voice synthesis features

Official support for: xoxos
RELATED
PRODUCTS

Post

:o

That's sounding really snazzy so far. :O

Post

stanlea - thanks for sussing my link error! :O

link's fixed - http://www.breathcube.com/syng2demo01.mp3
you come and go, you come and go. amitabha neither a follower nor a leader be tagore "where roads are made i lose my way" where there is certainty, consideration is absent.

Post

birrbits wrote:this thread made me take my spoken language processing book of the shelf (mighty dusty). It seems I learned the phonemes using the sun representation, so I googled that and came up with this link, which may aid in your quest.

http://www.ibiblio.org/sounds/phonemes/
that list has a few phonemes which have no connection to the english language, so i presume they're purely american phonemes.

http://www.ibiblio.org/sounds/phonemes/AO
http://www.ibiblio.org/sounds/phonemes/ER

i guess they're like british english 'o' and 'oe.'
you come and go, you come and go. amitabha neither a follower nor a leader be tagore "where roads are made i lose my way" where there is certainty, consideration is absent.

Post

beautiful :)

it sounds like the mass-spring plosives make a real difference!

one thing which has always bothered me with speech synthesis is the dipthongs - morphing between formant signatures is just maths and not really what's-happening-in-the-mouth, if you see what i mean..

how close to real-world are your dipthongs and phonemes?
or shouldn't i have asked..? ;)

Post

CinningBao wrote: one thing which has always bothered me with speech synthesis is the dipthongs - morphing between formant signatures is just maths and not really what's-happening-in-the-mouth, if you see what i mean..
that's ~the gooey bubblegum effect you get with klaat-style synths. here the crossfade time is dynamically adjusted based on the phoneme, which reduces the duration of 'artifacts.'

for my diphthongs i think it works well, though evaluation varies with any given transition.

using a mass-spring for plosives isn't perfect.. there's one part in the file where the higher frequencies of a 't' are masked by percussion and only the lower frequencies are heard, so it sounds a bit like an 'n.'

for some consonants, like 'd' and 't,' i amplitude modulate the noise source by the mass-spring and blend the two, which is more stable/recognisable than using a short envelope of noise.

still, consonants are hard to analyse and recreate.
you come and go, you come and go. amitabha neither a follower nor a leader be tagore "where roads are made i lose my way" where there is certainty, consideration is absent.

Post

xoxos wrote:i don't know who chubby brown is.
chubby was the nearest vocal example for where i come from,but i guess the beatles accent is similar. uk and us robots always have soft 'u' in fact most vowel sounds are soft compared to northern ones.
i think,though i've not studied it all too deeply.i'll have to give it some thought.

Post

xoxos, I did answer in the other thread - maybe you can also add la "jota", spanish specific.
You can't always get what you waaaant...

Post

Hmmmnnn...it appears to me that interest would be quite a modulator (or unmodulator :hihi: ).

The "interest" slider/knob could dial in and/or half-arse of any presentation- just a like a real vocalist! :-o :dog: :D
I've got nothing to sell...am I on the right site?

Post

stanlea wrote:xoxos, I did answer in the other thread - maybe you can also add la "jota", spanish specific.
it would be better for someone familiar with french and english to evaluate the french phonemes, i don't know enough french. i'm not going to institute another set, but i will add a few extras :)
you come and go, you come and go. amitabha neither a follower nor a leader be tagore "where roads are made i lose my way" where there is certainty, consideration is absent.

Post

xoxos wrote:
that list has a few phonemes which have no connection to the english language, so i presume they're purely american phonemes.

http://www.ibiblio.org/sounds/phonemes/AO
http://www.ibiblio.org/sounds/phonemes/ER

i guess they're like british english 'o' and 'oe.'
The dusty book uses the examples of dog, lawn, and caught for 'ao'
and turn, fur and meter for 'er'

very cool demo btw
my sig will go here

Post

birrbits wrote: The dusty book uses the examples of dog, lawn, and caught for 'ao'
and turn, fur and meter for 'er'

very cool demo btw
somehow even living in the states i ended up with a british dictionary, which provides contrasting taxonomic amusement. in british english, 'dog' uses a short 'o' and 'caught' uses an 'aw.'

british english uses 'oe' for 'turn' and a schwa for the second vowel in meter, so there again there is a difference.

i've added breathy voices and 'sidechaining.' the external signal only replaces the glottal source oscillator, which means sybillants and plosives are still sourced internally, which is more understandable.

the 'external signal' consists of synths from the mix.. it's mono.

http://www.breathcube.com/syng2demo01vox2.mp3
you come and go, you come and go. amitabha neither a follower nor a leader be tagore "where roads are made i lose my way" where there is certainty, consideration is absent.

Post

btw it's supposed to have the long 'j' in 'technology..' i thought it sounded kind of cool and foreign.. a shorter 'j' sounds like the english pronounciation :)
you come and go, you come and go. amitabha neither a follower nor a leader be tagore "where roads are made i lose my way" where there is certainty, consideration is absent.

Post

For French, the french wikipedia page seems very complete but the sounds are in ogg format: http://fr.wikipedia.org/wiki/Prononciat ... n%C3%A7ais

This page from the University of Texas is a good summary with the a sound file for consonants, vowels, nasal vowels and semi-vowels including all reference words: http://www.laits.utexas.edu/fi/html/pho/02.html
Image

Post

nice work
my sig will go here

Post

Thanks Antoine, better than my post.
You can't always get what you waaaant...

Post Reply

Return to “xoxos”