Some more experimentation with the SSML tags, a set of soothing whispering voices, in spanish. Transition from normal voice to whispering voice is not always super natural but there were ways to improve it slightly.
As correctly guessed, this is Polly. The level of customization in the mark up language is fairly decent. Although, I would like to get her to show some temper.
While having spoken words serves a tangible purpose in its own right. Overarching goal here is to find ways to hook the lady into something like Blender ( the chat-AI, not the 3D modeling suite ). Its worth doing if the response returned can be transformed to proper sounding voice and spoken through the VaM model. This part seems achievable.
Hi Guest!