Changing the voice used in Speech Synthesis

Hello,

I am a web developer and I am really interested and have been playing
around with the Web Speech API.

I think I understand the Speech recognition part, but Speech synthesis is
still somewhat troubling to me.

This is because the spec that I am looking at only has one example of
Speech Synthesis and is from 2012. Is there a more recent spec?

Please correct me if I am wrong, the voice Synthesised Speech is spoken is
dependent on what "voices" the user has installed, correct?

However, there is a way to detect what voices are installed and then choose
a voice from those installed, correct?

How do I do that? I can't find any examples on how to get a list of voices
installed and how to set the voice out of one of those voices? Is there a
way to serve a voice with the code so it is guaranteed it is spoken in that
voice?

How are Google Conversational Search results spoken in such a nice voice?
And the same voice on every time and every machine?

I would greatly appreciate it if someone could help me and answer these
questions.

Thanks in Advance!

Best Regards,
Irfan.