RadianceKit

Product Introduction

1. Definition: RadianceKit is a native macOS desktop application for 3D Gaussian Splatting. It is a complete, self-contained photogrammetry software solution that enables users to create photorealistic 3D scenes and models directly from photographs or videos using the computational power of their Mac's Apple Silicon GPU.
2. Core Value Proposition: RadianceKit provides a fast, private, and accessible workflow for 3D Gaussian Splatting reconstruction, eliminating the need for complex cloud services, Python environments, or CUDA hardware. It transforms standard 2D media into interactive 3D digital assets entirely locally on a Mac, making high-fidelity 3D scanning available to a broader audience.

Main Features

1. Apple Silicon Native GPU Compute: The core engine is built for and runs natively on Apple Silicon (M1, M2, M3, M4 chips). It leverages the Metal graphics API to perform all Gaussian Splatting training and rendering calculations directly on the Mac's GPU, ensuring optimized performance, speed, and power efficiency for the 3D reconstruction pipeline.
2. Simple and Expert Modes: RadianceKit caters to different user skill levels. Simple Mode provides a streamlined, one-click interface for quick results, ideal for beginners. Expert Mode offers a professional three-panel layout with a project navigator, interactive 3D viewport, and a detailed inspector panel, granting full control over training parameters, live loss curve monitoring, and advanced export options for technical 3D scanning workflows.
3. Comprehensive 3D Gaussian Splatting Workflow: The application integrates the entire process into a single interface. Users import photos or video, the app automatically aligns camera positions using photogrammetry algorithms, trains the Gaussian Splatting model to create millions of 3D ellipsoids (splats) representing the scene, allows for interactive preview and editing (like artifact removal), and finally exports the finished 3D reconstruction.
4. Interactive Gaussian Editor: Includes in-app tools for post-training refinement. Users can select and delete unwanted regions, floating artifacts, or background elements directly in the 3D viewport using a brush tool. This allows for precise cleanup and fine-tuning of the 3D scene without needing external 3D editing software.
5. Multiple Export Formats and Sharing: Supports exporting final scenes in six industry-standard formats: PLY, Compressed PLY, SPZ, glTF, .splat, and SOG. Additionally, it can generate orbit videos and create self-contained, interactive HTML web viewers for easy sharing of 3D scenes online without requiring recipient software.

Problems Solved

1. Pain Point: RadianceKit solves the barriers of complexity, cost, and privacy associated with traditional 3D reconstruction (NeRF, photogrammetry) and cloud-based Gaussian Splatting services. It removes the need for command-line interfaces, Python environment setup, GPU clusters (CUDA), and the requirement to upload sensitive photos/videos to external servers.
2. Target Audience: The product is designed for 3D Artists & Game Developers needing quick asset creation, Product Designers & Architects requiring rapid 3D prototyping from physical objects or spaces, Real Estate & Tourism Professionals for virtual tour content, Content Creators & YouTubers for 3D content, and Mac Enthusiasts & Hobbyists interested in computational photography and 3D technology.
3. Use Cases: Essential for creating 3D product visualizations for e-commerce, scanning real-world objects for AR/VR applications, documenting architectural spaces or cultural artifacts, generating 3D environments for game development, producing interactive content for websites, and educational projects involving computer vision and 3D reconstruction.

Unique Advantages

1. Differentiation: Unlike cloud-based NeRF or photogrammetry services (e.g., Luma AI, Polycam), RadianceKit offers 100% local processing, guaranteeing data privacy and eliminating ongoing subscription or upload costs. Compared to open-source Gaussian Splatting repositories or Python-based workflows, it provides a polished, integrated GUI with no setup complexity, making the technology accessible to non-developers.
2. Key Innovation: The key innovation is the complete, professional-grade 3D Gaussian Splatting pipeline packaged as a native macOS application optimized for Apple Silicon. By harnessing Metal for GPU acceleration and integrating photogrammetry (camera alignment) and splatting training into a unified, user-friendly app, it democratizes a cutting-edge 3D reconstruction technique that previously required significant technical expertise.

Frequently Asked Questions (FAQ)

1. What exactly is 3D Gaussian Splatting and how does RadianceKit use it? 3D Gaussian Splatting is a modern technique that represents a real-world scene as a collection of millions of tiny, colored 3D ellipsoids (called "splats") reconstructed from ordinary photos or video. RadianceKit automates this entire process: it analyzes the input media to calculate camera positions, trains the Gaussian model on your Mac's Apple Silicon GPU via Metal, and lets you interact with the resulting photorealistic, explorable 3D scene within the app.
2. Do I need a powerful Mac or specific software like Python or Blender to use RadianceKit? No. RadianceKit is a self-contained macOS application requiring no external tools, command lines, or Python environments. It requires an Apple Silicon Mac (M1, M2, M3, or M4) running macOS 12 (Monterey) or later, with 16GB of RAM recommended for optimal performance on larger scenes. All processing happens natively on your machine using the GPU.
3. How does RadianceKit compare to other 3D scanning apps like Luma or Polycam for Mac? RadianceKit differs primarily in its processing location and technology. While apps like Luma or Polycam often rely on cloud servers for reconstruction, RadianceKit performs all Gaussian Splatting training and processing locally on your Mac, ensuring complete data privacy and no internet dependency. It focuses specifically on the Gaussian Splatting method, offering a streamlined workflow for this particular technique compared to broader photogrammetry tools.
4. What is the pricing model and is there a free trial? RadianceKit is available as a free download from the Mac App Store, which includes a fully-featured 3-day trial of all its capabilities. After the trial period, users can unlock the complete version with a one-time purchase of $7.99. There are no subscriptions or recurring fees.
5. Can I import footage from a 360-degree camera like an Insta360? RadianceKit currently does not directly ingest 360-degree equirectangular or dual-fisheye (.insv) files, as the camera alignment process expects standard perspective camera models. The recommended workaround is to use the camera's companion software (e.g., Insta360 Studio) to reframe the 360 footage into standard, flat perspective video clips, which can then be imported into RadianceKit for 3D scene reconstruction.