Glad to see that most automatic announcers are update with the new TOCS names - seems a bit strange at Birmingham New Street with London Midland and CrossCountry. Is computer System automatically updated or does Phil Sayer have to record it from scratch once for most stations and copy is sent out?
Its interesting though how the automatic announcer differentiate between a Virgin Pendolino and Virgin Voyager and Virgin Trains [Mks3].
The way that the system works is by having several sound files which the system strings together. When a franchise changes hands, the system gets updated pretty quickly and consists of changing the old TOC sound file to the new one in each train.
For example, originally there would have been a train like this in the system (using BHM as an example - though all the Ditras work in the same way):
The train now standing | at | platform 3A | is the | 10 | 57 | Central Trains service to | Walsall.
Now it would be:
The train now standing | at | platform 3A | is the | 10 | 57 | London Midland service to | Walsall.
Phil and Celia would record each new company name (in two different ways for each version of the software - one just needs 'London Midland' as 'service to' is a file by itself), and send it to Ditra.
The new sound file (in this case 'London Midland service to') would be put onto each of the systems and someone would have to change the way each train is set up (i.e. Central Trains service to --> London Midland service to).
The announcements only differentiate between the different Virgin services as it is the way they have been set up - obviously someone decided it would be a good idea to tell passengers what train they were getting so they were recorded, put into the system and used (rather than just 'Virgin Trains service to'). IIRC, Ditra has never differentiated between 'Virgin CrossCountry' and 'Virgin West Coast' services.
As for thefab444's comment, I suppose when Ditra decided to use professional voiceovers they advertised for them and the applicants were interviewed, auditioned, and shortlisted. In the end, Phil Sayer and Celia Drummond got the jobs.
When an update is needed they get the original artist to record the new phrase for the system rather than someone new - it takes a lot of time (and hence money) to record the names of 2500 stations twice, as well as all the other fancy stuff Phil can do.
I remember reading somewhere that the system dates back to 1996 - obviously there are bits that still haven't been updated and things are just patched over and hence sometimes it sounds 'odd'.
Phil can say 'Micheldever' correctly and always has done - Celia was corrected (from 'mit-chell-DEE-ver' to 'mit-chell-DEV-ver) after a complaint to SWT. Phil has pronounced most (if not all) odd places correctly (e.g. Leigh as 'Lie' rather than 'Lee') and I think he researches the pronunciation (obviously it isn't provided as Celia wouldn't make mistakes!).
