Phone (‘oir’) missing from unilex-gam?

[Zoë B]

Hello,

I’ve noticed a potential mismatch between the list of possible phones and one particular pronunciation in the lexicon:

coir::NN: { k * oir r } :{coir}:111

The phone ‘oir’ is absent from the phonelist for ‘General American English’ accents — though it only occurs once in the entire lexicon.

Is ‘oir’ a valid phone for GAM? Or am I perhaps misunderstanding how to read the Unilex dictionary?

I downloaded the Unisyn lexicon from https://www.cstr.ed.ac.uk/projects/unisyn/ and converted it into its ‘gam’ version using the provided Perl scripts. When accessing the pronunciation for ‘coir’ directly through Festival, I get the same pronunciation (i.e., including ‘oir’).

Thank you for your help!

[Korin Richmond]

The ‘oir’ phone is in the Festival phone set definition for unilex (see /Volumes/Network/courses/ss/festival/festival_linux/festival/lib/unilex_phones.scm).

Or do you mean the phone_list file that gets put in the alignment directory when doing force alignment with the multisyn build tools? If so, you’re right, that vowel doesn’t seem to be there. Who exactly created that file is now lost in the mists of time, but it *is* a very rare phone!

If that phone appears in the phone transcriptions of sentences you need to align with HTK/multisyn-build tools, then just add it to the phone_list file (in the “alignment” subdirectory).

[Zoë B]

Thank you, Korin!

Yes, I mean the file referenced under Section 8.1) of the ‘Build your own unit selection voice’ exercise. In my case, it’s ‘phone_list.unilex-gam’:

/Volumes/Network/courses/ss/festival/festival_linux/multisyn_build/resources/phone_list.unilex-gam

‘oir’ happens to be the only phone in the accent-specific dictionary that’s missing from the list. I stumbled upon it while trying to compile a list of ‘impossible’ diphones, i.e. diphones which do not occur in the dictionary (within words/across word boundaries).

[Author]

Posts