Each recording had a series of rhyming words that the volunteers had to identify, with and without the intelligent microphone, for different levels of background noise.