WELLESNET MESSAGE BOARD

Orson Welles reads "The Raven" by Edgar Allan Poe (Speech Synthesis):
https://youtu.be/OQ4A7krOIqE

Even if they can get to the point of improving the artifact-laden quality of the voice (which comes across like over-compressed streaming audio from twenty years ago) to approximate a warmer, more human tone, they still have a long way to go to get a good "performance" out of the AI. This is the sound of a bored Welles distractedly reading a grocery list!

I agree with Roger's assessment above. This doesn't sound great. But I am still pretty impressed with it. The technology surely has a long way yet to go, but it does mostly sound like Welles (poorly) reading "The Raven." One could probably mix this (or something like this) with the voice of an actor who sounds like Welles and have the final product sound like a good performance by Welles. Eventually, we'll be able to re-create the narration from the lost reels of "Ambersons" with technology like this.

jbrooks wrote:...Eventually, we'll be able to re-create the narration from the lost reels of "Ambersons" with technology like this.

Despite my disappointment in this example (and, to be fair, the samples found here using a non-celebrity voice are significantly better), I immediately thought of Ambersons as well; not just a way to recreate the lost narration by Welles, but all of the missing dialogue from Cotten, Moorehead, Collins, et al.

Maybe they could create a soundtrack for The Deep too.

Here are some other speech synthesis readings of Welles:

Navy Seals copypasta:
https://www.youtube.com/watch?v=PqbkFdqDUmU

Timecube:
https://www.youtube.com/watch?v=EikJf1FaUA0

Norm Macdonald's Moth joke:
https://www.youtube.com/watch?v=rGjlus0htB8

Book of Genesis:
https://www.youtube.com/watch?v=Fj_M2Cm737A