Sign In
Register
Help
This topic is locked
Prince Syo, on 16 February 2012 - 09:52 PM, said:
First of all. You have no idea what those terms mean seeing as you can't properly use them.
PHONETICS: The representation of sounds through written symbols
DIAPHONE(ME): A sound in a language that varies by dialect. For example the different pronunciations of the word "eye" in American, English, Australian, Irish and Scottish pronunciations.
Example: [_ e][e g][g z][z e][e @][@ m][m p][p @][@ l][l _]
That's how a Vocaloid 2 voicebank would put the sounds together. (If you alter it to an American accent, which would use /e@/ instead of /{/)
In Vocaloid 3 it would be [_ e][e g][gz e][e@ m][mp @][@l _] It would require less files to be put together but require much more recording due to the vast amounts of consonant clusters in English.
If you ask what way SeeU's korean voicebank would create English, it would be more similar to the first due to the fact that Korean contains almost no consonant clusters in comparison to English, the largest cluster possible in Korean being two consonants put together (an ending consonant and an initial consonant coming after it) while in English a single syllable can have a shit ton of consonants at the beginning of it.
In SeeU's voicebank, the word example would be best written out as:
[e][g z e 7 mp][p 7]
which would be put together by the sounds:
[_ e][e g][gz e][e 7][7 mp][mpp 7][7 _]
Note that SeeU's voicebank uses atleast two-part consonant clusters due to the syllable syntax of Korean while the Vocaloid 2 method of puting English consonants treats each phoneme transition individually, thus making the output less clear in the latter.
TLDRx2: Bro you have no idea what you are talking about.
PHONETICS: The representation of sounds through written symbols
DIAPHONE(ME): A sound in a language that varies by dialect. For example the different pronunciations of the word "eye" in American, English, Australian, Irish and Scottish pronunciations.
Example: [_ e][e g][g z][z e][e @][@ m][m p][p @][@ l][l _]
That's how a Vocaloid 2 voicebank would put the sounds together. (If you alter it to an American accent, which would use /e@/ instead of /{/)
In Vocaloid 3 it would be [_ e][e g][gz e][e@ m][mp @][@l _] It would require less files to be put together but require much more recording due to the vast amounts of consonant clusters in English.
If you ask what way SeeU's korean voicebank would create English, it would be more similar to the first due to the fact that Korean contains almost no consonant clusters in comparison to English, the largest cluster possible in Korean being two consonants put together (an ending consonant and an initial consonant coming after it) while in English a single syllable can have a shit ton of consonants at the beginning of it.
In SeeU's voicebank, the word example would be best written out as:
[e][g z e 7 mp][p 7]
which would be put together by the sounds:
[_ e][e g][gz e][e 7][7 mp][mpp 7][7 _]
Note that SeeU's voicebank uses atleast two-part consonant clusters due to the syllable syntax of Korean while the Vocaloid 2 method of puting English consonants treats each phoneme transition individually, thus making the output less clear in the latter.
TLDRx2: Bro you have no idea what you are talking about.
I'm not going to argue mostly because it will derail the track, plus I have my mind on other things, if it makes you happy "you win" I can't be bothered to argue since its a waste of time. I'm quoting things I learnt years ago and I'm not as sound on nor do I wish to really get back into it. Plus arguing on the internet is generally a waste of time and this is still small talk. This will this effect SeeU's voicebank? No. Give me a topic dedicated to this, I'll debate with you on this matter. Quoting stuff I learnt several years ago only gives me a headache and I admit I make mistakes, what I don't know this stuff off by heart anymore. And you will always win in terms of X-AMPA conversion because the Vocaloid system is not something I handle well (hence why my Voclaoids never have perfect English anyway, then again I don't use mine to sound natural anyway).
This also does not change the fact I can't argue back on SeeU because I don't know Korean phonetics nor do I have a interest to learn about that ppart of Vocaloid. I knwo their two different languages they don't work the same and that SeeU wil have a Korean accent evem f a full English voicebank. What I am curious about and impatient on, is to learn what affect it will have on english because this is a new thing to discuss nd another example for us to use to describe oneEnglish capable Vocaloid against another for comparisons.
Virtual > Real has said anything elser I want to say on Korean English.... Its never going to sound as good as a native speaker but that doesn't mean we won't learn anything new from it.
MultiQuote
