New voice CIS on SWT

[swt class 450]

If that is the case then why did they get Celia to re-record the announcements and not get "Amey woman" to do them instead? I personally think that "Amey woman" sounds better than these new re-recorded Celia announcements although I agree that I think both are better than the "Anne" announcements.



Could they (Stagecoach) know something we don't?

Perhaps Stagecoach have already been told that they have won the new franchise? Is it possible they have been told before it gets announced to the public?

I imagine if First win they will probably replace the Phil Sayer and Celia Drummond station announcements with Anne Jermy (Atos Anne) and the on board Celia Drummond announcements with Faye Dicker (like they have done on their GWR franchise).

 

[abn444]

Is it possible they have been told before it gets announced to the public?

HH suggested that could be possible over on the SWT franchise thread.

 

[NorthernSpirit]

Whose the woman that is used in Leeds, Sheffied, Trowbridge, Westbury, London Vic and probably a dozen or so other places?

Going by by the MP3 files on the other page, I've noticed that the announcements have been slowed down but there's probabaly a good reason. Shame that GWR didn't do the same as it would give an almost uniformed voice or voices (in the way that Southern has done with their Celia Phil combos) across the network.

 

[capital12]

Have to say I find the Celia/Phil combo the most clear announcements of them all, these new ones do sound too slow (to me) and leave me thinking 'just get on with it' while listening!

Although the worst has surely got to be GWR with that god awful male voice that sounds completely moronic!

 

[Parallel]

Whose the woman that is used in Leeds, Sheffied, Trowbridge, Westbury, London Vic and probably a dozen or so other places?

Going by by the MP3 files on the other page, I've noticed that the announcements have been slowed down but there's probabaly a good reason. Shame that GWR didn't do the same as it would give an almost uniformed voice or voices (in the way that Southern has done with their Celia Phil combos) across the network.

I think the one in use at London Vic is different to the one at Trowbridge and Westbury. Trowbridge and Westbury definitely use Anne.

Have to say I find the Celia/Phil combo the most clear announcements of them all, these new ones do sound too slow (to me) and leave me thinking 'just get on with it' while listening!

Although the worst has surely got to be GWR with that god awful male voice that sounds completely moronic!

The only thing worse than the GWR male station announcer's voice is the one onboard refurbed 150/2s and 158s. They've used a separate voice to record the 'west' station names too, so it says something like "The next station is... FRESHFORD?" And I want to claw my face off every time the announcement about the doors comes on when the train stops at a station. Awful.

There seems to be a problem relating to the 158s too, for some reason the system doesn't announce when a train is about to arrive at a station. Very poor seeing as vocal announcements are meant to aid passengers who are hard of hearing. The 150s however do announce when the train is approaching a station.

 

[godfreycomplex]

Nobody seems to know how Micheldever is pronounced! I regularly hear both MITCHELL-DEAVER (to rhyme with "beaver") and MITCHELL-DEVER (to rhyme with "never")!

Rhymes with never. The real dilemma locally is whether it's pronounced "Michell" (as in Roger) or "Mitchell" (as in Phil)

 

[HowardGWR]

Rhymes with never. The real dilemma locally is whether it's pronounced "Michell" (as in Roger) or "Mitchell" (as in Phil)

Never heard of the first, wish I hadn't heard of the second.

Do you say the first as 'Michelle my belle'?

 

[swt class 450]

Whose the woman that is used in Leeds, Sheffied, Trowbridge, Westbury, London Vic and probably a dozen or so other places?

Going by by the MP3 files on the other page, I've noticed that the announcements have been slowed down but there's probabaly a good reason. Shame that GWR didn't do the same as it would give an almost uniformed voice or voices (in the way that Southern has done with their Celia Phil combos) across the network.

The automated announcements at Leeds / Sheffield / Trowbridge / Westbury are done by Anne Jermy (Atos Anne) who seems to be appearing everywhere. She is probably the most widely used automated announcer now. The ones at London Victoria are a different unknown female who for now is just known as "Amey". However Atos Anne does now do the safety and security announcements at London Victoria station.

Have to say I find the Celia/Phil combo the most clear announcements of them all, these new ones do sound too slow (to me) and leave me thinking 'just get on with it' while listening!

Although the worst has surely got to be GWR with that god awful male voice that sounds completely moronic!

The new GWR male voice is Rodger (not to be confused with Roger who was at most Southeastern stations until around 2011 or 2012). He does sound very robotic.

I think the one in use at London Vic is different to the one at Trowbridge and Westbury. Trowbridge and Westbury definitely use Anne.




The only thing worse than the GWR male station announcer's voice is the one onboard refurbed 150/2s and 158s. They've used a separate voice to record the 'west' station names too, so it says something like "The next station is... FRESHFORD?" And I want to claw my face off every time the announcement about the doors comes on when the train stops at a station. Awful.

There seems to be a problem relating to the 158s too, for some reason the system doesn't announce when a train is about to arrive at a station. Very poor seeing as vocal announcements are meant to aid passengers who are hard of hearing. The 150s however do announce when the train is approaching a station.

So is the voice on the refurbished 150s and 158s is a different person to Faye Dicker on the 165s and 166s and 387s units? Do you have any recordings or videos of the new 150 and 158 automated announcements?

 

[Parallel]

So is the voice on the refurbished 150s and 158s is a different person to Faye Dicker on the 165s and 166s and 387s units? Do you have any recordings or videos of the new 150 and 158 automated announcements?

The generic parts of the announcements I believe are Faye Dicker, just the recording of the station names in the west sound like they have been done by someone else - I'll try to grab a recording when I can.

 

[swt class 450]

The generic parts of the announcements I believe are Faye Dicker, just the recording of the station names in the west sound like they have been done by someone else - I'll try to grab a recording when I can.

Ok. Thanks. That seems crazy to get two different people to do the announcements! I dont know why Faye Dicker couldnt announce the station names in the West!

 

[greaterwest]

The generic parts of the announcements I believe are Faye Dicker, just the recording of the station names in the west sound like they have been done by someone else - I'll try to grab a recording when I can.

All of GWR's on-board announcements are Faye Dicker I believe.

 

[Parallel]

All of GWR's on-board announcements are Faye Dicker I believe.

Might be because the West names have been recorded at a later time to the rest of the announcements, so don't flow very well.

 

[Bletchleyite]

One thing I've wondered in all of this is why someone (possibly even them, if it's patented) hasn't come up with an AviaVox style system for rail announcements? The closest I've heard is the fully synthesized voice TfL use for some "engineering work" type announcements at some stations - but it is obviously fully synthesized and sounds quite odd.

For the uninitiated, Aviavox is the woman with a Dutch accent (actually one of the Schiphol manual announcers, it's a bit odd when you hear her speaking more naturally) and the characteristic bing-bong you get at most airports. It differs from the typical UK systems in that rather than recording phrases from a person, you record a phoneme, and it then strings the phonemes together to announce anything, even a word it hasn't heard before.

Building on top of that, if you do try to announce something it hasn't heard before, it feeds that phrase (and the phonemes) it uses back to the company's HQ, who then analyse it, see if it sounds right, and if it doesn't either select which phonemes it should use or record a special one.

While the Dutch accent grates a bit (I don't understand why they didn't use a native English speaker to record it; they did for other languages it does) it is really quite effective and very flexible - for example, a new TOC name would be put together automatically.

More info: http://aviavox.com/#watch-our-video

Or the video: https://www.youtube.com/watch?v=usT0efA6XLk

 

[southern442]

One thing I've wondered in all of this is why someone (possibly even them, if it's patented) hasn't come up with an AviaVox style system for rail announcements? The closest I've heard is the fully synthesized voice TfL use for some "engineering work" type announcements at some stations - but it is obviously fully synthesized and sounds quite odd.

For the uninitiated, Aviavox is the woman with a Dutch accent (actually one of the Schiphol manual announcers, it's a bit odd when you hear her speaking more naturally) and the characteristic bing-bong you get at most airports. It differs from the typical UK systems in that rather than recording phrases from a person, you record a phoneme, and it then strings the phonemes together to announce anything, even a word it hasn't heard before.

Building on top of that, if you do try to announce something it hasn't heard before, it feeds that phrase (and the phonemes) it uses back to the company's HQ, who then analyse it, see if it sounds right, and if it doesn't either select which phonemes it should use or record a special one.

While the Dutch accent grates a bit (I don't understand why they didn't use a native English speaker to record it; they did for other languages it does) it is really quite effective and very flexible - for example, a new TOC name would be put together automatically.

More info: http://aviavox.com/#watch-our-video

Or the video: https://www.youtube.com/watch?v=usT0efA6XLk

I think the Heathrow T5 satellite transit train's on board announcements uses this system. It seems to work quite well, perhaps a native English speaker could record some for the UK network, it would be interesting to see the results. At present there is at least one Text To Speech system in use at various stations in the UK, particularly around the East Midlands, although this isn't really the same and is far more unclear and garbled than your example. The closest thing to this I can think of is the on-board announcements used by DLR, which I have heard are TTS but sound quite clear and flow very well.

 

[Uzair]

Just listened to the new recordings and I don't like them at all! However, they are clearer.

 

[Bletchleyite]

I think the Heathrow T5 satellite transit train's on board announcements uses this system. It seems to work quite well, perhaps a native English speaker could record some for the UK network, it would be interesting to see the results. At present there is at least one Text To Speech system in use at various stations in the UK, particularly around the East Midlands, although this isn't really the same and is far more unclear and garbled than your example. The closest thing to this I can think of is the on-board announcements used by DLR, which I have heard are TTS but sound quite clear and flow very well.

Aviavox is of course a phoneme-based TTS system and not the only one on the market, but the thing that sets it apart is the back-end service they provide to ensure quality of the output even for words it hasn't heard before.

As an example, the Google/Waze synthesized voice always mispronounces Bletchley - it can't do the "tch" sound properly and sort-of slurs it. In Aviavox that would have been fixed the very first time it was used without any need to report it.

 

[southern442]

Aviavox is of course a phoneme-based TTS system and not the only one on the market, but the thing that sets it apart is the back-end service they provide to ensure quality of the output even for words it hasn't heard before.

As an example, the Google/Waze synthesized voice always mispronounces Bletchley - it can't do the "tch" sound properly and sort-of slurs it. In Aviavox that would have been fixed the very first time it was used without any need to report it.

Indeed, if you listen to several recordings of the 'Acapela Rachael' system that is in use at a few stations around the country, it sounds horribly disjointed, and someone who doesn't know the network well around that area could potentially miss a stop or something due to the very bad pronunciation of place names. Something like Aviavox would be much better in this case.

 

[rg177]

Indeed, if you listen to several recordings of the 'Acapela Rachael' system that is in use at a few stations around the country, it sounds horribly disjointed, and someone who doesn't know the network well around that area could potentially miss a stop or something due to the very bad pronunciation of place names. Something like Aviavox would be much better in this case.

Highlights include the one at MetroCentre pronouncing Seaham as "Sea" and James Cook as "J C H"

Sent from my SM-A310F using Tapatalk

 

[swt class 450]

Acapela Rachael sounds very robotic and not very natural in my opinion. Southeastern have programmed their Acapela Rachael announcements very poorly. For example she says "St Leonards W S" instead of "St Leonards Warrior Square" and she says "W St Leonards" instead of "West St Leonards" and there are probably others as well.

 

[Bungle965]

Acapela Rachael sounds very robotic and not very natural in my opinion. Southeastern have programmed their Acapela Rachael announcements very poorly. For example she says "St Leonards W S" instead of "St Leonards Warrior Square" and she says "W St Leonards" instead of "West St Leonards" and there are probably others as well.

Acapela Rachael is extremely poor, when she got put into Smithy Bridge she announced Manchester Victoria as Manchester Vic and Todmorden as Tod-MOR-don.
Horrible and disjointed.
Sam

 

[Thisisajm]

Same wording and voice artist, but entirely new recordings. If louis97 spots this thread I'm sure he'll post his recordings that were otherwise posted privately in the automated announcements section.

Are there any specifics to joining this section? I get prompted give a reason for joining.

 

[greaterwest]

Are there any specifics to joining this section? I get prompted give a reason for joining.

Nothing in particular, I don't even remember what I put for my request, or what the criteria is for joining.

 

[swt class 450]

Nothing in particular, I don't even remember what I put for my request, or what the criteria is for joining.

Do you know how long it takes to get accepted to the group? I applied just over a week ago but i dont think i have been accepted yet.

 

[Parallel]

Whereabouts do you apply for this section? Feel free to drop me a DM. I've got quite a few announcements saved to my phone.

 

[swt class 450]

Whereabouts do you apply for this section? Feel free to drop me a DM. I've got quite a few announcements saved to my phone.

You have to click on "User CP" and then click on "Group Memberships" and then you can select the group you want to apply to join.

 

[Jordy]

I've approved the requests that were waiting!

 

[swt class 450]

I've approved the requests that were waiting!

Thank you! Sorry if it is a silly question but how do i access the group now i have been approved?

 

[swt class 450]

Thank you! Sorry if it is a silly question but how do i access the group now i have been approved?

Actually dont worry i have found it now! I didnt realise the section was at the bottom of the page!

 

[Parallel]

You have to click on "User CP" and then click on "Group Memberships" and then you can select the group you want to apply to join.

Thanks! - I've just requested to join.

 

[Thisisajm]

Also so have I.