It's not your average audio editor. Adobe is working on a piece of audio manipulation software that has a peculiar artificial intelligence built in. It's so smart that it allows you to fix verbal mistakes in a way never seen before. This functionality can be so helpful and, at the same time, massively abused.
VoCo is the code name of the application presented at Adobe MAX 2016. In this live demo they took a voice recording of the co-host of the conference and ran it through the software. The app could interpret the audio and turn it into a text. OK, that's something we've already seen these days. It's when we use simple voice commands and our devices understand what we say. But this app goes a lot further. It interprets the audio and creates a voice profile. Then you can type any text you want and play it back with this person's voice as if they actually said that.
There are many possibilities this technology gives for the good of humanity. For example, you don't have to re-record a voiceover. You can simply fix the mistakes as if you fixed them in text. Slips of the tongue can also be easily fixed in post. However, there are many scary possibilities this app opens. A radio news host may no longer have their job because once there's a voice profile, the news can be automatically read by a machine. One could also manipulate audio recordings and submit it as evidence in a court. I'm sure you can think of many more cases.
I admit that this piece of technology is so fascinating. I envision in the near future we will say: "It's fake. The audio is voco-ed." Photoshop changed our visual world in a groundbreaking way. What will happen when we have Photoshop for audio?
Well, that's gonna save a lot of time editing interviews.
Yes, indeed. And celebrities can do radio interviews only by email :) Actually they don't have to write the email themselves at all.
Actually they will say "i never said that ! "
After they release such software no audio evidence in court will hold...
A lot of voice-over work will be lost to this, and technically illegally. I can see backlash from this the voice-over industry.
It depends on the quality of the software. From the demo I can't see a very realistic speech replacement but it could be the small sample they've created a profile from.
It's like the current 3D modeling used for food and interior shots. It's cool but it has its trade offs. It's like the VFX. People always like the real thing but sometimes budget constraints require the use of VFX.
Legally photoshopped images have the same issues as would "photoshopped" audio do.
It all depends on how it's contracted, and if union people are involved. Legal or not it will take work away from voice-over people, especially as the app evolves.
It might be happening.
http://motorcitymuckraker.com/2016/12/20/warren-mayor-fired-radio-show-r...