对于多路语音输入生成的音视频文件，要同时播放需要在生成的时候进行混音处理，有些播放器不支持多个音轨同时播放，ffmpeg可以用命令对于已经生成的来回改革或多个音频进行混合，使用以下命令

ffmpeg -i test-1.aac -i test2-3.aac -filter_complex amix=inputs=2:duration=first:dropout_transition=2 output.aac

参数可以参考下面代码、

static const AVOption amix_options[] = {
    { "inputs", "Number of inputs.",
            OFFSET(nb_inputs), AV_OPT_TYPE_INT, { .i64 = 2 }, 1, INT16_MAX, A|F },
    { "duration", "How to determine the end-of-stream.",
            OFFSET(duration_mode), AV_OPT_TYPE_INT, { .i64 = DURATION_LONGEST }, 0,  2, A|F, "duration" },
        { "longest",  "Duration of longest input.",  0, AV_OPT_TYPE_CONST, { .i64 = DURATION_LONGEST  }, 0, 0, A|F, "duration" },
        { "shortest", "Duration of shortest input.", 0, AV_OPT_TYPE_CONST, { .i64 = DURATION_SHORTEST }, 0, 0, A|F, "duration" },
        { "first",    "Duration of first input.",    0, AV_OPT_TYPE_CONST, { .i64 = DURATION_FIRST    }, 0, 0, A|F, "duration" },
    { "dropout_transition", "Transition time, in seconds, for volume "
                            "renormalization when an input stream ends.",
            OFFSET(dropout_transition), AV_OPT_TYPE_FLOAT, { .dbl = 2.0 }, 0, INT_MAX, A|F },
    { "weights", "Set weight for each input.",
            OFFSET(weights_str), AV_OPT_TYPE_STRING, {.str="1 1"}, 0, 0, A|F|T },
    { "normalize", "Scale inputs",
            OFFSET(normalize), AV_OPT_TYPE_BOOL, {.i64=1}, 0, 1, A|F|T },
    { NULL }
};