Accurate timbre and localization of binaural recordings through headphones using sound pressure measurement at the eardrum.
David Griesinger
Abstract
David Griesinger
Abstract
Individual HRTFs are almost universally measured with blocked or partially blocked ear canals, and individual headphone equalization is obtained in the same way. Using this method binaural reproduction without head tracking usually results in a significant alteration of timbre as well as inaccurate and/or in the head localization. The error is important—auralizations using incorrect timbre can lead to misleading conclusions. This problem was studied with the help of a dummy head that precisely models the author’s pinna, ear canals, and eardrum resonance. A new type of probe microphone was also developed that allows comfortable recordings and measurements at the eardrum. Data from this equipment show that while the spatial variation in HRTFs can be captured with a blocked ear canal up to a frequency of 7 kHz, errors in headphone equalization made with a blocked canal are typically in error by more than 10 dB at midfrequencies. Eardrum measurements of HRTFs and headphones result in superior performance for an individual, and a pair of headphones correctly equalized for an average individual produces out-of-head localization for at least 50% of other individuals in our tests. Data and theory and a noninvasive method for equalization will be presented.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Individual HRTFs are almost universally measured with blocked or partially blocked ear canals, and individual headphone equalization is obtained in the same way. Using this method binaural reproduction without head tracking usually results in a significant alteration of timbre as well as inaccurate and/or in the head localization. The error is important—auralizations using incorrect timbre can lead to misleading conclusions. This problem was studied with the help of a dummy head that precisely models the author’s pinna, ear canals, and eardrum resonance. A new type of probe microphone was also developed that allows comfortable recordings and measurements at the eardrum. Data from this equipment show that while the spatial variation in HRTFs can be captured with a blocked ear canal up to a frequency of 7 kHz, errors in headphone equalization made with a blocked canal are typically in error by more than 10 dB at midfrequencies. Eardrum measurements of HRTFs and headphones result in superior performance for an individual, and a pair of headphones correctly equalized for an average individual produces out-of-head localization for at least 50% of other individuals in our tests. Data and theory and a noninvasive method for equalization will be presented.
Key concepts: Eardrum, Headphones, Binaural recording, Acoustics, Ear canal, Equalization (audio), Timbre, Computer science