Processing the acoustic effect of size in speech sounds