System Architecture
System Architecture
Generated from profile: studio Last Updated: 2026-01-20
High-Level Overview
This pipeline ingests 4K video from DSLR/mirrorless cameras, synchronizes audio via Dante network, enables low-latency NDI caller integration, and outputs to streaming platforms (YouTube, Twitch) via OBS.
Design principle: Separation of concerns — capture, sync, mixing, output are independent subsystems connected via standardized protocols (HDMI, Dante, NDI, RTMP).
Signal Flow Diagram
┌─────────────────────────────────────────────────────────────┐
│ INPUT LAYER │
├─────────────────────────────────────────────────────────────┤
│ │
│ Cameras (HDMI) Cameras (Audio) │
│ ├─ [Camera 1] ├─ USB or XLR │
│ ├─ [Camera 2] └─ Dante converters │
│ ├─ [Camera 3] │
│ └─ [Camera 4] Callers (Remote) │
│ ↓ └─ NDI OBS stream │
│ │
│ () │
│ ↓ │
│ │
├─────────────────────────────────────────────────────────────┤
│ CAPTURE LAYER () │
├─────────────────────────────────────────────────────────────┤
│ │
│ │
│ ├─ Scene 1: Camera quad (4K mosaic) │
│ ├─ Scene 2: Single camera + caller │
│ ├─ Scene 3: Screen share (graphics) │
│ └─ Audio Input: │
│ │
│ │
│ Dante Network (audio sync) │
│ ├─ (Dante I/O) │
│ ├─ Dante AVIO nodes (camera audio converters) │
│ └─ Sample clock: 48 kHz (broadcast standard) │
│ │
├─────────────────────────────────────────────────────────────┤
│ PROCESSING LAYER │
├─────────────────────────────────────────────────────────────┤
│ │
│ (audio mixing) │
│ ├─ Input: Camera feeds (via Dante) │
│ ├─ Master clock: Drives all Dante devices │
│ └─ Output: Master mix → Audio Interface → OBS │
│ │
│ (optional) │
│ ├─ Input: Generative parameters (MIDI from DAW) │
│ └─ Output: Screen share → OBS │
│ │
├─────────────────────────────────────────────────────────────┤
│ OUTPUT / ENCODING LAYER │
├─────────────────────────────────────────────────────────────┤
│ │
│ OBS Encoder (H.264 or H.265) │
│ ├─ Video bitrate: 8–25 Mbps (codec-dependent) │
│ ├─ Resolution: 4K or 1080p mosaic │
│ └─ Framerate: 30fps or 60fps │
│ ↓ │
│ RTMP Output (streaming protocol) │
│ ↓ │
├─────────────────────────────────────────────────────────────┤
│ DISTRIBUTION LAYER │
├─────────────────────────────────────────────────────────────┤
│ │
│ Streaming Platforms: │
│ ├─ YouTube │
│ └─ Twitch │
│ │
│ Archive & Analytics: │
│ ├─ Local recording (OBS → H.264 file) │
│ └─ Optional: Cloud backup of raw recordings │
│ │
└─────────────────────────────────────────────────────────────┘
Key Components
1. Capture Subsystem (HDMI + Video Capture)
- **: 4× HDMI input, outputs to PCIe
- **: thunderbolt3 enclosure, PCIe slot for capture card
- Protocol: Uncompressed HDMI or HDCP-encrypted video
- Latency: <1ms (hardware-based capture)
- Bandwidth: ~1.5 Gbps per 4K30 stream
2. Audio Synchronization (Dante)
Why Dante?
- All audio sources share a global sample clock
- Avoids drift between video and audio
- Supports flexible routing without hardware reconfiguration
- Industry standard for broadcast
Components:
- Dante Network: Dedicated managed Ethernet switch
- **: 24 in / 28 out channels
- **: Acts as Dante clock master
3. Computer System
| Specification | Value |
|---|---|
| Model | |
| Chip | |
| RAM | GB |
| OS |
4. Network Configuration
| Interface | Purpose |
|---|---|
| en0 | Primary network (streaming) |
| Dante Network | Dedicated audio network |
Operational Thresholds
| Parameter | Value |
|---|---|
| Minimum Disk Space | 50 GB |
| Recommended Disk Space | 100 GB |
| Max CPU Usage | 70% |
| Min Memory Free | 20% |
This document was generated from profile: studio