I am using the IBM Text to Speech to process some German texts. Pronounciation of the text with a German voice is working ok. However, the text contains some English words and phrases. For those, the pronounciation sometimes is incorrect. How can I fix that?
Example:
Er kommt aus Amerika und nennt sich TJ.
TJ
should be pronounced like someone from California would do.
I was able to make it work by using Speech Synthesis Markup Language (SSML) on my text. The German text remains the same. For the English expressions I provided some "phonemes". Here is the one for TJ
:
<phoneme alphabet="ibm" ph="tIZE">TJ</phoneme>
I had to use German symbols which can be found in the docs.