TTS assets/modules in unity

I'm currently designing a game that will have a consistent amount of text. Given my own experience reading, I've decided to research using Text-to-Speech (TTS) to create a more engaging mean to reach the player.

This text can include unknown elements such as names, acts, etc. that can be pulled from a large list or entered directly by the player. So I would need a solution that can do TTS on the fly.

In terms of preferences outside of this:
+Having as natural voices as possible
+Having access to a variety of voices
+Be able to handle expressions and emtions
+(minor) Not rely on internet, would be nice, but not a must

I've looked at using other solutions such as RT-voice, but the demo couldn't detect any voices despite me having many IVONA 2 voices. Highlighting a weakness in relying the end-user's voices.

Many thanks to any thoughts and help that can be given on this.

 

It reminds me of this:

Spoiler

Using TTS to make a game sound is not something I saw done a lot.

You should probably try amazon polly. They do not offer built-in unity solution, but would allow you to create voice recording for the text you provided.

As far as I remember, expression and tones are handled through SSML, so if the TTS system supports SSML, it should be able to hanlde that to some degree.

There are some opensource engines such as eSpeak that could be duct-taped to unity on way or another, but creation of a new voice is not trivial, quality isn't great and there aren't a lot of voices.