Microsoft Technology Licensing, LLC

Kevin Juho Venalainen

Christopher Ian Borowski

Dana William Bourke

Brian Walter Aznoe

Stanley Thomas Birchfield

Avronil Bhattacharjee

Arthur Baker

Chun Beng Goh

Jani Samuli Kuivalainen

Eric Lee Middleton

Sylvain Angrignon

Eric Middleton

Techniques performed by a data processing system for operating a speaker disposed within a sealed enclosure herein include obtaining a first input signal to be output by the speaker; determining a first volume level associated with the first input signal; selecting a first Linkwitz Transform and a first Multiband Compressor (MBDRC) from volume-dependent configuration data based on the first volume level; generating a first intermediate signal by applying the first Linkwitz Transform to the first input signal to increase a low-frequency response of the speaker; generating a first output signal by applying the first MBDRC to the first intermediate signal by compressing the at least a portion of the first intermediate signal; and driving the speaker to produce first audio output using the first output signal.

Increasing low frequency extension for microspeakers using a volume dependent linkwitz transform and multiband compressor

A frequency-dependent signal limiter using a paired combination of a bandpass filter (or 'selection filter') and an attenuating equalizing filter (or “cutting filter”) limit the amount of distortion generated by an audio output device in an acoustic rendering system. A bank of one or more of these pairs in series can be used to limit selected frequencies identified as primary causes of distortion in the audio output device to specified levels that avoid distortion.

Limiting distortion of acoustic rendering

A host device is connected via universal serial bus (USB) connections to one or more external devices with respective speakers. The host device synchronizes playback of audio through the speakers of the USB-connected external devices by requiring devices to implement a deterministic latency between receipt of a start playback request, and requiring devices to slave their audio clocks to USB start-of-frame tokens, and determining a buffered start time for starting data transmission to the respective USB devices. The buffered time equates to the current time of the USB host plus an additional time buffer. The host device waits until the buffered start time, and then starts transmitting the audio stream to the external devices, which, in turn, begin playing the audio data on the next available clock cycle, frame, or microframe.

Architecture for USB-synchronized array of speakers

A computing device includes a touch-sensitive user interface configured to present a unified collaborative session for two or more users, and an authentication module configured to simultaneously identify and authenticate multiple users physically co-located within a collaborative environment, allowing each of the multiple users to interact with the touch-sensitive user interface. A content module is configured to simultaneously provide one or more content portals within the unified collaborative session for each authenticated user. Each content portal is configured to enable an authenticated user to access, retrieve, and present user-owned content files within the unified collaborative session. In this way, multiple users may simultaneously access, retrieve, and present their own content files on a single computing device.

Simultaneous authentication system for multi-user collaboration

Techniques are described herein that are capable of performing sound source localization (SSL) confidence estimation using machine learning. An SSL operation is performed with regard to a sound to determine an SSL direction estimate and an SSL-based confidence associated with the SSL direction estimate based at least in part on a multi-channel representation of the sound. The SSL direction estimate indicates an estimated direction from which the sound is received. The SSL-based confidence indicates an estimated probability that the sound is received from the estimated direction. The multi-channel representation includes representations of the sound that are detected by respective sensors (e.g., microphones). Additional characteristic(s) of the sound are automatically determined. A machine learning (ML) operation is performed based at least in part on the SSL direction estimate, the SSL-based confidence, and the additional characteristic(s) to determine an ML-based confidence associated with the SSL direction estimate.

Sound source localization confidence estimation using machine learning

An electronic device with directional MEMS microphone assembly is provided, including a MEMS microphone capsule with a PCB affixed thereto, and a housing affixed to the PCB. The microphone assembly includes a first internal port and a second internal port through the PCB, wherein the first and second internal ports fluidically communicate with the MEMS microphone capsule. The microphone assembly further includes first and second external ports through the housing, wherein the first external port is offset from the first internal port in an offset direction perpendicular to a thickness direction of the microphone assembly. The microphone assembly further includes first and second cavities located between the PCB and the housing, wherein the first cavity fluidically communicates with the first internal port and the first external port, and the second cavity fluidically communicates with the second internal port and the second external port.

Electronic device including directional MEMS microphone assembly

Suppression of tapping noise caused by tapping an acoustically coupled touchscreen. When a tapping event is detected on the touchscreen, a tapping noise suppressor is alerted of the event, and responds by at least temporarily mitigating the tapping noise in the audio stream. The suppression occurs temporarily for at least part of the duration of surge in audio levels that occurs as a result of the tapping event, and in a manner that reduces the psychoacoustic impact on the conversation. Suppression may be performed by first placing the tapping noise suppressors in an alert mode when a tapping event occurs. The tapping noise suppressor then monitors the audio stream generated by the microphone for the beginning of the tapping event&#x2014;which will be represented in the form of a surge in volume. The tapping noise suppressor then temporarily applies the suppression window.

Growing community of inventors

Seattle, WA, United States of America

Kevin Juho Venalainen