Why it is worth attention

It rescues a discontinued premium webcam with full manual camera control, low-latency processing, and a virtual camera system—all built on open hardware and software.

Who it is for

Owners of the Opal C1 webcam
macOS users seeking advanced webcam control (exposure, focus, white balance)
Open-source enthusiasts interested in DepthAI and camera pipeline hacking
Video conferencing users wanting a virtual camera with background blur

Use cases

Using the Opal C1 as a daily driver webcam with manual settings and background blur
Replacing Opal's discontinued Composer software with a maintained alternative
Integrating the camera as a virtual device in Zoom, Meet, or FaceTime
Learning about USB camera control, ISP downscaling, and Metal-based effects

Strengths

Full hardware control (exposure, focus, white balance) via direct USB communication
Low glass-to-screen latency (~45 ms at 1080p30) thanks to on-camera ISP downscaling
Background blur using Metal and Vision segmentation, with optional depth‑graded mode
Works as a system‑level virtual camera for any app when installed and signed

Considerations

Requires Apple silicon Mac on macOS 26+ with Xcode 26 (very new and specific system requirements)
Building from source involves multiple steps and dependencies (depthai‑core, XcodeGen)
Only supports the Opal C1 webcam; virtual camera requires a signed/notarized build

README quick start

Open Opal

A native macOS app for the Opal C1 webcam. Opal has discontinued the C1 and its Composer software; this project keeps the camera working.

The C1 is built on Luxonis' DepthAI platform, so it can be driven with open-source tools. Open Opal talks to the camera directly over USB.

Features

Exposure: auto, or manual shutter and ISO, with EV compensation and AE lock
Focus: autofocus modes, manual lens position, click anywhere to focus
White balance: presets or manual Kelvin
Anti-banding for 50/60 Hz lighting
Sharpness, denoise, brightness, contrast, saturation
4K / 1080p / 720p, up to 42 fps
Background blur, rendered in Metal with Apple's Vision segmentation
Optional exposure metering on your face instead of the whole frame

Frames are downscaled on the camera's own ISP before crossing USB, which keeps glass-to-screen latency around 45 ms at 1080p30. The toolbar shows the live number.

Building

Needs an Apple silicon Mac on macOS 26 or later, Xcode 26, and brew install cmake ninja xcodegen.

./scripts/bootstrap.sh     # fetches and builds depthai-core
./scripts/fetch-models.sh  # downloads the Core ML depth model
xcodegen generate
open OpenOpal.xcodeproj    # then build & run from Xcode (⌘R)

To build from the command line and install to /Applications:

xcodebuild -project OpenOpal.xcodeproj -scheme OpenOpal \
  -configuration Release -derivedDataPath build/DerivedData build
cp -R build/DerivedData/Build/Products/Release/OpenOpal.app /Applications/

Use Release builds for day-to-day use — Debug builds noticeably stutter in UI animations.

The hardware

Little of this is documented elsewhere, so for the record:

The C1 is a Luxonis DepthAI device: an Intel Movidius Myriad X VPU (USB VID 0x03E7) paired with a Sony IMX582 48MP sensor (module name LCM48) and a real autofocus lens.
The sensor has no native 1080p mode. Its readout modes are 3840×2160 (2–42 fps), 4000×3000 (2–30 fps), and 5312×6000 (1–10 fps). That's also why the app caps out at 42 fps.
Full 4K NV12 at 30 fps is ~370 MB/s, which saturates USB 3 and costs ~300 ms of latency. So the app always captures 4K and downscales on the camera's ISP before the frame crosses the wire — that's the ~45 ms figure.
The sensor is mounted upside down in the housing; the stock firmw

Description

Modernized Opal C1 software for macOS with an improved bokeh model

Related repositories

Similar projects matched by category, topics, and programming language.

Blaizzy

Featured

nativ

Nativ is a native macOS app that lets you run AI models locally on Apple silicon, offering chat, model management, performance analytics, and an OpenAI/Anthropic-compatible API server.

AI & Machine LearningAI Vision & Audio

985

uzairansaruzi

Featured

hermex

Hermex is a native SwiftUI iPhone app that lets you control a self-hosted Hermes AI agent directly from your phone, with no subscriptions, tracking, or third-party relay.

AI & Machine LearningLarge Language Models

941

simonw

Featured

speech-analyzer-cli

A macOS command-line tool that uses Apple's on-device speech APIs to transcribe prerecorded audio into plain text, JSON, JSONL, SRT, or WebVTT.

Developer ToolsCLI & Terminal