I've seen games like Moonbase Alpha where they have TTS which looks like its way more fun to play with then the actual game. How complex do you guys think it is to set up a TTS system for sc2?
i dunno if this help , but the game `` MARCHENT RPG`` has text-to-speach in the game ,. like .( red, blue , epic , fail , go ,win ) all thes words and more when u type them , theirs a speach that repeats it after you type it . maybe you can ask the guy that made that game to tell you how he did it.
i dunno if this help , but the game `` MARCHENT RPG`` has text-to-speach in the game ,. like .( red, blue , epic , fail , go ,win ) all thes words and more when u type them , theirs a speach that repeats it after you type it . maybe you can ask the guy that made that game to tell you how he did it.
He's just got a datatable setup to play premade sounds when the predetermined text is registered in the chat string event. Which is basically what I said up above.
Probably better would be to have a record (We will call it speech) and each word would have a variable in speech named the same as the word it accounts for. So the sound file where someone says Hi would be named Hi etc. Then, when someone types something (Eg Hi, how are you?) it would pick each word, remove any punctuation (So Hi is Hi and not Hi,) and play a sound of Speech.<word>. So it would play Speech.Hi then Speech.How then Speech.Are etc. If it can't find it then it plays Speech.s_Null (A pause of the average length of a word. You could also add full stops and comma pauses). Then just add each word to the variable and it should play what the person types.
You would have to have a loooooot of space though for all those sound files :/. Even better would be to have syllables and somehow determine what syllables go in what word and construct the words from that, but that would be really hard to do and the code for it would be massive.
duno if its possible in sc2 to access a 3rd party website, but there are several websites that you type in what u wana say, and an audio file is made or whatever..., ofc as said before the actual phonetic pronounciation and stuff is still fairly limited even with the thousands of manhours put into it globally
Rollback Post to RevisionRollBack
To post a comment, please login or register a new account.
I've seen games like Moonbase Alpha where they have TTS which looks like its way more fun to play with then the actual game. How complex do you guys think it is to set up a TTS system for sc2?
@DrHu: Go
Why and how?
@Vexal: Go
why? just bringing the idea to the table. and how can this be done? first got to find someone who happens to be an expert of the English language.
@DrHu: Go
Decades of research have been put into current text-to-speech technologies and it still has a lot of problems.
Best you can do is assign sounds to variables then pull them dynamically I.E.
Wave Sound[1] = "guysaying1.mp3"
Wave Sound[2] = "guysaying2.mp3"
Wave Sound[3] = "guysaying3.mp3"
Play Sound "Guy saying Wave.mp3"
Play Sound Wave Sound[WaveNumber]
etc
@Dustin374: Go
That's not the best you can do. In fact, that's probably close to the worst you can do.
The best you can do is a real, working, viable text-to-speech system.
EDIT: Actually, the worst you could do is probably a blue-screen.
You need to study linguistics.
Decades you say? I think ill give it a shot anyways with the help of pronunciation symbols and look for basic letter combination.
i dunno if this help , but the game `` MARCHENT RPG`` has text-to-speach in the game ,. like .( red, blue , epic , fail , go ,win ) all thes words and more when u type them , theirs a speach that repeats it after you type it . maybe you can ask the guy that made that game to tell you how he did it.
He's just got a datatable setup to play premade sounds when the predetermined text is registered in the chat string event. Which is basically what I said up above.
@Dustin374: Go
Probably better would be to have a record (We will call it speech) and each word would have a variable in speech named the same as the word it accounts for. So the sound file where someone says Hi would be named Hi etc. Then, when someone types something (Eg Hi, how are you?) it would pick each word, remove any punctuation (So Hi is Hi and not Hi,) and play a sound of Speech.<word>. So it would play Speech.Hi then Speech.How then Speech.Are etc. If it can't find it then it plays Speech.s_Null (A pause of the average length of a word. You could also add full stops and comma pauses). Then just add each word to the variable and it should play what the person types.
You would have to have a loooooot of space though for all those sound files :/. Even better would be to have syllables and somehow determine what syllables go in what word and construct the words from that, but that would be really hard to do and the code for it would be massive.
duno if its possible in sc2 to access a 3rd party website, but there are several websites that you type in what u wana say, and an audio file is made or whatever..., ofc as said before the actual phonetic pronounciation and stuff is still fairly limited even with the thousands of manhours put into it globally