James Farrar wrote:
It depends where the breaks between phrases are. The problem is that
the breaks between fixed words and variable words aren't where real
people naturally breathe. So instead of a natural "this is a
Piccadilly line train / to Cockfosters", you get a very
unnatural-sounding "this is a Piccadilly line train to ///
Cockfosters", unless the person doing the recording is told where to
breathe.

But this is speech synthesis, and the science is pretty mature now
considering a decent effort is available on a SatNav that costs under
£400, or mobile phones that cost £200 or less (and free on a
contract). TomTom does a pretty good job speaking clearly (see
http://www.tomtom.com/products/produ...tegory=0&Lid=1 for
a chance to hear for yourself). In fact, why not compose an example
announcement to see if it does a better job than the lame LUL version!
Nokia has software that can read text messages, although admittedly
this is rather hit and miss.

Navteq now stores phonetic information for all place names/streets on
their maps, which should improve the pronounciation even more. Surely
LUL could at least ensure that the system can know exactly how to
pronounce station names and different lines properly and clearly!!

As I said before, it's just a matter of tweaking the software. Once
this is done, it's an excellent way of ensuring passengers can get all
the information they could ever need. I am sure LUL would agree with
everything I have said too.

Jonathan