15 August 2023

Full bandwidth wave-based simulations using Treble

Abstract

The study measured the capabilities of Treble’s wave-based solver in a medium-sized room of 80 cubic meters. The solver was tested for full-bandwidth up to 8kHz and both mono and binaural room impulse responses (RIR) were measured using a directive source and a KEMAR mannequin at various positions. The simulations were carried out on a cluster with six Nvidia A100 GPU cards running in parallel. Results are compared and presented for Energy Decay Curves and derived acoustical parameters, reverberation time T20 and early decay time, for the mono RIR. The binaural RIR results are compared both in the frequency domain and in the time domain.

Download full paper

What you'll find

Full bandwidth wave-based virtual acoustics with directive sources and receivers.
Simulated binaural rendering using 16th order ambisonics using an virtual spherical microphone array.
Simulated results showing an excellent match with measurements up to 8kHz.

Read full paper

For more information about our technology, you can read our knowledge base.
For more information about the importance and future of acoustic engineering, read our insight paper.

You can also find more validation studies such as:
-Treble simulation of empty rectangular rooms with porous absorbers

Webinar: A new era for audio AI data

Join our webinar on February 26th unveiling new Treble SDK capabilities for advanced voice processing and audio AI, including dynamic scene simulation, a new source receiver device for own voice and echo cancellation scenarios and powerful tools for large scale synthetic data generation. Industry leaders in audio technology and voice AI will join to discuss how these advancements elevate product performance and accelerate research and development.

03 February 2026

Meet Treble at ISE 2026

We will be at ISE 2026 in Barcelona on February 3-6 where wewill showcase the Treble Web Application for high-fidelity room acoustic simulations, giving visitors a hands-on look at our streamlined cloud-based approach to acoustic prediction and auralization.

28 January 2026

Synthetic realism in speech enhancement: Training models for the real world with ai-coustics

Tim Janke, Co-founder and Head of Research at ai-coustics, examines why voice AI systems struggle outside controlled environments and how training data is often the limiting factor. The article introduces synthetic realism as a data driven approach to teaching models what truly matters in real acoustic scenes, from distance and geometry to room acoustics. It also shows how physics based simulation with the Treble SDK enables spatially accurate training data that translates into more reliable speech enhancement and voice agent behavior in production.

Back to insights

Full bandwidth wave-based simulations using Treble

Abstract

What you'll find

Recent posts