5 ms·

I've wasted (counting) about 300 seconds of my life listening to these audio files and they all sound and seem fake...

• tagyro 3y ago
  just to be clear, one could mistake them for some (voice) actor reading a book (maybe) but even to my untrained ear they sound fake and artificial. Am i missing something?
  • kvn8888 3y ago
    It's meant to sound artificial. The focus is on speed and consistency
• svantana 3y ago
  I found that in my (high quality) studio monitors, the audio sounded fine and hard to distinguish from 24kHz wav. But in headphones, the artifacts were pretty obvious. So probably some reverberation will do a lot to cover up artifacts. In the paper, they only do a subjective comparison between the generated audio and the soundstream-encoded original audio, which seems a bit disingenuous. Listening to soundstream audio in headphones, I can hear those same artifacts.
• jeffbee 3y ago
  Did you read the paper? They intentionally steered the quality to ensure they sound fake. Their generated speech is "very easy to detect" according to the reference at the end of the paper.