Great! I have not implemented any timestamping yet, so I'll do whatever exists when I get around to add support for it.

I would also want to add waveform to this specification, same as voice messages.

https://github.com/nostr-protocol/nips/blob/master/A0.md

It's part of the "imeta" tag though for this, but we could just call it "waveform" as a tag. I'm sure many would appreciate being able to add that.