- General
-
Questions
import voices or changing them to match teens
Is it possible to import voices into Site Pal. None of the voices match any of the child or teen avatars. If not, what can I change to make the current voices sound like a teen/child? If I change the pitch, the voices sound robotic.
Answer
Hi JNJ -
You can't import voices - but you can get good results with pitch & rate adjustment by using SSML prosody tags in the text. They don't work the same way for all voices, best when used with engine 3 female voices.
Here are some examples - try these & let me know what you think.
1. Voice: Ashley - 3,1,6 <prosody pitch="+100%" rate="-5%"> My mom says too much chocolate is not good for you. So I won't give you any. </prosody>
Would I include these tags in the "sayText" function like
sayText ("<prosody pitch="+20%" rate="+5%"> My mom says too much chocolate is not good for you. So I won't give you any. </prosody>",6,1,3);
Yes exactly.
You can also use them in the text entered in the UI when you create audio in the editor.
I'm not using the editor, only the javascript API. I don't see any change when changing the pitch and rate values though.
sayText ("<prosody pitch='low' rate='+3%'> My mom says too much chocolate is not good for you. So I won't give you any. </prosody>",6,1,3);
is no different from
sayText ("<prosody pitch='x-high' rate='+3%'> My mom says too much chocolate is not good for you. So I won't give you any. </prosody>",6,1,3);
Got it. we're looking into it.
To hear what it is supposed to sound like try it here -
https://www.vocalware.com/index/demo
The JS API should sound the same.
I tested your link and I was able to tell the difference. I put the effect string into sayText and it sounds different from what it does on https://www.vocalware.com/index/demo.
It's as if the prosody tag is being ignored in the sayText function
Yes - we deployed an update this morning that fixed this problem.
This issue was missed in our new HTML5 characters, now corrected.
Thanks for calling our attention to it!
Gil
Customer support service by UserEcho
Hi JNJ -
You can't import voices - but you can get good results with pitch & rate adjustment by using SSML prosody tags in the text. They don't work the same way for all voices, best when used with engine 3 female voices.
Here are some examples - try these & let me know what you think.
1. Voice: Ashley - 3,1,6 <prosody pitch="+100%" rate="-5%"> My mom says too much chocolate is not good for you. So I won't give you any. </prosody>