Evaluation of voxel-based rendering of high resolution surface descriptions

Dorte Hammershøi, Søren Krarup Olesen, Milos Markovic

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

Abstract

The auditory rendering in a VR system either entail real, acoustical measurements of the room, or – more commonly – a model of the room rendered. The models are usually based on relatively coarse approximations of the geometry, and the calculated room impulse responses deviate from measured impulse responses by having more “distinct” representations of the individual reflections. When rendered audible, such descriptions will lead to a sound quality that can best be described as “canned”. For the rendering of real rooms, as e.g. in "teletransporting", this problem may be addressed by using high resolution scans of the room surfaces as basis for the room models. In the present work this approach is evaluated with a voxel-based method (described in earlier publications), and compared to measured impulse responses. The results are compared objectively by visual inspection of the impulse responses, and by comparison of statistical parameters calculated from measured and modelled data. The work is ongoing.
Original languageEnglish
Title of host publicationProceedings of Forum Acusticum 2014
Number of pages7
PublisherEuropean Acoustics Association - EAA
Publication date7 Sept 2014
Article numberR05_8
Publication statusPublished - 7 Sept 2014
EventForum Acusticum 2014 - AGH University of Science and Technology, Kraków, Poland
Duration: 7 Sept 201412 Sept 2014
Conference number: 7

Conference

ConferenceForum Acusticum 2014
Number7
LocationAGH University of Science and Technology
Country/TerritoryPoland
CityKraków
Period07/09/201412/09/2014

Fingerprint

Dive into the research topics of 'Evaluation of voxel-based rendering of high resolution surface descriptions'. Together they form a unique fingerprint.

Cite this