On Wed, 28 Nov 2007 00:37:00 -0800 (PST), MIG
wrote:

On Nov 28, 1:37 am, Barry Salter wrote:
Boltar wrote:
What I don't understand is why companies bother using recordings when
speech synthesizers cost buttons, are about the size of one and can
say anything you want - you don't have to pay some actor to record new
phrases if things change , you just type it in. Even in the 70s Texas
Instruments could do a 1 chip version and they sound a lot better
these days.

Because 99% of the time speech synthesisers are utter crap. Take the one
at King's Cross St. Pancras for example, which used to announce a Good
service on the Pick-addle-y Line when first installed. (Though that's
now been fixed).

The main problems being either mispronunciation or a complete lack of
inflection where you'd normally expect it. At least where there's a
(human) automated announcer it generally sounds more "natural"...Though
you can't beat a *properly trained* "live" announcer.


A bizarre thing I heard on Southern recently was an obviously
synthesised voice saying "London Underground have informed me ...".

Oh, that winds me up, along with "I am sorry for the delay to this
service". You can't be sorry, you're a computer. "We are sorry" would
be far better.