Outer Echo is a research-driven company grounded on scientific innovation and fueled by a deep passion for sound and music. We engineer technologies that prioritize sound quality, physical coherence, and computational efficiency, with a clear focus on real-time applications.
Much of our work is rooted in Physical Modeling: the simulation of physical phenomena applied to sound, music, and acoustic environments. This foundation shapes how we think about synthesis, spatialization, interaction, and immersion.
Besides pushing the boundaries of well-established techniques we are developing a portfolio of proprietary technologies, having obtained patents in major international jurisdictions. Although these technologies are developed and get validated through our own engines and products, they are not limited to a single use case or implementation: they are designed from the outset to be modular, adaptable, and extensible.
Physical Modeling
Physical Modeling lies at the core of Outer Echo’s approach to audio technology. By physical modeling, we refer to the numerical simulation of physical phenomena—such as wave propagation, coupled resonant behavior, and interactions between sound sources, structures, and environments—applied directly to sound synthesis and spatial rendering.
Physically-modeled sound rendering is inherently challenging, as it requires:
- Mathematically sound formulations
- Stable and efficient numerical methods
- A deep understanding of both physics and human perception
When physical modeling is designed carefully and implemented efficiently, it stops being exotic and becomes uniquely powerful. Our philosophy is simple: the emulation of complex physical behavior should not result in complex tools. Through mathematical abstraction, optimization, and perceptually informed design, we couple time, frequency, and modal representations to translate acoustic phenomena into lightweight, flexible simulation technologies that are:
- Efficient enough for real-time execution
- Robust under dynamic and interactive conditions
- Practical to integrate into production environments
Our commitment to physical modeling provides a natural ground for spatial audio in media, interactive applications, games, and XR. As we do not limit spatial audio to merely placing sound sources in three-dimensional space, standing on physics-supported representations enable us to effectively introduce unconventional physical modeling abstractions that allow the efficient simulation of diverse wave propagation and diffusion phenomena, to support:
- Real-time navigation of virtual acoustic environments
- Dynamic interaction between sources, listeners, obstacles, and coupled spaces
- Extremely low latency for immersive and XR applications
6D Spatial Audio
At Outer Echo we develop groundbreaking physical-modeling 6D Spatial Audio technologies that enable unprecedented fidelity without the typical efficiency compromises. By 6D Spatial Audio we understand six-degree-of-freedom (6DOF) immersion that goes beyond standard 3D audio—where sound sources are merely placed around a stationary listener. Thanks to approaching spatial audio from a physical modeling perspective, it is possible to enable true volumetric movement where the entire acoustic environment responds coherently to listener and source motion, behaves according to real-world acoustic principles and sustains a seamless and believable sense of space.
Instead of relying on brute-force numerical simulations and/or interpolated convolution of pre-computed impulse responses, we combine physics-based abstractions and perceptual compression techniques to construct a self-balancing physical model of the acoustic environment, defined from first principles. This allows to dynamically regulate the trade-off between perceived spatial resolution and required computational cost while faithfully simulating, in real time and low latency, diverse acoustic phenomena such as:
- Reflection
- Diffraction
- Diffusion
- Reverberation
- Large-Scale Sound Propagation
- Acoustic Coupling
- Emitter and Receiver Directivity
- Occlusion
- Doppler
- Near-Field Effects
- Head-Related Transfer Function (HRTF)
Our framework enables the creation of virtual acoustic environments—where multiple obstacles, sources and listeners are in continuous movement and rotation—that can be flexibly modified and freely navigated, and empowers artists with the ability to apply updates instantly without baking. To achieve this, we take advantage of integrating two key proprietary technologies on which Outer Echo has spent years of research and development: Adaptive Wavefront Resolution Auralization (AWRA), and Acoustic Resonator Graph Optimization (ARGO).
AWRA — Adaptive Wavefront Resolution Auralization
AWRA (pronounced “aura”) lies at the center of our sound rendering pipeline. In a dynamic scene, hundreds (if not thousands) of individual wavefronts—such as those arising from reflection, diffraction, diffusion, and occlusion caused by multiple moving sound sources and obstacles interacting with the acoustic environment—must be processed simultaneously. In addition, numerous reverberation and propagation multi-directional components contribute to the final auditory result. Each of these elements requires dedicated, time-varying processing under continuously changing conditions, making the overall computation highly demanding.
To address this complexity we devised AWRA, a highly efficient time-domain rendering algorithm able to dynamically manage computational resources by way of context-aware and content-aware real-time adaptive strategies. On the one hand, AWRA considers the listener’s context within the virtual acoustic environment—such as spatial position and scene dynamics. On the other hand, it analyzes the audio streams associated with each wavefront in real time. Based on this combined analysis, the system performs high-frequency reassignment of processing resources within a multi-resolution auralization module, ensuring perceptual quality while maintaining computational efficiency.
ARGO — Acoustic Resonator Graph Optimization
ARGO is the core technology responsible for physics-based modeling of the acoustic environment, operating in tight coordination with AWRA. Given a physics-consistent representation of a large scene provided by the artist —such as one derived from detailed geometric primitives and material properties, from perceptually motivated frequency-dependent decay parameters, or from other hybrid or abstracted acoustic-structural descriptors— ARGO constructs a large-scale physical model comprising a large number of coupled acoustic resonators organized in a dynamic, graph-like structure that can be edited both at design time and during run-time.
This graph-based physical modeling abstraction not only guarantees numerical stability and provides time-varying contextual information to AWRA, but also enables run-time optimization driven by scene dynamics and perceptual constraints. Run-time iterative optimization is one of the key capabilities of ARGO, which automatically determines physical modeling reorganization strategies to dynamically reallocate computational resources at audio rate in what effectively constitutes a self-balancing physical model of the acoustic environment.
3D Ambisonics
One of the well-established techniques exploited in some of our 3D Spatial Audio work is Higher-Order Ambisonics (HOA). Ambisonics provides a powerful and widely supported framework for representing surrounding sound fields in 3DOF (three-degrees-of-freedom) immersive applications. At Outer Echo, we do no treat Ambisonics as a static format, but as a flexible spatial representation capable of supporting advanced real-time processing.
Our Ambisonics portfolio comprises highly efficient implementations of modular Encoding, Rotation and Decoding pipelines up to HOA Order-7, including diverse encoding and decoding configurations under custom beamforming patterns for dynamic spatial reassignment, and supporting fast modulation schemes on hundreds of processing channels by means of platform-specific vectorization routines. By extending the efficiency and flexibility of Higher-Order Ambisonics, we enable dense, detailed spatial 3DOF scenes that remain practical in real-time interactive contexts. Our current Ambisonics work is focused on:
- Increasing spatial resolution without prohibitive computational cost
- Improving robustness under fast listener rotation and dynamic scenes
- Enabling time- and frequency-domain transformations in signal chains involving multiple adaptive beamformers
- Integrating Ambisonics with physically-based acoustic models