Abstract: Intensifying contextually relevant sound events for time-delayed broadcast uses multiple directional microphones that capture sound events from a specific location within regions of a stadium. The events are contextually relevant to the environment, such as a referee's whistle sound. A processor executes signal processing of the captured events on each channel to produce audio samples each having a signal profile. The profiles are automatically compared to reference feature templates stored in a database, correspond to pre-identified sound events of contextual relevance. The signal processing accentuates characteristic traits in the events, reflective of contextually relevant events that should be included in a final composite audio output for transmission. If the comparison of samples to the feature templates suggests a high probability of a correspondence, then buffered audio for that channel is introduced into a final audio mix.

Abstract: Intensifying contextually relevant sound events for time-delayed broadcast uses multiple directional microphones that capture sound events from a specific location within regions of a stadium. The events are contextually relevant to the environment, such as a referee's whistle sound. A processor executes signal processing of the captured events on each channel to produce audio samples each having a signal profile. The profiles are automatically compared to reference feature templates stored in a database, correspond to pre-identified sound events of contextual relevance. The signal processing accentuates characteristic traits in the events, reflective of contextually relevant events that should be included in a final composite audio output for transmission. If the comparison of samples to the feature templates suggests a high probability of a correspondence, then buffered audio for that channel is introduced into a final audio mix.