In this paper a large amount of absolute mean opinion scores obtained within a single listening test are presented. Naïve listeners’ preference on different speech signal properties such as mono/stereo and bandwidth was studied. Various codecs were ranked by their subjective voice quality. The listened speech sequences were recorded and selected to represent several realistic stereo audio capture and background noise configurations, where there are either one or several speakers. The number of conditions was selected to be as large as possible to be listenable in a single two hour session. Due to the test size, the results are divided into smaller graphs where interesting comparisons between different conditions can be easily evaluated.