Personalized three-dimensional audio
A headphone system includes a calibration microphone for performing a calibration routine with a user. The calibration microphone receives a stimulus signal emitted by the headphone system and generates a response signal indicating variations in the stimulus signal that arise due to physiological attributes of the user. Based on the stimulus signal and the response signal, the calibration engine generates response data. The calibration engine processes the response data based on a headphone transfer function (HPTF) associated with the headphone system in order to create an inverse filter that can reduce or remove acoustic variations caused by the headphone system. The calibration engine generates a personalized HRTF for the user based on the response data and the inverse filter. The personalized HRTF can be used to implement highly accurate 3D audio and is thereby well-suited for applications to immersive audio and audio-visual entertainment.
1. A computer-implemented method for generating a personalized head-related transfer function (HRTF) for a user, the method comprising:
generating response data based on a stimulus signal and a response signal, wherein a first audio driver within a headphone system transmits the stimulus signal towards a first ear of the user, and wherein the response signal is captured at the first ear of the user in response to the stimulus signal;
generating a target HRTF based on the response data, wherein the target HRTF characterizes at least one physical attribute of the user;
generating an inverse filter based on the response data and based on a headphone transfer function (HPTF) associated with the headphone system; and
generating the personalized HRTF for the user based on the inverse filter and the target HRTF.
2. The computer-implemented method of claim 1 , wherein the response data includes a partial near-field HRTF associated with the first ear of the user.
3. The computer-implemented method of claim 1 , wherein the stimulus signal comprises a sine wave sweep across a range of frequencies and a range of amplitudes.
4. The computer-implemented method of claim 1 , wherein generating the target HRTF comprises:
transforming the response data based on a location where the response signal is captured to generate transformed response data; and
selecting the target HRTF from a set of default HRTFs based on the transformed response data.
5. The computer-implemented method of claim 1 , wherein generating the target HRTF comprises:
transforming the response data based on the at least one physical attribute of the user to generate transformed response data; and
causing a machine learning model to synthesize the target HRTF based on the transformed response data, wherein the machine learning model is trained based on response data associated with a plurality of users and HRTFs generated for the plurality of users.
6. The computer-implemented method of claim 1 , wherein the inverse filter is configured to reduce acoustic artifacts caused by the headphone system.
7. The computer-implemented method of claim 1 , wherein the response signal is captured by a calibration microphone that is disposed within an inner portion of the first ear of the user.
8. The computer-implemented method of claim 7 , wherein the calibration microphone is coupled to the headphone system in place of a boom microphone that is configured to capture sounds produced by the user.
9. The computer-implemented method of claim 1 , further comprising modifying a perceived point of origination of a sound emitted by the first audio driver based on the personalized HRTF.
10. The computer-implemented method of claim 1 , wherein the at least one physical attribute of the user is captured via one or more sensors.
11. A non-transitory computer-readable medium storing program instructions that, when executed by a processor, cause the processor to generate a personalized head-related transfer function (HRTF) for a user by performing the steps of:
generating response data based on a stimulus signal and a response signal, wherein a first audio driver within a headphone system transmits the stimulus signal towards a first ear of the user, and wherein the response signal is captured at the first ear of the user in response to the stimulus signal;
generating a target HRTF based on the response data, wherein the target HRTF characterizes at least one physical attribute of the user;
generating an inverse filter based on the response data and based on a headphone transfer function (HPTF) associated with the headphone system; and
generating the personalized HRTF for the user based on the inverse filter and the target HRTF.
12. The non-transitory computer-readable medium of claim 11 , wherein the response data includes a partial near-field HRTF associated with the first ear of the user.
13. The non-transitory computer-readable medium of claim 11 , wherein the stimulus signal comprises a sine wave sweep across a range of frequencies and a range of amplitudes.
14. The non-transitory computer-readable medium of claim 11 , wherein the step of generating the target HRTF comprises:
transforming the response data based on a location where the response signal is captured to generate transformed response data; and
selecting the target HRTF from a set of default HRTFs based on the transformed response data.
15. The non-transitory computer-readable medium of claim 11 , wherein the step of generating the target HRTF comprises:
transforming the response data based on the at least one physical attribute of the user to generate transformed response data; and
causing a neural network to synthesize the target HRTF based on the transformed response data, wherein the neural network is trained based on response data associated with a plurality of users and HRTFs generated for the plurality of users.
16. The non-transitory computer-readable medium of claim 11 , wherein the inverse filter is configured to reduce acoustic artifacts caused by at least one of position variations and leakage associated with a first ear cup that includes the first audio driver and is worn over the first ear of the user.
17. The non-transitory computer-readable medium of claim 11 , wherein the response signal is captured by a calibration microphone that is disposed within an inner portion of the first ear of the user, and wherein the calibration microphone is coupled to the headphone system in place of a boom microphone that is configured to capture sounds produced by the user.
18. The non-transitory computer-readable medium of claim 11 , further comprising the step of modifying a perceived point of origination of a sound emitted by the first audio driver based on the personalized HRTF.
19. The non-transitory computer-readable medium of claim 11 , wherein the at least one physical attribute of the user is determined based on a user interaction with a graphical user interface.
20. A system, comprising:
a memory that stores a calibration engine; and
a processor that executes the calibration engine to perform the steps of:
transmitting, using a first audio driver within a headphone system, a stimulus signal towards a first ear of a user;
capturing a response signal at the first ear of the user in response to the stimulus signal;
generating response data based on the stimulus signal and the response signal;
generating a target HRTF based on the response data, wherein the target HRTF characterizes at least one physical attribute of the user;
generating an inverse filter based on the response data and based on a headphone transfer function (HPTF) associated with the headphone system; and
generating a personalized HRTF for the user based on the inverse filter and the target HRTF.