To provide an audio signal processing system, a surround signal generation method and so on that are capable of obtaining stable output signal level and feeling of soundscape without depending on an audio input signal. A DSP inputs audio input signals respectively corresponding to a plurality of channels, and generates surround signals that have reflection sound and reverberate sound and that respectively correspond to the plurality of channels. The DSP generates the surround signal corresponding to one channel based on an audio input signal corresponding to the one channel, a variable number changing within the predetermined range in response to a signal level of the surround signal at a preceding predetermined time corresponding to the one channel, and the audio input signal corresponding to the other channel.