Hi Martin,
I am here again to request a small additional feature in
Speech Output. Previously I have ever request about
Speech Output to sound file.
This time, while developing the speech flow, I am dealing with multiple language. In Input Speech, the
Language field support variable, So it is possible for me to use part of the map database to change the input speech language as I want (using {lcode}). But in Speech Output, the
Voice field doesn't support variable yet. I can only pick one of them.
- Speech Output variable.png (124.39 KiB) Viewed 7407 times
I use 5 languages in my flow. So to cover all possibility, I have to use a pair of expression and Speech output for each language, resulting in total 10 elements, linking all of them in parallel expression. This number will increase if I use more languages. It works, but takes up too much space. And each round of execution will consume 6 elements at once (5 parallel expression check + 1 true speech output). It makes me hit the AES easily (yes I hit the default 60 within 20 seconds).
- Multiple Speech Output.png (87.4 KiB) Viewed 7407 times
If the Voice field support variable,
then I only need single Speech Output, instead of 5 pairs of expression-speech output. Since I can change the voice language just in the script before passing it as variable to speech output. No matter how languages I wanna use, I always need single Speech output. I only need to add the additional language to my main mapping definition. Save the space and save the element execution, thus lower AES.
Since I have requested the
Voice field, I think if it is possible, the
Audio Stream Type should be using variable too. So I can control the speech output volume by changing the stream type, instead of changing the volume itself.
Regards,
Desmanto
