On Tue, Aug 9, 2016 at 11:06 PM, Daniel Hoffman via bitcoin-dev <bitcoin-dev@lists.linuxfoundation.org> wrote:
I have updated the GitHub a lot (changed tones to be less chirpy, fixed some smalls) and made a couple of samples (see attachment for MP3 and FLAC of both tone tables, first 16 then 4). Is this good enough to warrant an official BIP number? I haven't built a decoder yet, but it seems like the encoder is working properly (looked at Audacity, seems like it is working), and some people on reddit want to "allow for decoding experiments"

What suggestions do you all have for it?


With DSP hat on, your decoder for noisy/distorted channels will be 99.9% of the complexity and will completely control the design of the encoder.

It's not a proposal yet without a decoder, it's just an idea.  FSK modems microphone-channel (terrible multipath) is quite challenging and several other parties have tried to do bitcoin info over audio in the past without success.

It's very interesting, but I think you do need to go through and get the whole thing working to really gauge viability.