AVOLO-CAM

Multi-iPhone NDI streaming system for OBS with desktop remote control.

Overview

AVOLO-CAM enables multiple iPhones to stream high-quality, low-latency video to OBS via NDI|HX, controlled from a single desktop application. Perfect for multi-camera productions, live events, and professional streaming setups.

Key Features:

📱 Multiple iPhone cameras streaming NDI to OBS
🎛 Desktop controller for unified management
🔄 Real-time telemetry (FPS, bitrate, battery, temperature)
👥 Group control for batch operations
📡 Automatic camera discovery via mDNS
🎨 Rec.709 Full color pipeline end-to-end
⚡ Low latency (≤150ms glass-to-glass target)

Project Status

✅ LOT A - MCP Core: Structure and architecture complete

iOS app structure with all core managers implemented
Tauri controller with mDNS discovery and group control
Full API contracts defined
Ready for NDI SDK integration and testing

📝 Next Steps:

Complete Xcode project creation
Integrate NDI SDK for iOS
Implement SwiftNIO HTTP/WebSocket server
Test with physical iPhones and OBS
Measure and optimize latency

Architecture

flowchart TB

%% ---------------- LAYERS ----------------
subgraph Layer1["🎛 Control Layer"]
  Tauri["Tauri Controller (Desktop)"]
  CBus["HTTP/WS Control & Telemetry Bus"]
  Tauri --> CBus
end

subgraph Layer2["📱 Capture Layer — AvoCam Devices"]
  direction LR
  P1["iPhone #1"]
  P2["iPhone #2"]
  P3["iPhone #3"]
end

subgraph Layer3["🖥 Video Ingest Layer"]
  VBus["NDI HX Network"]
  OBS["OBS (NDI Receiver)"]
  VBus --> OBS
end

%% ----------- FAN-OUT / FAN-IN -----------
CBus --> P1
CBus --> P2
CBus --> P3

P1 --> VBus
P2 --> VBus
P3 --> VBus

Components

1. iOS App (ios-app/)

Swift iOS application for iPhone cameras.

Features:

AVFoundation video capture (1080p30 baseline, up to 4K)
VideoToolbox H.264 encoding (low-latency, no B-frames, GOP=fps)
NDI|HX transmission with metadata
HTTP REST API for control (Bearer token auth)
WebSocket telemetry (1Hz updates)
Embedded web UI for standalone control
Bonjour/mDNS advertisement
Thermal management with auto bitrate adjustment

Status: 🚧 Core structure complete, needs Xcode project + NDI SDK integration

→ iOS App README

2. Tauri Controller (tauri-controller/)

Rust + Svelte desktop application for multi-camera control.

Features:

mDNS camera discovery
Grid view with real-time telemetry
Group control (start/stop/settings for multiple cameras)
Bounded concurrency (Semaphore-based parallelism)
WebSocket telemetry subscriptions
Manual camera addition (fallback for restricted networks)
Profile management (LOT B)

Status: ✅ Fully implemented and ready for testing

→ Tauri Controller README

3. OBS Integration

Standard NDI Source plugin (no custom development required).

Configuration:

Install NDI Plugin for OBS
Project settings: Rec.709, Full range
Disable per-source rescale (use scene-level only)
Add NDI Source, select "AVOLO-CAM-XX"

Quick Start

Prerequisites

iOS Development: macOS with Xcode 15+, iOS 15+ iPhones
Desktop App: Rust 1.70+, Node.js 18+
OBS: OBS Studio with NDI Plugin
Network: All devices on same WiFi (with multicast support)

Setup

iOS App:

cd ios-app
# Follow ios-app/README.md to create Xcode project
# Integrate NDI SDK
# Build and deploy to iPhones

Tauri Controller:

cd tauri-controller
npm install
npm run tauri:dev

OBS:
- Install NDI Plugin
- Configure project: Rec.709 / Full
- Add NDI sources

Usage

Launch AvoCam on iPhones (same WiFi network)
Launch Tauri Controller on desktop
Cameras appear automatically (or add manually)
Click "▶️ Start" to begin streaming
Add NDI sources to OBS scenes

Technical Specifications

Video Pipeline

Capture: AVFoundation, 720p-4K @ 24-60fps
Encode: VideoToolbox H.264, CBR, real-time, no B-frames
Bitrate: 8-12 Mbps (configurable, thermal-aware)
Color: Rec.709 Full Range end-to-end
Latency: ≤150ms glass-to-glass target

API (iOS → Controller)

REST: HTTP on port 8888 (configurable)
Auth: Bearer token (generated per camera)
WebSocket: ws://<ip>:8888/ws for telemetry (1Hz)
Discovery: mDNS _avolocam._tcp.local.

Endpoints:

GET /api/v1/status - Current status + telemetry + capabilities
GET /api/v1/capabilities - Supported formats (per-lens)
POST /api/v1/stream/start - Start NDI stream
POST /api/v1/stream/stop - Stop NDI stream
POST /api/v1/camera - Adjust settings (WB, ISO, shutter, zoom)
POST /api/v1/encoder/force_keyframe - Force IDR frame
GET /api/v1/logs.zip - Download rotating logs

WebSocket Telemetry:

{
  "fps": 29.97,
  "bitrate": 9800000,
  "queue_ms": 6,
  "battery": 0.78,
  "temp_c": 38.4,
  "wifi_rssi": -55,
  "ndi_state": "streaming",
  "dropped_frames": 0,
  "charging_state": "unplugged"
}

Group Control

Bounded Concurrency: Max 10 parallel operations (tokio::Semaphore)
Atomic Fan-out: All selected cameras receive command
Per-camera Results: Individual success/failure reporting
Target Latency: <250ms for batch operations

Acceptance Criteria (LOT A - MCP)

Multi-cam Streaming: ≥3 iPhones streaming 1080p30 @ 8-12 Mbps
Stability: ≥2 hours continuous, <1% frame drops
Latency: Glass-to-glass ≤150ms median
Resolution Switch: <3 seconds
Group Control Latency: <250ms to all targets
Reconnect: <2 seconds after network blip
Color Accuracy: Rec.709 Full maintained end-to-end

Development Roadmap

LOT A - MCP Core (Current) ✅ Structure Complete

iOS video pipeline with low-latency encoding
NDI|HX transmission
HTTP/WebSocket control API
Tauri controller with mDNS discovery
Group control with bounded concurrency
Basic web UI (embedded in iOS)

Remaining: NDI SDK integration, SwiftNIO server implementation, device testing

LOT B - Stability & Multi-Cam Hardening

Reconnect logic (keep-alive, fast resume)
Thermal guard with UI warnings
Network quality indicators (RSSI)
Settings profiles (save/recall/copy)
Per-cam rename, tagging, groups
Read-only monitoring mode

LOT C - Image Quality & Ops

Orientation lock, lens selection (ultra-wide/wide/tele)
Anti-banding (50/60 Hz), WB presets (3200K/4300K/5600K)
IDR on demand, optional B-frames toggle
Test patterns (SMPTE bars, focus chart, 1kHz tone)
On-device preview refinement

LOT D - Diagnostics & Admin

Diagnostics endpoint (dropped frames, queue depth, temp timeline)
Rotating logs with downloadable zip
Telemetry charts (sparklines)
Config backup/restore (full fleet snapshot)
Stealth mode (API only, no mDNS)

LOT E - Polish & Extensions (Optional)

Adaptive bitrate ladder (10→8→6 Mbps)
NTP time-stamping, basic tally return
Optional TLS (self-signed)
Ethernet guidance (USB-C→RJ45)
LUT pipeline / HDR→SDR

Documentation

CLAUDE.md - Architecture and implementation guidance for AI assistants
LOT-A-CHECKLIST.md - Detailed task breakdown (150+ items)
docs/specs.md - Complete project specifications
ios-app/README.md - iOS app setup and development
tauri-controller/README.md - Desktop controller setup

Testing

Network Matrix

Wi-Fi 6 single AP vs mesh
PC on gigabit Ethernet
1080p30 @ 8/10/12 Mbps
3-6 camera configurations

Validation Checklist

mDNS discovery across WLAN↔LAN
OBS NDI Source configuration (Rec.709 / Full)
Latency measurement (LED clock method)
Bitrate stability (<1% drops over 2h)
Resolution switch speed (<3s)
Group command latency (<250ms)
Reconnect after AP blip (<2s)
Thermal behavior monitoring
Color accuracy validation

Troubleshooting

mDNS Discovery Issues

Cameras not appearing: Check network allows multicast, not on guest VLAN
Firewall: Allow port 5353 UDP
Fallback: Use manual camera addition in controller

Streaming Issues

High latency: Check WiFi signal, verify GOP=fps and no B-frames
Frame drops: Reduce bitrate, check network congestion
Color mismatch: Verify OBS project is Rec.709 / Full
Thermal throttling: Lower bitrate/fps, improve iPhone ventilation

API Issues

401 Unauthorized: Check Bearer token matches
Timeout: Verify iPhone is on network, check firewall
Rate limit 429: Slow down camera settings updates (50ms min interval)

Contributing

This is an internal project. For development:

Check LOT-A-CHECKLIST.md for current tasks
Read architecture in CLAUDE.md
Follow code style in existing files
Test on physical devices before committing

License

Internal use only. NDI SDK license applies to NDI components.

Resources

Support

For questions and issues:

Review inline code comments (especially NDIManager and NetworkServer)
Check CLAUDE.md for architectural guidance
See component READMEs for specific setup instructions

Project Timeline: 4-6 weeks for LOT A MCP completion

Target: Professional multi-camera NDI streaming system with desktop control

Name		Name	Last commit message	Last commit date
Latest commit History 134 Commits
.claude		.claude
docs		docs
ios-app		ios-app
tauri-controller		tauri-controller
.gitignore		.gitignore
.mcp.json		.mcp.json
CLAUDE.md		CLAUDE.md
LOT-A-CHECKLIST.md		LOT-A-CHECKLIST.md
OPTIMIZATION_SUMMARY.md		OPTIMIZATION_SUMMARY.md
PERFORMANCE_OPTIMIZATIONS.md		PERFORMANCE_OPTIMIZATIONS.md
PERF_CHANGES_SUMMARY.md		PERF_CHANGES_SUMMARY.md
PROJECT-STATUS.md		PROJECT-STATUS.md
README.md		README.md
REDESIGN_SUMMARY.md		REDESIGN_SUMMARY.md
TODO.md		TODO.md
WARP.md		WARP.md
package.json		package.json
pnpm-lock.yaml		pnpm-lock.yaml

JulienCr/avolo-cam

Folders and files

Latest commit

History

Repository files navigation

AVOLO-CAM

Overview

Project Status

Architecture

Components

1. iOS App (ios-app/)

2. Tauri Controller (tauri-controller/)

3. OBS Integration

Quick Start

Prerequisites

Setup

Usage

Technical Specifications

Video Pipeline

API (iOS → Controller)

Group Control

Acceptance Criteria (LOT A - MCP)

Development Roadmap

LOT A - MCP Core (Current) ✅ Structure Complete

LOT B - Stability & Multi-Cam Hardening

LOT C - Image Quality & Ops

LOT D - Diagnostics & Admin

LOT E - Polish & Extensions (Optional)

Documentation

Testing

Network Matrix

Validation Checklist

Troubleshooting

mDNS Discovery Issues

Streaming Issues

API Issues

Contributing

License

Resources

Support

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Contributors 2

Uh oh!

Languages

Packages