http://www.fileden.com/files/2008/4/21/1876835/TSPEECH3.BIN Samples aren't optimal yet, but its a fairly good beginning IMO :v: Regen or Fusion only (Fusion preferred). Should work on real HW, not yet tested it.
Neat, though I only understand the first "1 2 3". I tried it in both Gens and Kega and it sounded pretty much the same.
I'm going to guess the person who recorded the samples isn't a native English speaker. Might wanna re-do those with someone who is simply for sake of clarity. Nice start though! I could make some words out. :v:
Hahaha, awesome! It sounds a lot like an old Spectrum game I used to have, called Cobra's Arc... except that Spanish is much easier to reproduce phonetically, of course. Here's the text version. =P
Spanish has more ways of spelling things and thus more complex. That ununderstandable text string is how you have to write Estonian to get approximate English pronounciation... (and that capital N in the beginning of string is not part of it) Yeah, I do not speak english natively... and this is only my first alpha, 2nd version would sound much better. It will still have some of its issues because its like 100 bytes of code and some 65KB of samples :P Anyway, I hope to clear out the samples by tomorrow and then things should be good enough. Not as good as "Microsoft Sam" thing, but my synth is just so simple and works entirely around Estonian pronounciation system. oh, and how things sound depend on how I talk English myself, and I have quite strong accent :P
Just tried it out. It's kinda cool. Mac OS X: It works in Generator and Kega with the same audio quality; but doesn't play a thing in Genesis Plus. Which is odd, because Genesis Plus is supposedly at the same level as Kega, but I think I did hear from some people that Genesis Plus's sound emulation isn't the hottest thing in the world. Good luck with this!
Record my voice for the samples. Nah, kidding, people complain my voice is 'too deep' anyway. Anyway, for people that do not completely understand it yet: But yeah, I am loving this, it does indeed need more work in the voice samples, but after listening to it several times, it begins to make sense. And yes, I did steal ICEknights translation as a base, but I still needed to figure out some words by myself anyway .
The problem with the voice synthetizer is that Tiido just recorded his voice and slapped parts of what he recorded directly. And moreover, this isn't exactly the best of his recordings (I was talking to him about this). If someday he manages to have enough time, he should try recording his voice (maybe several times, heck, even the voice from other people too), see what waveforms are generated and create new samples based on the common patterns of those (which of course would as well join perfectly). Then it'll sound good. Add pitch and volume control (dammit, even tempo, these waveforms are played like a single-channel module!), and pretty much you have a quite good control over the voice (making it not being so emotionless, change the gender, etc.).
Hmm? We have only one way to read each letter, just like in Japanese. It's much more simple than English, where each letter can be pronounced quite differently depending on the case (gaol, anyone?). There's always exceptions, but they're only a few cases.
In Estonian its same, a letter in any word is always pronounced same (with few exceptions). So you guys can potentially make it speak English or any other language after a while :P
of course I agree with him xD, as english it's not pronounced as written... maybe, tiido, you should try an spanish version xD by the way, I got a headache.. :S and I think this needs some phonetic knowledge
Mmm... I don't think Spanish is easier, think this: We (native spanish speakers) use "Perro" for a male dog, or "Perra" for a female dog... also, we got "La perra" o "El perro", in english you use "The dog" and it doesnt matter the gender... Also, we got the volcals with " ´ " ... and that is what gives the pronunciation... is not the same to say "Por qué?" or "Porque" ...almost the same word for different things... Heck, we also got some hard tenses like "Preterito plus cuan perfecto" ... (No me acuerdo si se escribía así xD)... Altough english got tenses like past perfect its not the same thing you know...
A volume change can deal with it really (think on it as a higher note on attack if you consider the phonems as notes in a song). And everything else has absolutely nothing to do with how written->spoken stuff is done.
I think it was pretérito pluscuamperfecto and it does exist in english. Yo LO había hecho. I had done IT. and the "Tilde", well that is a problem..
