Thanks Greg for the comments. I'm not a specialist, far from it. I've just seen
ML applied in other domains and it did not inspire much confidence, however
maybe this will be indeed different. If there is anyone who can pull this off,
it is certainly Google. They do have access to a very large
On Sun, Feb 28, 2021 at 12:41 PM Adrian Musceac wrote:
> While interesting and newsworthy, I'd assume from the start that this codec
> has the same advantages and pitfalls as other ML applications, i.e. works
> very well in 90% of cases and fails dramatically in 10% of edge cases. Even
>
Greg - LPCNet quantised to 3000 kbit/s would indeed be an interesting
comparison; Jean-Marc and I have been quantising LPCNet at beneath 2000
bit/s, in my case to squeeze it through a HF Radio channel. I suspect
the focus on 3000 bit/s is because that is the lowest rate that makes
sense for
While interesting and newsworthy, I'd assume from the start that this codec has
the same advantages and pitfalls as other ML applications, i.e. works very well
in 90% of cases and fails dramatically in 10% of edge cases. Even though the
page specifies it is aimed at a completely different