AES New York 2013
Thursday, October 17, 2:30 pm — 4:30 pm (Room 1E10)
W6 - Design and Usage of Anchors in Listening Tests for Audio Quality EvaluationChair:
Frederik Nagel, Fraunhofer Institute for Integrated Circuits IIS - Erlangen, Germany; International Audio Laboratories - Erlangen, Germany
Schuyler Quackenbush, Audio Research Labs - Scotch Plains, USA
Francis Rumsey, Logophon Ltd. - Oxfordshire, UK
Thomas Sporer, Fraunhofer Institute for Digital Media Technology IDMT - Ilmenau, Germany; Ilmenau University of Technology - Ilmenau, Germany
Nick Zacharov, DELTA SenseLab - Iisalmi, Finland
Listening tests for the evaluation of audio or speech quality (like MUSHRA or P800) employ anchors in order to stabilize rating scales and facilitate comparison among testing sites. As coding technology progressed, traditional anchors (including band-limited or noisy signals) are no longer related to artifacts occurring in state-of-the art audio codecs. Expert listeners probably have an internal reference for basic audio quality that may allow them to evaluate audio quality even without any anchors in case of mono and two–channel stereo material. However, especially for newly emerging technologies (including multichannel audio) anchors could have a huge impact as many listeners are still building their internal reference scale.
We will discuss:
• How should anchors be created in order to exhibit typical artifacts of modern audio codecs?
• Is it possible to create anchors with an expected quality for a variety of different input material?
• Which features should anchors have that are used in multichannel testing?
• How important are anchors at all, compared to internal anchors and references of experts?
• What effect has modification of anchors to the results?
|This session is presented in association with the AES Technical Committee on Perception and Subjective Evaluation of Audio Signals|