The great thing about computers – and computerised speech synthesis systems – is that they work tirelessly, never stopping to draw breath… try that last part again: the trouble with speech synthesis systems is that they never stop for breath according to Technology Review, Douglas Whalen, of Haskins Labs in New Haven, Connecticut has found that synthetic speech can be made much more intelligible by adding the sound of a breath before each sentence – and test subjects came up with the same answer whether the breathless was followed by the breathy or vice-versa; he now plans to vary the length of the breaths and to insert them in the middle of sentences where a human would normally pause for breath.
