There are now several well-established ways to create artificial voices – some take a purely electronic approach, while others, like Professor Hideyuki Sawada at Kagawa University, Japan, have explored electro-mechanical methods. The video above shows how pneumatically-driven artificial lips, tongues, voicebox and nasal resonating cavities might simulate speech. Unfortunately, the mpeg videos linked-to via the professor’s website are now ‘404 not-found’ (missing) – but fortunately we can view more Japanese pneumatic voice synthesis progress at other Japanese research websites, like that of Professor Atsuo Takanishi at Waseda University. The Takanishi Laboratory provides a selection of online videos which demonstrate progress (also in mpg format) – may we recommend :
BONUS: Mashup artist Vonyco at Youtube has made a musical video compilation of some of the featured bots :
[Advisory : May induce earworms]