Marked Tonicity Audio Samples
-
Don't forget what they said yesterday,twopolicemen is excessive.
Reference Human Speech Copy Style and Voice Copy Style, but convert to different Voice Just copy the Voice, ignore style -
I said,Ihave won, when they asked.
Reference Human Speech Copy Style and Voice Copy Style, but convert to different Voice Just copy the Voice, ignore style -
Everybody here in this room can agree, thattwolovebirds is annoying.
Reference Human Speech Copy Style and Voice Copy Style, but convert to different Voice Just copy the Voice, ignore style -
As soon as we arrived at the farmyesterday, we saw that they were milking cows.
Reference Human Speech Copy Style and Voice Copy Style, but convert to different Voice Just copy the Voice, ignore style
Preserving Meaning in Resynthesis
[sarcastic]
Wow, what a beautiful day! Reference Human Speech | |
Naive Resynthesis | |
Different voice, same Prosody | |
Voice Unification
»Quellen hattest du ihm, hattest dem Flüchtigen kühle Schatten geschenkt.«
Reference 1 Poetry Reading | |
Reference 1 converted to shared Voice | |
Reference 2 Poetry Reading | |
Reference 2 converted to shared Voice | |
Audiobook to Audioplay
STAGE INSTRUCTIONS:
Ghost beckons Hamlet
HORATIO:
It beckons you to go away with it, as if it some impartment did desire to you alone.
MARCELLUS:
Look, with what courteous action it waves you to a more removed ground: But do not go with it.
HORATIO:
No, by no means.
HAMLET:
It will not speak; then I will follow it.
HORATIO:
Do not, my lord.
HAMLET:
Why, what should be the fear? I do not set my life in a pin's fee; and for my soul, what can it do to that, being a thing immortal as itself? It waves me forth again: I'll follow it. Reference Human Speech | |
Copy Style, but give each Character a unique voice | |
Copy Style, but give each Character another unique voice | |
Just for Fun: Playing all 3 simultaneously | |
Copyright Notices
The text we used in the Audiobook to Audioplay section is taken from William Shakespeare’s Hamlet.
The reference samples from the Marked Tonicity section is taken from the ADEPT corpus (https://arxiv.org/abs/2106.08321)
The voice used for generating the samples in the Voice Unification section is based on Hanns Zischler’s recordings which come with the following copyright notice:
Florian Michel (Ton) und Hanns Zischler (Sprecher) im Auftrag des Deutschen Literaturarchivs Marbach (2020)