Real world audio is difficult to reproduce
Audio systems need to perform across changing environments, devices, and listening conditions. Physical recordings and lab tests take time to set up, while simplified simulations can miss acoustic effects that influence performance.
For engineering teams, that means gaps in training data, slow design iterations, and uncertainty about the situations they haven’t tested. Exploring a new idea or investigating an edge case often means another round of measurements or prototypes.
What Treble Enables
Bring in your device, configure the acoustic scene, and connect simulation to your development workflow. Explore designs, generate data, and evaluate performance with less reliance on physical prototypes and recording campaigns.
Real world acoustics
Simulate
Import device geometry, define microphones and sources, and configure your environment. Model how sound interacts with the design and, for wearables, the wearer before investing in physical hardware.
High fidelity data at scale
Generate
Vary acoustic conditions to create labelled datasets tailored to your device and use case, with detailed metadata. Cover scenarios and edge cases that are difficult to record at scale.
Benchmark & evaluate
Validate
Use the generated audio in your evaluation pipeline to test hardware and algorithms across repeatable scenarios. Understand where performance holds up and where it needs work.
Accelerate your workflows
Optimize
Automate design studies and parameter sweeps to compare more configurations with less manual effort. Use the results to refine hardware and audio processing together.
Inside Treble
Use Treble’s Python interface to set up simulations, virtually prototype audio products, and generate acoustic data. Connect the results to your existing analysis and machine learning workflows.
Platform Capabilities
Treble combines advanced acoustic simulation capabilities into a single platform, enabling engineers to model complex environments, generate synthetic data, develop audio AI, and accelerate virtual prototyping through physics based simulation.
Model complex acoustic sources.
Advanced Source Modeling
Simulate directional sound sources, speech directivity, and CAD based source models with high physical accuracy.
Bring acoustic scenes to life.
Dynamic Acoustic Simulation
Simulate moving sources and listeners, changing orientations, and the acoustic conditions experienced along a trajectory.
Design for the devices that matter.
Device Simulation
Model device geometry and microphone arrays to understand how your hardware captures sound. Render audio specific to your device for design exploration, training, and evaluation.
Create immersive spatial experiences.
Spatial Audio Rendering
Generate high order ambisonics alongside binaural and multi channel rendering for realistic spatial audio applications.
Run thousands of simulations in parallel.
Simulations at Scale
Generate large scale acoustic datasets using cloud native infrastructure, extensive environment libraries, and automated scenario generation.
Build realistic virtual environments.
Scene Generation
Build acoustic scenes with environments, sound sources, and recording devices. Vary their configurations to generate diverse training and test scenarios.
Built for modern engineering workflows
From simulation to your development workflow
A hybrid engine built for acoustic accuracy at scale
Treble combines accelerated wave based simulation with advanced geometrical acoustics to deliver accurate results across the full frequency spectrum, balancing physical realism with computational efficiency.
Wave-based FEM
Explore geometry, material, and configuration variations with advanced parametric studies for optimized, precise acoustic designs and sensitivity analysis.
Geometrical acoustics
Efficiently models high frequency propagation in large and complex environments using advanced phased ray methods, preserving critical reflection and phase behaviour at scale.
Built for cloud scale simulation
Run simulations in parallel, automate design studies, and generate large acoustic datasets without maintaining your own simulation infrastructure.
Massive Parallel Compute
Launch thousands of simulations simultaneously.
Cloud Native
Access scalable compute without specialised hardware.
Enterprise Ready
Secure collaboration, project management, and scalable infrastructure built for engineering teams.
Improving multichannel speech enhancement through accurate room-acoustic simulations
Training on the high-fidelity dataset results in an up to 38% relative reduction in median word error rate compared to the lower-fidelity alternatives. These results show that augmentation with high-fidelity room-acoustic simulations directly translates into improved multichannel speech enhancement performance.
Synthesis of Room Acoustics for Speaker Distance Estimation
Explore how simulated acoustic data can support models that estimate a speaker’s distance from a microphone. This study presents a benchmark for generating room acoustics and evaluating their value for speaker distance estimation.
Room-Acoustic Simulations as an Alternative to Measurements for Audio-Algorithm Evaluation
Evaluate audio algorithms with simulated acoustic data. See how wave-based simulations closely match measured results.
