why always compare to MP3, a 30+ years old format? There a plenty alternatives and most communication platforms from zoom, jitsi, discord, … are not using MP3. Even podcast are moving away from it.
It’s a good baseline, since it’s well known. I agree that a more illustrative test would be a comparison with modern codecs.
They are just making propaganda to try to convince people to invest in their terrible company.
So something similar to Google’s Lyra codec? Main difference seems to be that it works well with music and is not optimized for speech only. Is it open-source like Lyra?
Edit: https://github.com/facebookresearch/encodec but not an open-source license.
I wonder if this will especially benefit poor and rural areas, especially in the Global South, that may not have great coverage.
You need more compute power at the client then, which means cheap devices will not be able to do it for a while at least. It’s neat tech for sure, but for now it mainly benefits bandwidth heavy centralized services that can save costs that way.
