6.4 Stereo to mono conversion

26.2903GPPAudio codec processing functionsExtended Adaptive Multi-Rate - Wideband (AMR-WB+) codecRelease 17Transcoding functionsTS

In the case the application or hardware does not support stereo output, the output of the synthesised output stereo bit stream needs to be mixed to mono.

6.4.1 Low-Band synthesis

The low band stereo output is converted to mono by omitting the stereo information in the bit stream and skipping the stereo decoder. Hence, the mono output is synthesised only according to Section 6.1.

6.4.2 High-Band synthesis

The stereo bit stream contains separate bandwidth extensions for left and right High-Band synthesis. To minimise the decoder complexity, the stereo to mono down mixing is done on parameter level.

First the High-Band gain factors for both left and right channel are decoded as explained in Section 6.2. The gain difference of left and right channels is calculated in dB scale. When the difference is greater than 20 dB, the LP synthesis filter coefficients of the higher energy channel are selected for the mono output synthesis filtering and the LP parameters from the lower energy channel are omitted completely. On the other hand, when the gain difference is less than the 20 dB threshold, the LP synthesis filter coefficients of left and right channel are averaged in ISP domain. The gain parameters from left and right channels are averaged in both cases. When the LP synthesis filter coefficients and gain factor is determined, the High-Band synthesis is done according to mono signal High-Band synthesis in Section 6.2.