Determining and adapting to changes in microphone performance of playback devices
Systems and methods for determining and adapting to changes in microphone performance of playback devices are disclosed herein. In one example, an audio input is received at an array of individual microphones of a network microphone device. Output microphone signals are generated from each of the individual microphones based on the audio input. The output microphone signals are analyzed to detect a trigger event. After detecting the trigger event, the output microphone signals are compared to detect aberrant behavior of one or more of the microphones. Optionally, corrective actions can be taken or suggested based on the detection of aberrant behavior of one or more microphones.
1. A network microphone device (NMD) comprising:
a microphone array comprising a plurality of individual microphones;
a network interface;
one or more processors; and
a computer-readable medium storing instructions that, when executed by the one or more processors, cause the NMD to perform a method comprising:
receiving an audio input via the individual microphones;
producing output microphone signals from each of the individual microphones based on the audio input;
capturing a voice input based on at least one of the output microphone signals;
deriving frequency-domain sound data for each of the individual microphones based on the output microphone signals;
after capturing the voice input, sending, via the network interface, the voice input to one or more remote computing devices associated with a voice assistant service (VAS)
receiving, at the NMD, a response from a remote evaluator based on the frequency-domain sound data, the response (1) identifying one of the individual microphones and (2) indicating that the identified microphone has an aberrant signal; and
modifying a beamforming algorithm of the network microphone device to at least partially compensate for the aberrant signal.
2. The NMD of claim 1 , wherein the method further comprises sending, via the network interface, the frequency-domain sound data to the remote evaluator without exposing the output microphone signals from which the frequency-domain sound data is derived.
3. The NMD of claim 1 , wherein the one or more remote computing devices comprises the remote evaluator.
4. The NMD of claim 1 , wherein the method further comprises temporarily storing the output microphone signals in a lookback buffer.
5. The NMD of claim 4 , wherein the method further comprises:
analyzing the output microphone signals stored in the lookback buffer to detect a wake word associated with the VAS;
after detecting the wake word, capturing the voice input; and
after detecting the wake word, deriving the frequency-domain sound data based on the output microphone signals stored in the lookback buffer.
6. The NMD of claim 1 , wherein the indication of the aberrant signal comprises an indication of excess noise detected via one or more of the individual microphones.
7. The NMD of claim 1 , wherein the remote evaluator is distinct from the VAS.
8. A method comprising:
receiving an audio input via a plurality of individual microphones of a network microphone device (NMD);
producing output microphone signals from each of the individual microphones based on the audio input;
capturing a voice input based on at least one of the output microphone signals;
deriving frequency-domain sound data for each of the individual microphones based on the output microphone signals;
after capturing the voice input, sending, via a network interface of the NMD, the voice input to a first one or more remote computing devices associated with a voice assistant service (VAS); and
receiving, at the NMD, a response from a remote evaluator based on the frequency-domain sound data, the response (1) identifying one of the individual microphones and (2) indicating that the identified microphone has an aberrant signal; and
modifying a beamforming algorithm of the network microphone device to at least partially compensate for the aberrant signal.
9. The method of claim 8 , further comprising sending, via the network interface, the frequency-domain sound data to the remote evaluator without exposing the output microphone signals from which the frequency-domain sound data is derived.
10. The method of claim 8 , wherein the one or more remote computing devices comprises the remote evaluator.
11. The method of claim 8 , further comprising temporarily storing the output microphone signals in a lookback buffer.
12. The method of claim 11 , further comprising:
analyzing the output microphone signals stored in the lookback buffer to detect a wake word associated with the VAS;
after detecting the wake word, capturing the voice input; and
after detecting the wake word, deriving the frequency-domain sound data based on the output microphone signals stored in the lookback buffer.
13. The method of claim 8 , wherein the indication of the aberrant signal comprises an indication of excess noise detected via one or more of the individual microphones.
14. The method of claim 8 , wherein the remote evaluator is distinct from the VAS.
15. Tangible, non-transitory computer-readable medium storing instructions that, when executed by one or more processors of a network microphone device (NMD), cause the NMD to perform a method comprising:
receiving an audio input via a plurality of individual microphones;
producing output microphone signals from each of the individual microphones based on the audio input;
capturing a voice input based on at least one of the output microphone signals;
deriving frequency-domain sound data for each of the individual microphones based on the output microphone signals;
after capturing the voice input, sending, via a network interface of the NMD, the voice input to one or more remote computing devices associated with a voice assistant service (VAS);
receiving, at the NMD, a response from a remote evaluator based on the frequency-domain sound data, the response (1) identifying one of the individual microphones and (2) indicating that the identified microphone has an aberrant signal; and
modifying a beamforming algorithm of the network microphone device to at least partially compensate for the aberrant signal.
16. The computer-readable medium of claim 15 , wherein the method further comprises sending, via the network interface, the frequency-domain sound data to the remote evaluator without exposing the output microphone signals from which the frequency-domain sound data is derived.
17. The computer-readable medium of claim 15 , wherein the one or more remote computing devices comprises the remote evaluator.
18. The computer-readable medium of claim 15 , wherein the method further comprises temporarily storing the output microphone signals in a lookback buffer.
19. The computer-readable medium of claim 18 , wherein the method further comprises:
analyzing the output microphone signals stored in the lookback buffer to detect a wake word associated with the VAS;
after detecting the wake word, capturing the voice input; and
after detecting the wake word, deriving the frequency-domain sound data based on the output microphone signals stored in the lookback buffer.
20. The computer-readable medium of claim 15 , wherein the indication of the aberrant signal comprises an indication of excess noise detected via one or more of the individual microphones.