SAPI phoneme data

[JRowe]

Is there a way to grab the phoneme data for a line of text? For example,

D @ 0 k w I k b r aU n f Q k s dZ V m p t @U 1 v @ D @ 0 l eI 1 z i: 0 d Q g 1

is this:

The quick brown fox jumped over the lazy dog.

I'd like to input a string and get a phoneme string in return.

What I'd like to do is have a talking head lipsync to entered text. I can do the animation, the 3D and the rest, but SAPI stuff is new to me. The idea would be to create animations for phonemes, queue the animation sequence according to the phoneme setup, and then fire both events, so that the head would appear to be speaking synced with the speech data.

Any help would be appreciated!

[JRowe]

It looks like this is pretty much done, and the phoneme to mouth position relationship is encoded in Visemes. Which are helpfully denoted and automated by the SAPI already.

http://www.autoitscript.com/forum/index.php?showtopic=70079

Progandy has a wonderful little script here that more than meets my needs. Thanks for any searching or interest, I think that the script was it though

For search reference:

Viseme, Phoneme, Animation, Lip-Sync Synching Speech TTS Text to Speech.

expandcollapse popup

SP_VISEME_0
       Silence
   SP_VISEME_1
       ae, ax, ah
   SP_VISEME_2
       aa
   SP_VISEME_3
       ao
   SP_VISEME_4
       ey, eh, uh
   SP_VISEME_5
       er
   SP_VISEME_6
       y, iy, ih, ix
   SP_VISEME_7
       w, uw
   SP_VISEME_8
       ow
   SP_VISEME_9
       aw
   SP_VISEME_10
       oy
   SP_VISEME_11
       ay
   SP_VISEME_12
       h
   SP_VISEME_13
       r
   SP_VISEME_14
       l
   SP_VISEME_15
       s, z
   SP_VISEME_16
       sh, ch, jh, zh
   SP_VISEME_17
       th, dh
   SP_VISEME_18
       f, v
   SP_VISEME_19
       d, t, n
   SP_VISEME_20
       k, g, ng
   SP_VISEME_21
       p, b, m

Edited February 2, 2009 by JRowe