Comfort UK Voice
|
Comfort UK Voice
|
01-06-2007, 09:41 AM
We are planning to record a new voice for the UK, as some have complained that our standard voice is not \"english\" enough
We have 2 potential candidates here so that everyone can have the opportunity to review Charlie and Kate are the voices 
01-09-2007, 01:50 PM
In this form they both sound fine with my preference being Charile. The issue with these examples is that they do not replicate what the comfort voice programming and the keypad will do to the quality of playback.
I think part of peoples problem with the voice is clarity. The inconsistences in speed of speech and the gaps as the words are strung together or the wrong inflections etc. In that utopia, I think a \"text to speech\" engine would be great. Would be very easy to upload new voice styles etc. Additionally responses could then contain endless custom phases doing away for custom voices. It might even be that a TTS engine could be used to create a new recording rather than booking studio time? Examples: [size=2] http://actor.loquendo.com/actordemo/default.asp?page=id (As used in TomTom sat nav) http://public.research.att.com/~ttsweb/tts/demo.php [/size] http://www.neospeech.com/demo/demo_text.php http://www.nuance.com/realspeak/demo/
01-10-2007, 12:03 PM
Thats something to look into. Perhaps some of the voices can be licenced.
You seem to be a fount of information! 
01-25-2007, 08:03 AM
Does that mean we get Jon Cleese as a voice option like in TomTom :-)
01-25-2007, 08:39 AM
If Cytech and their customers really wanted John Cleese then it would be a case of paying the royalities and studio time. Considering the debate over the current voice I\'d guess that is going to be unlikley.
Would be a lovely thought that we could have that much flexibility but my suggestions above are really food for though for future versions of Comfort. The engine (electronics) is not there yet. No TTS engine or MP3 playback. On the older TomToms real voices and phrases were encoded but (and this is where I think the main difference is) they managed to keep the continuity and inflections in place. The comfort processing seems to loose some of this. The latest TTs now use TTS technology. My personal preference would be a TTS engine as this means that almost any word can be said rather than locking you in to specific ones. Of course there are good and bad versions of these and then being realistic the cost. Imagine the power of a Response \"Say Text <location> <text string>\". Where location is keypad, doorphone, ALARM SPEAKER and text string is a line of text. Anyway back to todays voice, whatever her name is.
05-16-2007, 07:24 PM
I am pleased to announce the creation of a new UK voice.
It is available from: http://www.cytech-technology.com/software_right.php?71 Note that the file is meant for use with Comfort II only. Further editing may be necessary for the words, this is just to get an initial feedback Regards
08-15-2007, 08:49 AM
Both Anita and Kate are in BIN format
VoiceProc and VoicePlay (Id like to hear them first) both require INI files. How can we get to hear the voice in action before any upload to Comfort ?
08-15-2007, 04:21 PM
I\'ve put Kate on to the system - works ok - although had to use Comfigurator 212 to get it working - 221 failed multiple times.
What I find tho is the voice is too slow. But, if you speed things up in comfigurator, she says the words faster, but thats not the issue - it\'s the pause length between each word that need shortening, but I dont see an option there. So, with things speeded up, she now says each word in a short sharp manner, but the gap between the words doesnt appear to change. Is there a way to shorten the gap between the words in a phrase, and not the speed of the word itself ? |
|
« Next Oldest | Next Newest »
|
