Wavez viewer showcase
Wavez is a real-time 3D scientific simulation and data visualization application.
Simulations export particles, uniform-grid scalar fields and triangle meshes to
Wavez's WVZ1 .wvz capture format. The interactive viewer lets you explore
those recorded layers together, navigate their sample times and export images
or video.
Comparison — same simulation in Python
Wavez is in pre-1.0 development. Scientific Selection and Analysis tools are experimental and have not yet been fully implemented or comprehensively validated. Independently verify results before relying on them for scientific conclusions.
- View particles, volumetric fields and meshes together through the Particles / Fields / Meshes interface.
- Inspect scalar patterns with scientific colormaps and legends, field ranges, opacity controls and slices.
- Play and seek recorded time; move freely through the scene or use camera orbit.
- Export viewport images and deterministic video of a chosen time range and view.
Source-based probing is available with documented precision and verification limits. Stored samples, derived measurements and rendered colors have different meanings. The Technical Documentation collects their contracts and limitations. For everyday controls, see the user handbook.
- Run a provided Linux x86_64 package, keeping
WAVEZ.x86_64andWAVEZ.pcktogether, or follow the source-build instructions. - Export your simulation's arrays with Python's
SceneCaptureto a new.wvzfile. Small particle, field, mesh and combined examples are included. - Choose File → Open. The capture opens paused while preparation continues.
- Explore the scene, adjust its presentation, then export an image or video.
Getting Started covers installation, complete Python examples and simulation integration.
| Workflow | Requirements |
|---|---|
| Packaged viewer | Graphical Linux x86_64 desktop and working Vulkan GPU/drivers. The Godot runtime is included; Python and a separate Godot installation are unnecessary for viewing. |
| Python export/integration | Python 3.10+, NumPy ≥1.23 and this repository's exporter. Zstandard compression is optional. |
| Optional tools | MP4 needs external FFmpeg with libx264. Experimental Selection/Analysis needs external python3 and NumPy; isosurface extraction additionally needs scikit-image 0.26.x, on Python 3.11+. |
| Source development/build | Tested with Godot 4.7.2; Linux packaging also needs Python and matching export templates. |
Use the installation guide for dependency setup. Forward+ Vulkan is the sole supported renderer.
Linux x86_64 is the currently validated platform. Plan for a modern multi-core 64-bit CPU, a Vulkan-capable GPU with working drivers, and enough free storage for captures and exports. The tiers below are provisional planning guidance, not certified minimums or performance guarantees.
| Tier / intended workload | System RAM | GPU memory | Storage |
|---|---|---|---|
| Minimum estimate — modest captures and light workloads | 16 GB | About 6–8 GB | SSD |
| Recommended — substantial particle/mesh data and moderate fields | 32–64 GB | 8+ GB | Fast SSD / NVMe |
| High-end — 192³+ fields, multiple channels and longer prepared coverage | 64–128+ GB | 16+ GB where practical | Fast NVMe, substantial free capacity |
Development and validation have used Fedora 44 / KDE, Ryzen 7 5800X, 48 GiB RAM, RTX 3070 Ti 8 GB and Godot 4.7.2 / Forward+ Vulkan. That workstation is evidence of tested use, not certification of the other tiers or every capture.
More RAM can extend prepared playback coverage and reduce re-preparation. CPU throughput and memory bandwidth help field preparation; bounded workers adapt to available headroom. GPU speed and VRAM affect rendering capacity, while faster storage helps uncached reads. VRAM does not replace the system-RAM capture cache. Costs vary with resolution, channel/particle count, recording cadence, playback speed and cache settings; preparation can still limit playback on a fast GPU.
Public read/clone access to the source and documentation does not require an invitation. Project development remains owner-directed: only authorized collaborators may push, administer or merge changes, and external contributions require explicit owner arrangement. No public contribution intake is offered. See the Development and Maintenance Policy.
Original Wavez material is source-available under the PolyForm Noncommercial License 1.0.0, not an OSI-approved open-source license. Commercial use requires a separate agreement. Separately licensed components retain their third-party terms, and the name and artwork follow the branding policy.