Photoshop for sound: new audio editor Adobe imitates the voice of any person after 20 minutes of training
Adobe has provided the public with unique audio editor, codenamed VoCo. The product is able to provide approximately the same capabilities for audio processing, which gives Photoshop for photos. This statement “audiofoto”.
One of the key functions VoCo is edit the voices so that the user can insert in the audio stream of the word that actually was not said by the source. Moreover, this procedure does not require any knowledge. Simply enter text in the text box and the program will automatically generate the audio voice of the original.
VoCo enough 20 minutes to listen to of any person (and if more, if not perfectly), after which he is able to utter the voice of anything. In other words, VoCo breaks the speech into separate phonemes, which are then able to make anything. On application Adobe, the company wants to bring a revolution in the world of voice processing.
At the presentation, which was led by comedian, actor and Director Jordan Peele revealed a striking demonstration of the application. The Adobe engineer, TSE-Yu Jin directly into the scene handled the interview with a partner Saw appearances by Keegan-Michael Key. Jin has modified the record, just typing the phrase, and made the Key “tell” that he kissed a Saw instead of his wife.
The developers admit that they understand that such a tool in the hands of fraudsters can be a dangerous weapon. But for those who are involved in podcasting, this tool can provide invaluable assistance. Now they don’t need to hire professionals to voice another program or the news.
Adobe said that although VoCo is not yet ready for release, people find it difficult to distinguish between real voice and arranged a fake, therefore, the company approached the issue with seriousness and think about how to make the difference easily detectable.