Adaptive equalization of acoustic transparency in an augmented-reality headset

Juho Liski, Riitta Väänänen, Sampo Vesa, Vesa Välimäki

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

7 Citations (Scopus)

Abstract

Headphones are commonly used in noisy environments. Insert headphones attenuate and color the spectra of ambient sounds and thus alter the auditory perception. When the ambient sounds are desirable, a hear-through function can be used to reproduce them naturally while wearing headphones, i.e. to make the headphones acoustically transparent. A novel adaptive hear-through algorithm is proposed, which estimates the isolation and fine-tunes the hear-though equalization for optimal acoustic transparency. Measurements on a prototype headset and simulations show that the proposed algorithm produces acoustic transparency with default settings when the fit is good, and that the adaptation improves the transparency by up to 6 dB when the headset is poorly fitted. Volume control with additional shelving filter adjustments reduces the comb-filtering effect at frequencies below 1 kHz. The proposed algorithm is a suitable premise for augmented reality audio applications and offers improved behavior when compared to fixed hear-through systems.
Original languageEnglish
Title of host publication2016 AES International Conference on Headphone Technology
EditorsAlexander Lindau, Jurgen Peissig
PublisherAudio Engineering Society
Number of pages8
ISBN (Electronic)978-1-942220-09-1
Publication statusPublished - 24 Aug 2016
MoE publication typeA4 Article in a conference publication
EventAES International Conference on Headphone Technology - Aalborg, Denmark
Duration: 24 Aug 201626 Aug 2016
http://www.aes.org/conferences/2016/headphones/

Conference

ConferenceAES International Conference on Headphone Technology
CountryDenmark
CityAalborg
Period24/08/201626/08/2016
Internet address

Keywords

  • Acoustics
  • audio signal processing
  • equalizers
  • headphones

Fingerprint

Dive into the research topics of 'Adaptive equalization of acoustic transparency in an augmented-reality headset'. Together they form a unique fingerprint.

Cite this