarm feed and the full grid of monitored channels, with rolling SLA at the base.
Capabilities

Continuous intelligence on every feed.

GPU-accelerated computer vision and neural networks reading every frame — fault isolation, compliance and verification, on every signal at once.

01

Multiviewer suite

An extensive 20+ suite of multiviewers across the entire ecosystem, with dynamic, customizable, easy-to-configure layouts.

20+ viewsCustom layouts
02

Penalty Box

Anomalies are isolated and held with forensic clip playback — so an operator can review exactly what happened, frame by frame, after the catch.

Anomaly isolationForensic playback
03

GPU-accelerated analysis

Frame-level computer vision and neural networks on CUDA, classifying what crosses the signal in real time.

CUDAComputer visionNeural networks
04

Compliance suite

Continuous validation against the standards distribution requires — with evidence captured automatically.

CALM · LUFS/LKFSSCTE-35CEA-608/708Nielsen
05

Ad verification

Automated as-run verification by audio fingerprinting — confirming what aired actually aired, automatically.

Audio fingerprintAs-run
06

VelaIQ

Natural-language query across the plant — ask a question of the signal and get an answer. The operator's single pane of glass.

Natural-language queryPlant-wide

Role-based dashboards tailor the same monitoring to each desk — NOC, engineering, executive and AdOps each see the view they need.

What it monitors

130+ parameters, every dimension of stream quality.

Deep, continuous analysis across four measurement domains — each parameter with configurable thresholds, alerting and historical trending.

130+
Parameters across audio, video, transport & metadata
26+
Protocols supported natively
<1s
Sub-second fault-to-alert latency
5–70+
Concurrent streams across three hardware tiers
Video30+
  • Black & freeze-frame detection
  • Macroblocking & compression artifacts
  • Resolution, frame rate & format
  • Color space & gamut verification
  • Letterbox / pillarbox / aspect ratio
  • Quality scoring via pattern recognition
Audio20+
  • LUFS/LKFS loudness (CALM Act)
  • True-peak & loudness range
  • Silence, clipping & distortion
  • A/V sync drift measurement
  • Channel layout & codec validation
  • Audio fingerprinting for ad verify
Transport40+
  • Packet loss & burst-loss detection
  • Jitter, latency & bitrate analysis
  • PID continuity & PCR accuracy
  • PAT/PMT table consistency
  • ETR 290 Priority 1/2/3 compliance
  • Redundancy-path verification
Metadata25+
  • SCTE-35 splice detection & timing
  • CEA-608/708 closed captions
  • Nielsen watermark verification
  • EPG / PSIP validation
  • EAS alert monitoring
  • FCC compliance logging
VisionEDGE — Compliance Data · LKFS
Compliance data view charting LKFS loudness over time with violation markers and a per-timestamp data table
↳Loudness compliance — continuous LKFS trending against CALM thresholds, violations flagged in-line, exportable to CSV, Excel or PDF.
Protocol coverage

26+ protocols, one unified pipeline.

The protocol-abstraction layer normalizes every supported signal type into a single monitoring framework — broadcast, baseband, IP and streaming alike.

Over-the-air
ATSC 1.0ATSC 3.0DVBISDB-T
Baseband & facility
SD/HD/3G-SDIST 2110ST 2022-6NDIASI
Contribution
SRTZixiRISTRTP/RTSP
Adaptive streaming
HLSLL-HLSDASHCMAFRTMPWebRTC
MPEG transport
MPTSSPTSUDP/RTPMPEG-2 TS
Cable & satellite
QAMDVB-CDVB-S/S2UDP multicast
IP transport
UDPTCPMulticastUnicast
Cloud & file
CDN endpointsCloud originFile-based
VisionEDGE — TS InfoLive
Transport-stream info view: PID table, video and audio parameters, PSIP guide and bitrate composition
↳Transport analysis — PID map, PSIP guide, ETR 290 / ETSI error counts and live bitrate composition, per input.
Forensic inspection

Drill into any signal, down to the packet.

The Media Inspector pairs synchronized playback and timeline snapshots with the exact measurement behind every event — loudness, ratings watermarks, splice markers and more, on one timeline.

Media Inspector · LKFS
Media Inspector LKFS tab: playback with timeline snapshots beside a loudness chart and per-second data
↳Loudness — LKFS with true-peak, frame by frame.
Media Inspector · Nielsen
Media Inspector Nielsen tab: watermark detections with type, SID, layer, quality and strength
↳Nielsen — watermark type, SID, layer, quality & strength.
Media Inspector · SCTE
Media Inspector SCTE tab: splice events on a timeline with embed timecode, type and direction
↳SCTE — splice events, timed, with cue direction.
SCTE-35 — decoded message
SCTE-35 splice message fully decoded and inspectable
↳Every SCTE-35 splice message fully decoded and inspectable — not just flagged.
The operator interface

Every stakeholder sees their view.

Multiple interface layers ensure each desk — from NOC operator to executive — sees exactly the information relevant to their role.

Channel Events · Live Alarms
Channel Events live-alarms table with status, duration, source, channel and alarm type
↳Alert Management — prioritized queue, severity, acknowledgment and resolution times.
Media Inspector · Penalty Box
Media inspector with playback, timeline snapshots and a time-aligned caption panel
↳Penalty Box & forensics — isolated anomalies with clip playback and frame-accurate caption review.
Dashboard · Digital Triggers
Digital triggers view listing SCTE-35 marker events with counts and a timeline
↳Digital Triggers — SCTE-35 / SCTE-104 and HLS marker detection, counted and timelined.
Settings · Storage
Storage settings showing drive usage, retention categories and an in-progress cleanup
↳Retention & storage — drive usage, retention policy and automatic cleanup, on-prem.

VelaIQ sits across all of it — natural-language queries giving plain-English access to every monitoring record, the operator's single pane of glass. Role-based dashboards tailor the same data to NOC, engineering, executive and AdOps desks.

Deployment & integrations

Drops into the plant you already run.

Deployment

On-prem GPU appliance · three hardware tiers · zero cloud dependency for core monitoring.

↳Three hardware tiers — sized to the facility, from 5 streams to 70+; scale by adding appliances.
↳On-prem, zero cloud dependency — core monitoring runs inside the facility; the signal never leaves.
↳Deploys in hours — operational within hours of hardware delivery, configured into existing workflows.
↳Optional VisionCLOUD — a VEDGE-to-VCLOUD bridge adds enterprise-wide visibility for multi-facility groups.

Integrations

Built to reconcile with the systems already running the station.

↳REST API — station automation hooks into the monitoring layer directly.
↳Traffic-system reconciliation — verified as-run checked against scheduling and billing.
↳Alert delivery — email, SMS and webhook, with escalation paths and on-call scheduling.
↳Self-service portals — advertiser access to verification data and playback confirmation.
Report library

Evidence, generated automatically.

Standard reports ship with the platform and run on configured schedules — daily, weekly, monthly — with export to CSV, PDF or API. Clients can customize and add their own.

▸Stream Health Summary
▸Protocol Performance
▸Alert Volume & Pattern
▸Compliance Status
▸Penalty Box Audit Trail
▸Uptime & Availability
▸Video Quality Trending
▸Audio Compliance
▸Transport Health
▸SCTE-35 Verification
▸Caption Compliance
▸ETR 290 Compliance
▸SLA Performance
▸Incident Resolution
▸Root Cause Analysis
▸Cross-Stream Correlation
▸Trend & Prediction
▸Executive Summary
Where in the chain

The station-edge node of the Vision ecosystem.

VisionEDGE is the first node in the signal chain — at the station, where the signal begins. Each Vision solution stands on its own; together they map the path the signal travels.

At the station

VisionEDGE

Monitoring where the broadcast chain begins.

→
Aggregation

VisionCLOUD

Cloud-scale rollup across the fleet.

→
Consumer edge

VisionPROBE

Verification out at the point of delivery.

Proven in the plant

Production-grade, not a prototype.

Built to run unattended in real master-control environments — on a codebase with a broad broadcast-distribution footprint, monitoring the whole signal around the clock.

~1,000
Stations on the broadcast-distribution codebase it’s built on
5–70+
Concurrent streams per appliance
24/7
Continuous unattended operation
Zero
Cloud dependency for core monitoring
Foundation verify

Built on the codebase behind a broadcast-distribution installed base of roughly 1,000 stations.

Autonomy

Once configured, monitors continuously — detecting, logging, alerting and capturing forensic clips without operator attention during normal operation.

Bring us a feed

See VisionEDGE on your own signal.

Engagements start with a signal. Bring a representative feed and we run it through the audit pipeline — you see exactly what VisionEDGE catches, before any commercial conversation.

01

Feed assessment

02

Platform audit

03

Operator review

04

Deployment scoping

Start with a signal →

info@vela.com · (727) 507-5300