At the 118th AES Convention in a workshop problems and solutions for automatic down-mixing have been summarized. The key question is for which items automatic algorithms provide acceptable quality. In close conjunction to this issue is the question how to evaluate the quality of the result. Standardized listening test procedures like ITU-R BS.1116 and ITU-R BS.1534 are designed to evaluate the difference between an unimpaired reference and a modified signal under test. They never had been intended to be used for the comparison of 5-channel signals to 2-channel signals. In this paper a new listening test procedure is described which is designed to judge the quality of down-mixing algorithms. First results of listening tests performed using this procedure are described.
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.