AN ENHANCED, EASY-TO-CONNECT, PORTABLE, 3D AUDIO CAPTURING MICROPHONE ARRAY
Researchers from ANU have developed a new type of microphone array, based on directional and omnidirectional microphones arranged in a unique, compact 2D grouping, that enables full 3D sound capture.
1) Flexible installation and placement – compact design enables easy mounting to the walls or ceilings of meeting rooms, studios, and aircraft and car cabins. 2) Better sound quality - removes unwanted background noise (using beamforming) and automatically focuses on the desired sound/speech source (using adaptive steering). 3) Inexpensive device – as compared to current microphone arrays with 3D
recording capability. 4) Plug and play - using USB or 3.5 mm microphone/audio jack. 5) Portable - compact, flat and lightweight; can easily fit into a laptop bag or handbag, can clip onto your laptop or sit unobtrusively on your desk. 6) Clear voice capture – noise reduction via: (i) spatial filtering, (ii) echo cancellation and noise suppression, (iii) electronic wind-noise reduction, and (iv) multichannel noise reduction processing. 7) More accurate speech recognition in high noise environments.
The 2D microphone
array system we have invented can
improve these activities without adding
significant costs and inconvenience
(due to its compact size and shape)
to the end-user. Due to the modular
nature of the design, we are able to
change the type and/or quality of the
components, altering the functionality of
the end product, and generating different
microphone-based products to meet
different needs. Possible model types
include: (i) desktop – features including
hemispherical field, thin moderate
directivity, USB or audio jack interface,
mono/multi-channel (7.1) recording, auto
voice tracking, USB powered; (ii) mobile/
tablet – features including hemispherical/
spherical field, small lower directivity, blue
tooth or audio jack interface, mono/multichannel
(5.1) recording, battery powered;
or (iii) high-end – features including full
3D capture, larger maximum directivity,
fire wire or USB interface, beamforming,
raw output recording, multichannel (7.1)
recording, 220V AC powered. Many
model options are available.
ANU researchers have developed a
novel, compact microphone array
arrangement and custom DSP
algorithms for enhanced 3D sound
capture and delivery. Through the use of
directional microphones, the researchers
have been able to significantly reduce the
size of the array (i.e. the new array has a
planar, 2D shape; Fig. 1B) as compared
to conventional spherical 3D arrays (Fig.
1A) which have similar functionality. This
array configuration will enable easier
integration into consumer electronics,
broadening the use of 3D sound
recording and analysis, in various audio
as well as radio-frequency applications.
The new array configuration consists
of multiple co-centred circular rings,
with two sensor types (directional
and omnidirectional microphones)
placed on a 2D plane (Fig. 1B). Vertical
alignment along the circular rings as
well as perpendicular placement to the
omnidirectional microphones enables
the directional microphones (also known
as first order sensors) to measure the
pressure gradient and vertical harmonic
components of an impinging soundfield.
The omnidirectional microphones (also
known as omnidirectional sensors)
which are also placed along the circular
rings (Fig. 1B) are used to measure the
sound pressure and horizontal harmonic
components of an impinging soundfield.
The two sensor types working together
in this 2D arrangement, enables the
microphone array to extract full 3D
soundfield information. This information
can then be utilised by DSP algorithms
for various sound refining purposes
such as beamforming and active noise
cancellation.
ANU researchers have also developed
an accompanying software solution for
the 2D microphone array which utilises
advanced DSP algorithms, including
sound source localisation, adaptive
beamforming and noise reduction,
to suppress background noise while
enhancing the speech signal. This
provides reliable, clear, crisp, audio
delivery which existing USB/desktop
microphone arrays cannot provide. For
example, current USB microphones (at
best) have cardioid and/or bidirectional
patterns. As such, they lack the spatial
resolution and the broader polar pattern
width of our new array, which means
that they still pick-up unwanted sound
from their surroundings. In addition, most
conventional microphone arrays only
measure sound pressure, whilst our new
array measures the pressure gradient
of a soundfield in addition to the sound
pressure.
To date, ANU researchers have produced
two differential microphone array (DMA)
prototypes based on the aforementioned
technology. DMAs use the spatial
derivative of the sound pressure field to
localise and track a desired speaker/
signal, and are particularly useful for
teleconferencing and speech pick-up
in noisy and reverberant environments.
The first prototype (Fig. 2A) consists of a
single pair of sensors (a directional and
omnidirectional microphone), whilst the
second prototype (Fig. 2B) encompasses
three pairs of sensors, with a working
first order system. Prototype materials
are comprised mainly of low cost
components, such as digital MEMs
microphones and micro-controllers.
Due to the modular nature of the
design, there is the ability to change the
quality of the components and thus the
Licensing/ commercial development
16/07/2015 00:00:00
WO2016011479
G01H 17/00 • G01S 3/00 • G06F 17/10 • G06F 19/00 • H01Q 21/00 • H01Q 21/24 • H01Q 21/29 • H04R 1/32
AU20140902837 [2014-07-23]
美国
