gnuspeech-contact
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [gnuspeech-contact] IPA vs gnuspeech phonetic alphabet


From: David Hill
Subject: Re: [gnuspeech-contact] IPA vs gnuspeech phonetic alphabet
Date: Wed, 20 Jul 2016 11:58:36 -0700

Hi Paul,

There is a table of equivalences between IPA and gnuspeech orthography, along 
with examples and commentary, in the "Gnuspeech_Monet_Manual" in appendices B1, 
B2, and B3.

Do you simply want to substitute IPA symbols for gnuspeech symbols in the 
parsed Monet strings ready for speaking (for the benefit of readers)?

Or do you want connected utterances written in IPA, rather than in standard 
English orthography, to be translated for input to the speech conversion 
process? This latter goal would involve significant work to sort out the 
rhythmic structure (based on salient syllables -- information drawn from the 
dictionary in the text-to-speech process); and access to appropriate timing 
data for the successive phones. The latter could presumably be transferred from 
the gnuspeech phonetic entities to an equivalent IPA symbol table, or the IPA 
could be translated into normal gnuspeech script. But you could obviously 
bypass dictionary look-up itself, since you'd already have the pronunciation. 
Presumably the IPA would include primary and secondary stress markers to allow 
the foot structure to be determined to allow generation of the rhythm and 
intonation framework and production of the control characters. The actual Monet 
input should be gnuspeech phonetic entities (so you wouldn't then need to 
change anything following that).

If only isolated words, rather than connected utterances, were the goal, the 
requirements would be less stringent, and a standard control character matrix, 
with just the word embedded, should do the trick. Then you'd just have to 
convert the IPA to gnuspeech symbols which would be trivial.

Time for some serious code reading?? :-)

An interesting project that should be doable fairly easily, IMO. I presume the 
goal would be linguistic research.

All good wishes.

david

On Jul 19, 2016, at 19:50 09PM, Paul Tyson wrote:

> Hi,
> 
> Can anyone comment on the suitability, benefits, challenges, etc. of
> equipping gnuspeech with a parser for International Phonetic Alphabet
> input?
> 
> Thanks and regards,
> --Paul
> 
> 
> _______________________________________________
> gnuspeech-contact mailing list
> address@hidden
> https://lists.gnu.org/mailman/listinfo/gnuspeech-contact
> 
> 




reply via email to

[Prev in Thread] Current Thread [Next in Thread]