This blog post is part of an article series that takes a deep dive into the science behind Microphone Mist™ technology. Each piece originally appeared in audioXpress Magazine.
Imagine calling in to a meeting, ready for an important hybrid session, only to hear muffled voices and choppy silences coming from the room. It’s a familiar frustration for anyone who has joined a large meeting or learning space remotely. Despite decades of innovation, large-room audio conferencing still struggles to match the effortless clarity of conference calls in small rooms.
Why is this so hard to solve? The answer lies in the way legacy systems are designed — and the compromises they force on users. For years, the industry has relied on strategies that work “well enough” but never truly deliver.
The solution requires a bold rethink of traditional approaches. Nureva’s newest patented capability opens a new technology frontier in microphone array architecture. It significantly improves the installation process, integration with other AV equipment and the audio performance across rooms of diverse size and usage. The outcome is a foundational platform to deliver a suite of technologies that previously had not been possible for relatively low-cost and high-volume system deployments.
When it comes to rooms of up to 10 persons in size (Figure 1), the in-room and remote participants using a UC&C client application rarely think about the audio experience, assuming the equipment was installed and configured correctly. Installation requirements are relatively easy to meet, as the audio equipment is not operating at its technological limits. The use of small rooms is confined to closely spaced participants who are seated or standing at proximity to a wall, and one microphone and speaker device is all that is required as the demand on the technology is relatively low and easy to manage. There are no complex logic decisions or handoffs between audio devices to cover and manage the room. In short, small rooms are simple to install, configure and use — and they just work.
In a way, small-room microphone and speaker systems are like singles tennis. One player handles all aspects of game play on the court. No extra coordination or inter-player communication is required, which means it won’t break down during the game. The ball does not “hit the floor” because of mis-coordination with the other player, who was supposed to or was going to take it, or so they thought. Everything is (relatively) straightforward.
Figure 1: Small-room standard typical for up to 10 persons.
Unfortunately, audio systems for larger spaces rarely deliver this smooth experience. To continue my sports analogy, a larger room (supporting more than 10 and often over 20 people) is like playing a doubles racquet sport — perhaps tennis again — or a sport on an even bigger court such as volleyball. Everyone has seen a ball drop to the floor with the players saying, “I thought you had it!”
Well, the same thing occurs with larger spaces requiring audio coverage, when individual microphone devices try to coordinate operation through a digital signal processor in the room. Which device has the active talking participant? Who has the ball? Complex logic strategies and plans are developed to try and cover all situations, but — as in dynamic game play — the room situation is fluid and complex, and the best strategy and decisions don’t always work as intended.
This is such a consistent problem that participants go into larger rooms expecting poor-quality audio, based on past experiences. Audio performance decreases further when the room cannot be limited to a single-use scenario, such as a basic conference table with seated participants. A collaboration environment such as a classroom or hybrid multiuse room creates difficult requirements for the system to meet. If the classroom is configured for in-class and remote learning simultaneously, the students and teacher may be seated or moving concurrently at any location in the room.
With some systems, attempts to solve the problem of inadequate microphone audio pickup are made by anticipating the coverage areas and installing microphone arrays into specific parts of the room. For every coverage area required, a microphone array will need to be installed and configured (Figure 2).
As we can see, it is very challenging to provide adequate microphone coverage for all situations in the room. Inevitably, there will be gaps in coverage, which then translates into a subpar, hard-to-hear experience for participants at the far end of the UC&C call — audio performance that is well below the small-room experience. Furthermore, as speech-to-text models, AI facilitators and note-taking tools are incorporated into the audio/video sessions of today and tomorrow, the issues of such a system become even more glaring. When we consider factors such as participants speaking loudly and softly, and discussions taking place across the room — where audio pickup is challenging to track and maintain — we can expect inaccurate results, missed words and poor talker identification. In effect, “dropping the ball” can be an all-too-common experience.
Issues with large-room mic coverage start when the audio peripherals (microphones and speakers) are installed. Many rooms present a challenge in placing the critical microphone arrays into the optimal locations. Uncontrollable factors to work around can include HVAC, room fixtures, ceiling and wall materials, and, eventually, cabling. The larger the room, the more microphone arrays that are required to cover the space, increasing installation complexity considerably. Unlike sport players, who can move, adapt and reconfigure their court placement and actions dynamically, these systems are fixed and inflexible.
Once installation is complete, the job of the technician is just getting started. The audio system needs to be configured and calibrated for each specific microphone array coverage area and for the plurality of audio parameters. This applies to both the individual microphone array and the digital signal processor. The system shown in Figure 2 is a collection of disparate devices that are manually integrated and configured to operate as individual devices. Each must manage a separate section of the room while the digital signal processor tries to switch and/or blend the separate audio streams to send to the remote participant of the UC&C session.
In effect, each microphone array acts like a player on a team. Audio targeting and management becomes a complex negotiation-based decision-making exercise in an attempt to cope with the ever-present issues of hybrid rooms.
And if the room is divisible and/or reconfigurable, the demands on the technology are even greater. Inevitably, when you have separate devices (players) working individually and together at the same time, performance is compromised — forming unintended gaps in coverage, inconsistent audio pickup and low-quality UC&C streamed performance.
Figure 2: The old paradigm is that for every coverage area required, a microphone array will need to be installed and configured.
The best sports teams work as one, where the players can seamlessly perform as if they were one superplayer covering the court, scoring points with no balls being dropped. Reaching that magical state of team play rarely happens in sports, and it rarely happens in large-room audio conferencing systems either.
Nureva has changed the game, as discussed in my 2023 audioXpress article on the pioneering patent-pending unified coverage map, a feature of Microphone Mist technology. It addresses the challenges presented by large spaces (Figure 3) by taking this superplayer approach. Full-room coverage is accomplished using virtual microphones distributed throughout the room. There are no disparate coverage zones to manage and no gaps in coverage or missed situations with poor microphone negotiation and coordination — which means “no balls hit the floor.”
Installation, configuration and operation have been simplified to the point that a hybrid large room performs like a small room. The advantages become clear when a single coverage map is utilized for complex hybrid rooms supporting a variety of usage scenarios.
Figure 3: This shows an example of a unified coverage map.
Nureva set a goal to develop microphone array architecture built for the multifaceted audio spaces of today and tomorrow. It needed to be easy to install and able to auto-calibrate, going beyond basic echo canceler performance, to adapt its configuration and coverage area to support diverse room dimensions and usages.
Simple to operate, it had to be robust and scalable to stay relevant, while supporting deployment of next-generation technology such as the patent-pending unified coverage map. To go back to our sports analogy, the microphone system needed to operate like a single superplayer on the court, providing high-quality and robust audio performance in a simple and adaptable solution.
To accomplish this, the group of individual microphone arrays — which worked like a multi-player team with their own internal logic decision processes coordinated by a digital signal processor (DSP) — needed to be replaced. Instead, all the available individual microphone elements in each Nureva microphone array are combined into a single large 3D aperture array as shown in Figure 4 (the single court superplayer strategy), which is managed holistically by the audio processor.
If the individual microphone arrays are allowed to make and drive the room audio decisions on their own, there is opportunity for poor choices and missed audio data (i.e., “balls dropped on the court”). With the combined large 3D aperture array arrangement, all the microphone elements can be utilized from a pool as needed, regardless of the individual microphone array locations in the room. The audio processor will have access to all the audio, parametric and metadata parameters, with the highest quality audio stream, to make the best targeting, focusing and streaming processing decisions.
Legacy systems, with their underperforming approaches that focus on installing and mapping a single array to a specific coverage zone in the room, as illustrated in Figure 2, choose a specific microphone array based on the talking participant’s approximate location in the room (targeting parameters). At the same time, the other microphone arrays in the room are idled, leaving unused microphone elements that could be used to increase audio system performance. This is a poor asset utilization strategy and would be considered wasteful and inefficient in any other environment. The number of microphone elements that can be used for targeting determination and focusing for audio streaming is a limitation of the smaller array aperture of the individual device, which constrains gain structure, targeting accuracy and signal-to-noise performance parameters.
The problems of legacy systems are further exacerbated by the need to switch between physical microphone arrays as the talking participant moves throughout the room, a common challenge in hybrid spaces. To overcome the limitations, overly sophisticated targeting and audio management techniques are incorporated into both the individual microphone arrays and DSP system to coordinate the dynamic situation in the room. This usually shows up as suboptimal performance at the far-end (remote participant) side of the UC&C conference session, leading to the common experience of large-room performance that’s below the small-room standard.
Figure 4: Nureva’s patented single combined large 3D aperture array covers the complete room, wall to wall, all the time.
In contrast, Nureva’s patented single combined large 3D aperture array (US Patent 18/116,632) covers the complete room, wall to wall, all the time. It uses a subset — or all — of the available microphones in the room in real time dynamically, based on the talking participant’s location. It is obviously easier to manage a single microphone array compared to multiple individual and standalone microphone arrays, reproducing the small-room advantage in a large space.
This wide aperture array covers larger distances than would normally be possible, with the wide separated mounting locations — often on multiple mounting planes (walls and/or ceiling, across the room x, y, z dimensions) — and range of heights and orientations. The system is able to deliver more precise targeting of sound sources (talking participants) and x, y, z locations in the room, due to the microphone array system and the room coordinates as established during installation and calibration.
The proprietary algorithms of the single combined large 3D aperture array determine the individual microphone selection based on the distributed virtual microphones within the coverage grid. These algorithms optimize audio targeting, pickup and streaming to maximize the gain structure and signal-to-noise ratio (SNR) while minimizing room artifacts such as reverberation and echo.
All of this happens while the combined large 3D aperture microphone array tracks a specific moving sound source (the talking participant) or switches to another sound source (a second talking participant). All the microphone management is handled under one processing algorithm for seamless integration and remapping of the individual microphones in real time as needed, as any normal small aperture array would do within its operating parameters.
The next-generation technologies — specifically the patent-pending unified coverage map and patented position-based gain control (US Patent 11,635,937) built into Microphone Mist technology — utilize the foundational combined microphone array architecture. Instead of complex switching from microphone array to microphone array, the logic and audio processing functions are blended to provide the highest quality uninterrupted audio stream. By daring to reevaluate the problem and open up this new technological frontier, Nureva has developed large-room audio solutions that can now perform and deliver on the small-room audio standard that had been elusive and unobtainable till now.
To make this leap, Nureva needed fresh disruptive thinking. In applying this, the team focused on identifying and solving for problems that are within our control while looking for clear examples that work well today to riff off. For example, room size, décor, material choices, HVAC installation and available mounting locations of equipment are usually determined without a single thought given to the audio performance in the room. Audio technology companies cannot control those aspects; however, what can be controlled is how the equipment is integrated to work in the room within those constraints.
Nureva’s proprietary calibration procedure measures and builds the single combined large 3D aperture array at system start-up (power-up), which establishes the geometry of the large 3D aperture array in a space to a set of reference system coordinates (0x, 0y, 0z), as shown in Figure 5. This solves for the matter of array location in the room relative to room dimensions and for sound source (talking participant) location (P1x, P1y, P1z) in the room relative to those dimensions, which creates a robust platform to enable sound source location and camera steering applications (as shown in Figure 4).
Since the combined large 3D aperture array derives and configures its aperture geometry each time the system is powered up, any changes in microphone device locations and/or room configurations can be accounted for in real time, allowing the system to adapt to multi-use hybrid rooms and, in more challenging cases, divisible room deployments.
It does not matter if the microphone element is part of a separate microphone and speaker bar or microphone array pod. Each microphone array enclosure contains individual microphone elements (transducers) that can range from a few to hundreds, depending on the design used. If the microphone array modules are connected to the audio system, they can be utilized and mapped for location (x, y, z) to form the combined large 3D aperture array, as shown in Figure 5.
Figure 5: The autocalibration procedure dynamically measures and locates all transducer devices in the 3D space and builds the large 3D aperture array, defining the aperture geometry.
A loudspeaker in the system is chosen as the reference loudspeaker with location (0x, 0y, 0z) coordinates, which by default establishes the room reference coordinates for all other system proprietary logic and process functions. The height of the reference loudspeaker can be set to a default value for the standard recommended mounting height.
At power-up of the system, or when prompted by a user, calibration of the array is initiated. The autocalibration procedure dynamically measures and locates all transducer devices in the 3D space and builds the large 3D aperture array, defining the aperture geometry. Essentially, the system builds a single microphone array device for the audio processor to manage, just like in a small room.
The proprietary calibration signal consists of wideband white noise sequences that are sent first from a reference loudspeaker (Figure 5, S1) location to all individual microphone transducer elements in the room (illustrated by the orange arrows) that are enabled and integrated into the audio system. The individual microphone elements in the arrays are measured and accounted for in their orientations and positions on the walls and/or ceiling.
Then the calibration signal is transmitted from loudspeaker (Figure 5, S2), as illustrated by the red arrows, to the microphone elements in the arrays, completing the measurement process for the system illustrated. If more loudspeakers are present in the room, the process is repeated as needed for each loudspeaker in the system. A substantial set of system parameters are measured and recorded during the calibration measurement phase of the process.
The wideband white noise calibration signal measures:
Once the brief autocalibration measurement process is completed, a few minutes after start-up, a proprietary geometric solver processor constructs the combined large 3D aperture array in 3D space, forming the single large microphone array. All individual microphones in the system are referenced to a set of geometric aperture coordinates in the room, allowing for accurate and stable sound source targeting and focusing capabilities.
The 3D placement of the microphone arrays will ultimately inform the unified coverage map (Figure 3), which is automatically generated. Full-room coverage is achieved with virtual microphones dispersed into the room, eliminating the chance of technician configuration setup and calibration errors. Put another way, the calibration compensates for the placement and orientation of the microphone array enclosures, allowing for installation flexibility, automatic room configuration and support for troublesome rooms.
The system operates as a single microphone array, just like in a small room. No complex installation and configuration is required. As the room is reconfigured or the usage changes, the audio system instantly adapts, requiring no specialized personnel or knowledge. A simple power cycle or user-initiated calibration will have the system up and running again in minutes. Installation becomes a simple matter of mounting the units and powering them up. The system will build and configure itself automatically.
The realized system design and complexity was brought to bear on the real problems that needed to be solved rather than refining existing ineffective strategies that never really worked well to begin with. It turns out small rooms had the right idea for a large-room solution all along.
To close out with the sports analogy, we’ve enabled large rooms to play the small-room, single-player court game. Our technology overcomes the deficiencies of the multi-player court game so no ball will drop between players hesitating anymore. People working or learning in large rooms can now enjoy the small-room audio performance standard. The combined array design, found in Nureva’s HDL410 system and HDX audio series (coming soon), establishes a new technology frontier built on a significant architectural foundation. It makes it possible to deploy capabilities and solutions that were previously not possible — more of which we can look forward to in the future from Nureva!