Active noise reduction device

Abstract

An active Noise Reduction (ANR) device includes a plurality of inputs, a plurality of signal processing resources, an output for driving an earphone driver, a programmable switch arrangement capable of assigning any of the plurality of inputs to any of the plurality of signal processing resources, and a controller for programming the programmable switch arrangement in order to assign each of at least a subset of the plurality of inputs to a different one of the signal processing resources. The ANR device is dynamically configurable to vary which of the signal processing resources are selected to contribute to the output.

Claims

1. An Active Noise Reduction (ANR) device comprising: a plurality of inputs; a plurality of signal processing resources comprising: at least one analogue signal processing resource; and a plurality of digital signal processing resources; an output for driving an earphone driver; a programmable switch arrangement provided downstream of the plurality of inputs and upstream of the plurality of signal processing resources, the programmable switch arrangement being capable of assigning each of the plurality of inputs to each of the plurality of signal processing resources; and a controller for programming the programmable switch arrangement in order to assign each of at least a subset of the plurality of inputs to a different one of the signal processing resources; wherein the ANR device is dynamically configurable to vary which of the signal processing resources are selected to contribute to the output by dynamically configuring the programmable switch arrangement between at least first and second modes of operation to vary assignment of the signal processing resources to the plurality of inputs; wherein the ANR device is configured such that: in the first mode of operation one or more of the at least one analogue signal processing resource is selected to contribute to the output and one or more of the plurality of digital signal processing resources are selected not to contribute to the output, whereby a subset of the plurality of the inputs are assigned to the selected one or more of the at least one analogue signal processing resource; and in the second mode of operation one or more of the plurality of digital signal processing resources are selected to contribute to the output and one or more of the at least one analogue signal processing resources are selected not to contribute to the output, whereby a subset of the plurality of inputs are assigned to the selected one or more of the plurality of digital signal processing resources.

2. An ANR device according to claim 1, wherein the plurality of signal processing resources comprises a plurality of filters, the plurality of filters including: a plurality of analogue filters; and a plurality of digital filters.

3. An ANR device according to claim 1, wherein the ANR device is dynamically configurable so as to minimise a figure-of-merit or cost-function parameter.

4. An ANR device according to claim 1, wherein the plurality of inputs include: a plurality of analogue inputs comprising at least two analogue microphone inputs and at least one analogue audio input; and a plurality of digital inputs comprising at least two digital microphone inputs and at least one digital audio input.

5. An ANR device according to claim 4, wherein in the first mode of operation the selected analogue signal processing resource is configured as a feedforward ANR filter and the assigned input is an analogue feedforward microphone input.

6. An ANR device according to claim 4, wherein in the first mode of operation the selected analogue signal processing resource is configured as an analogue feedback ANR filter and the assigned input is an analogue feedback microphone input.

7. An ANR device according to claim 4, wherein in the first mode of operation the selected analogue signal processing resource is configured as an equalisation filter and the assigned input is an analogue audio input.

8. An ANR device according to claim 4, wherein in the second mode of operation the selected digital signal processing resource is configured as a feedforward ANR filter and the assigned input is a digital feedforward microphone input.

9. An ANR device according to claim 8, wherein in the second mode of operation the ANR device is further dynamically configurable to further select an analogue signal processing resource configured as an analogue feedback ANR filter and assigned to an analogue feedback microphone input.

10. An ANR device according to claim 4, wherein in the second mode of operation the selected digital signal processing resource is configured as an equalisation filter and the assigned input is a digital audio input.

11. An ANR device according to claim 1, wherein the ANR device is configured to power down or reduce power to signal processing resources that are not selected to contribute to the output.

12. An ANR device according to claim 1, wherein the ANR device is operative to provide a resource sharing output signal to an external device operative to provide an external signal processing resource.

13. An ANR device according to claim 12, wherein the programmable switch arrangement is operative to provide the resource sharing output signal to the external device.

14. An ANR device according to claim 13, wherein the resource sharing output signal is provided is via a dedicated output.

15. An ANR device according to claim 13, wherein the resource sharing output signal is provided using one of the plurality of inputs.

16. An ANR device according to claim 15, wherein the programmable switch arrangement is programmable to assign one or more of the plurality of inputs as an output for the resource sharing output signal.

17. An ANR device according to claim 12, wherein the plurality of signal processing resources is expandable to include an external signal processing resource assignable to an input by the programmable switch arrangement.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 is a schematic illustration of an ANR device in accordance with a first embodiment prior to configuration during manufacture.

(2) FIG. 2 is a schematic illustration of an ANR device accordance with a second embodiment prior to configuration during manufacture.

(3) FIG. 2A is a schematic illustration of a first configuration of an ANR device formed from the unconfigured ANR device of FIG. 2.

(4) FIG. 2B is a schematic illustration of a second configuration of an ANR device formed from the unconfigured ANR device of FIG. 2.

(5) FIG. 2C is a schematic illustration of a third configuration of an ANR device formed from the unconfigured ANR device of FIG. 2.

(6) FIG. 2D is a schematic illustration of a fourth configuration of an ANR device formed from the unconfigured ANR device of FIG. 2.

(7) FIG. 2E is a schematic illustration of a fifth configuration of an ANR device formed from the unconfigured ANR device of FIG. 2.

(8) FIG. 2F is a schematic illustration of a sixth configuration of an ANR device formed from the unconfigured ANR device of FIG. 2.

(9) FIG. 3 is a schematic illustration of an ANR device accordance with a third embodiment prior to configuration during manufacture.

(10) FIG. 3A is a schematic illustration of a first configuration of an ANR device formed from the unconfigured ANR device of FIG. 3.

(11) FIG. 3B is a schematic illustration of a second configuration of an ANR device formed from the unconfigured ANR device of FIG. 3.

(12) FIG. 3C is a schematic illustration of a third configuration of an ANR device formed from the unconfigured ANR device of FIG. 3.

(13) FIG. 3D is a schematic illustration of a fourth configuration of an ANR device formed from the unconfigured ANR device of FIG. 3.

DETAILED DESCRIPTION

(14) Described herein are methods and devices directed to the concept of constructing an ANR device from a new type of configurable device having architectural and processing resources for active control which are uncommitted at the time of manufacture. Unlike ANR devices familiar from the prior art, the inputs of the disclosed ANR devices are not uniquely hard-wired to internal processing resources; rather, it is possible to assign the inputs to the processing resources as best matches the demands of a particular application to which the device is targeted. This assignment is made through a flexible, programmable switching scheme and allows the device to be optimised for different applications, characterised by different balances of cost/power consumption/system functionality.

(15) FIG. 1 shows a simplified example of a configurable device 1 operating upon a plurality of inputs 2 to produce the single output 3 for driving an earphone (for simplicity only one channel of a stereo or binaural pair of channels is illustrated and discussed). The inputs 2 are coupled to a plurality of signal processing resources 5 (configurable filters) by programming a switching array 4 which maps the inputs to the plurality of signal processing resources 5, the output of which is the system response 3 (i.e. the control signal which drives the earphone actuator). The control signal is provided at appropriate amplitude and impedance by an amplifier 6, whose gain, along with the other configurable parameters of the system, is under the control of the supervisory block 7, which itself may respond to external control and programming inputs 8. The device will include other power and housekeeping functions 9.

(16) During manufacture a subset of the plurality of signal processing resources 5 is selected to contribute to the output of the device based on a design specification. Unselected signal processing resources are non-enabled so as to not contribute to the output of the device in any modes of operation. The selected signal processing resources are mapped to a subset of the plurality of inputs 2 in a one-to-one relationship via switching array 4 using supervisory block 7. Supervisory block 7 is additionally utilised to configure the selected signal processing resources to operate as desired filter types (e.g. feedforward, feedback or equalisation filters depending upon the type of input and the requirements of the specification). Advantageously, device 1 allows a range of differently specified ANR devices to be manufactured from a common device platform.

(17) FIG. 2 shows a more advanced device 1 based on device 1 (features in common are labelled accordingly) that operates on first and second pluralities of inputs 10, 11 to produce the single output 3. The first plurality of inputs 10 is a set of analogue signals whilst the second plurality of inputs 11 is a set of digital signals. Each of the first plurality of inputs 10, being an analogue signal, can only properly be processed directly by analogue means. Thus first plurality of inputs 10 is mapped through switching array 12 to a set of analogue processing resources 14. Similarly, the second plurality of inputs 11 is handled by its own switching array, 13 and processing resources 15. The output of the digital processing block 15 is converted to an analogue signal by a digital to analogue converter 16 before the weighted sum of the analogue and digital paths form the single (analogue) control output 3.

(18) The device as described introduces a divide between the two formats of analogue and digital. In some circumstances, it could be advantageous for a signal available in one format to be processed on a processing resource native to the other. This is provided by the introduction of data converters 17, 18 between the input switching arrays. A digital to analogue converter (DAC) 17 allows information encoded on a digital input to be applied to processing resources available in the analogue block, whilst conversely an analogue to digital converter (ADC) 18 allows analogue signals to be digitally processed.

(19) Each of the analogue and digital processing blocks 14, 15 shares a common basic architecture. Each consists of a series of programmable filters 19, which are summed at processing block 20 to form a single output, giving the block a multiple-input, single-output structure. Both the analogue and digital blocks 14, 15 have at least two inputs. It is the function of the switching arrays 12, 13 to populate these inputs appropriately, with signals from the input array. The summation at the end of each of the processing block 20 is an explicit weighted sum 21.

(20) In order to better manage gain distribution within a practical implementation of the device, a further pair of amplifiers and/or attenuators 22, 23 may extend the implementation of the weighted sum to the input of the final output amplifier 6.

(21) Device 1 is configurable for the purpose of optimising the noise cancelling performance of any product or system in which it is applied, the total cost of any system in which it is applied (where cost may be understood in terms of Bill-of-Materials, manufacturing and configuration cost, etc.) and the total power consumption of any system in which it is applied. In order to optimise power consumption, elements of the device not used in any configuration are capable of being powered down, to reduce power drain. Such elements include the ADC and DAC 17, 18 between the input switching arrays and elements of the input switching arrays 12, 13 and the analogue processing resources 14.

(22) Analogue processing block 14 includes a series of parallel filter paths, each of which potentially includes active circuits, which may consume power when not in use 32.

(23) The input switching arrays include interface circuits to support direct connections to microphones. These are provided in the digital switching array 33 to support the interface to digital microphones 34. The analogue switching array similarly includes interface circuits 35 specific to conventional analogue microphones 36. In both casesthough particularly in the case of the digital microphones and their interfacespowering down these sub-systems when not required represents a considerable and attractive power saving.

(24) The system of FIG. 2A shows (one channel of) an application of device 1 applied to a simple hybrid (i.e. feedforward and feedback) noise cancelling earphone application in which an analogue microphone 36 provides a signal for feedback control via analogue microphone input 35 and a digital microphone 34 provides a signal for feedforward control via digital microphone input 33. Analogue microphone input 35 is routed for filtering by programmable analogue filter 19A. Digital microphone input 34 is routed for filtering by programmable digital filter 19B.

(25) The system of FIG. 2B shows an application of the newly-disclosed device applied to a simple hybrid (i.e. feedforward and feedback) noise cancelling headphone application, in which analogue technology is used in pursuit of low overall system power consumption. Two analogue microphones 36, 37 provide the observation required for feedback and feedforward control, with the signals entering the newly-disclosed device at the analogue array's analogue microphone inputs 35, 38 and being routed to the two analogue processing channels 32, 39, where the two control signal components are designed.

(26) Audio program material enters the device as an analogue signal at 40 and is routed from the analogue input through the data converter 18 into the digital switching array, from where it is further routed to the digital processing block, where one of the filtering paths 41 applies compensation/equalisation. Notice that the other block in the digital path 42 is implemented on a numerical machine and there is little meaning in powering it down, despite the fact that it is not being used in this application. The digital microphone interface 33 on the other hand is explicitly powered down.

(27) The same device applied to a different target product, in which the highest possible hybrid noise cancelling performance is soughteven at the expense of higher power consumptionmay be configured differently, as suggested in FIG. 2C. In the application of FIG. 2C, the feedback noise reduction has been retained, but the higher differential order filtering possible with digital filtering has been exploited in the feedforward path. This has motivated the removal of an analogue microphone and the powering down of both its interface 38 and the analogue processing block 39, which was filtering the feedforward signal. The analogue microphone is replaced by a digital microphone 43 on the now powered-up interface 33, whose output is fed to the second digital filter path 42. Notice that the availability of a digital audio stream would allow the power hungry data converter 18 to be turned off.

(28) Assignment of signals from the input array to the processing resources is made at the time of configuration during manufacture. This assignment is made with reference to the requirements of the application, bearing in mind the functional demands of the application and the power implications of selecting any resource. For example, a low-cost product which is expected to draw low power from its battery might be forced to implement feedforward noise cancellation using a low-power analogue microphone, providing a signal which is filtered to relatively low levels of complexity by an analogue filter, itself consuming low power. However, application in a more exacting product may justify the specification of a more expensive and power-hungry digital microphone, whose signal is operated upon by a digital filter, able to operate at higher differential order and thereby able to deliver more complete noise cancellation. This flexibility of matching resources to application requirement across a wide range of target applications is not possible with prior art off the shelf noise cancelling devices. However, there is a further aspect of the disclosed device, which extends its flexibility still further.

(29) In addition to the ability to dispose the information gathered from the sensor inputs between the processing resources available on the device, as discussed above, it is an intended feature of the newly-disclosed device that it is further capable of exploiting processing resources located external to itself. By this means, an entire noise cancelling system may make use of processing means available on nearby sub-systems, in a resource-sharing strategy. This allows, for example, the entire system's power consumption to be optimised in an application where processing resources are at risk of duplication. It also allows a degree of future-proofing for the present device, allowing it to take advantage of resources which are not availableor conceived ofat the time of its design.

(30) This resource sharing strategy is best exemplified in the case of a wireless headphone, in which the newly-disclosed device is enabling the headphone in concert with a Bluetooth or similar wireless Codec. Such a Codec often is capable of digital filtering, which can be exploited to serve duty in any of the audio, monitor or feedforward roles made possible by the signal routing flexibility of the newly-disclosed device.

(31) As illustrated in FIG. 2D, in order to support the distribution of sensor information to remote processing resources, input switching arrays 10, 11 may provide outputs 24, 25 from the device for connecting processing means 26, 28 on remote resources. Results from remote processing resources are coupled back into the input vector of the device. Remote analogue processors 26 operating on the signal derived from 24 are typically themselves analogue signals and are fed back to an analogue input 27. Similarly for a digital remote processor 28 returning its result to a digital input 29.

(32) FIG. 2E shows an alternative configuration in which the cost and complexity of providing dedicated outputs 24, 25 are replaced by allowing the application to tap off the connection to the relevant transducer. As illustrated in FIG. 2E remote analogue processor derives its input from a tap on the (otherwise unused) analogue input 30. Alternatively, a digital remote processor is shown tapping off an application circuit line 31, which is making no connection to the newly-disclosed device.

(33) FIG. 2F shows a further alternative configuration in which a further output 44 and an input 45 provide an expansion path to allow the analogue processor to be expanded by external processing resource 48. As illustrated, output 44 allows a signal received via switching array 12 to be passed to external processing resource 48. Input 45 allows external processing resource 48 to return a processed signal component 46 into the output of analogue processor 14 (this component may optionally be capable of scaling by a constant such as shown at 47 or more elaborate linear filtering). Such expansion of the architecture of the analogue processor is seen to result in a different overall transfer function than is possible by routing a signal to an external resource and then returning the processed result through the input matrix and thence through the processing resources as previously described.

(34) A more detailed embodiment will now be described with reference to FIG. 3 and associated applications of the device illustrated in FIGS. 3A-D. These applications illustrate how the resources of not only the device alone, but all the resources of all devices in a system, can be shared so as to optimise performance with respect to application-critical parameters.

(35) FIG. 3 illustrates an unconfigured ANR device 1 (based on ANR device 1features in common are labelled accordingly) comprising two analogue and two digital filter paths, each of which is programmable, for each of two stereo/binaural channels. The filters are driven by a range of inputs, derived from analogue and digital microphone inputs and digital and analogue audio inputs.

(36) In the simplest, low-power application, the system is configured during manufacture as shown at FIG. 3A, in which hybrid noise cancellation (i.e. feedforward and feedback) is delivered using low-power analogue microphone technologies allied with simple analogue filtering. Despite its significant advantages (of low noise, zero latency and low power consumption), the analogue filtering is able only to operate with relatively modest differential order, so it delivers only a limited degree of noise cancellation in some applications. The audio signal in FIG. 3A is fed into an analogue to digital converter and through the digital path for equalisation.

(37) In a more ambitious application for a wired, stand-alone headphone, the same device could be configured as shown at FIG. 3B, in which hybrid noise cancellation (i.e. feedforward and feedback) is delivered using digital microphone technology, allowing the feedforward filtering to be implemented using digital filtering means at higher digital order. This will usually result in a higher level of total noise cancelling performance at the expense of higher overall power consumption and higher component cost.

(38) In the case of a wireless headphone application optimised for power consumption, as shown in FIG. 3C, feedforward noise cancellation would be provided by signals applied to the inputs of the Digital Audio Controller and filtered by resources on that device, before being passed into the newly-disclosed component's digital audio input, along with digital program material. Feedback control would again be realised by analogue means.

(39) In the case of a wireless headphone application optimised for noise cancelling performance, shown in FIG. 3D, feedforward noise cancellation would be generated by internal digital processing operations on signals obtained from a digital microphone. The same digital microphone's output would be shared by the Digital Audio Controller and filtered there to provide Monitoring (/talk through) signals and/or sidetone signals for telephony.