Can I locally compare echoprint audio file fingerprints?

It's possible, but we don't support it and would require some rethinking of the architecture. I'd take a look at the local_fp testing stubs in fp.py for a start. Maybe ask on the echoprint google group if anyone has done this.