Antares AI Voice Transformer

[Ploki]

https://www.kvraudio.com/news/antares-r ... eals-65280

so, yeah
"ethically-trained voice models"

[zerocrossing]

Ignoring ethics, I have issues.

First, there always seems to be some sort of noise. I assume this is an artifact of the render? If you doubled renter times, could you fix it like 3d renders anti-alias images?

Second, why are there no vintage pop/rock models? I want to hear a John Lennon and a Janis Joplin. Roger Daltrey. Freddy Mercury. Even a Billie Eilish or Regina Spektor. I hate this modern over-emotive "American Idol" b.s.

[Ploki]

Ignoring ethics, I have issues.

First, there always seems to be some sort of noise. I assume this is an artifact of the render? If you doubled renter times, could you fix it like 3d renders anti-alias images?

Second, why are there no vintage pop/rock models? I want to hear a John Lennon and a Janis Joplin. Roger Daltrey. Freddy Mercury. Even a Billie Eilish or Regina Spektor. I hate this modern over-emotive "American Idol" b.s.

1. i doubt more passes means better artefact control. the noise seems to be inherent to current AI "rendering" methods, nearly every genAI has this noisy/phasy undertone.

2. honestly the only thing that this thing does better than competition is being offline

[Cuauhtli]

Those demo voices brings to mind the George Carlin joke about suburban white boy junk played by bands like Men Without Testicles.