Trusted by industry leaders
Join the thousands of professionals using Treble to revolutionize their acoustic simulation workflows
Real World Evaluation for Speech AI
Sign up for our webinar introducing the Far Field ASR Leaderboard, developed with Hugging Face, including:
A new benchmark for evaluating ASR models under realistic far field acoustic conditions
Insights into how noise, reverberation, and distance impact real world performance
A walkthrough of how to submit models, interpret results, and use FFASR for evaluation and development
Featuring contributions from Hugging Face and Treble on advancing more reliable, real world speech AI evaluation.
The Treble Revit Plugin is live
The Treble Revit Plugin is now available, enabling a direct connection between Revit and the Treble Web Application. With a single click, project geometry can move from BIM to acoustic simulation without manual exports, rework, or intermediate file handling.
Hugging Face Collaboration to Advance Audio ML
Treble and Hugging Face are releasing Treble10, an open dataset of broadband RIRs and speech-convolved scenes from 10 furnished rooms, generated with the Treble SDK and available on the Hugging Face Hub for non-commercial research.
Our products
Treble Web Application
The Treble Web Application delivers next-generation acoustic simulation with a cutting-edge, cloud-native platform that combines speed, accuracy, and usability. Featuring a hybrid wave/geometrical acoustics engine, real-time collaboration, immersive auralization, and streamlined workflows. The Treble Web Application empowers professionals to design and optimize acoustic environments effortlessly at any scale.
Treble SDK
The Treble SDK is the first of its kind, offering a Python-based interface for scalable, high-fidelity simulations. It accelerates audio algorithm and hardware development, enabling synthetic data generation, virtual prototyping, and seamless integration with R&D workflows. It is trusted by leading tech companies to power next-generation audio experiences.
Wave-based FEM
One of the most effective methods for achieving accurate results in acoustic simulations is wave-based modeling. This method captures the complex physics of sound propagation, providing a level of detail that other approaches often miss.Experience how your spaces sounds
Generate perceptually authentic audio-visual listening experiences in seconds and share them with all stakeholders across the web with a single click.Seamless workflows
Our cloud-based platforms with zero hardware and software footprint plug into your 3D modelling tools enabling a new seamless and collaborative workflows, reducing time-to-results from weeks to minutes.Testimonials
Traditional acoustic modeling software would have taken 42 days to run one condition of the model, while Treble's raytracer only took an hour.
