4 ms·
That is a possibility. But the issue with that is complexity.
by iasay 4y ago
That is a possibility. But the issue with that is complexity.
- londons_explore 4y agoNote that modern compression[1] now allows decent fm-quality music or speech in just 375 bytes per second! That's fixed bitrate too! That means that compared to AM radio, you can have a 200x power reduction - or more likely, use the same power but far fewer transmitting masts, saving millions of $$$'s annually. > complexity Yes, you're right - but it's all software, so when you've done the one-time effort of writing said software, the complexity is 'resolved' - it's not like a complex circuit where every single device must have more components, more labour to assemble, more expense, etc. [1]: https://ai.googleblog.com/2021/08/soundstream-end-to-end-neural-audio.html https://ai.googleblog.com/2021/08/soundstream-end-to-end-neu...