This study explores the extent to which differences in voice quality with different bit rates become less perceptible when users are listening in a noisy environment. The individual rate modes of two multi-rate codecs were rated by listeners in various background noise conditions, including a quiet baseline, crowd babble, street noise, factory noise, and two levels of car noise. The results suggest that in some cases a lower bit-rate codec can be substituted without an associated drop in perceived quality when the listener is in a noisy location. Based on this effect, it would be possible to increase the system capacity or allow graceful handling of network overload by reducing transmission bandwidth allocated to receivers in high background noise without associated reduction in perceived voice quality.
展开▼
