-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 As far as i'm aware there are few voice & text language physical-machine based translator, that can convert language, and then speak it out with another speakers voice in another language.
That is what VoIP software over Tor needs to use and implement. For live person to person discussion or talk: Delay is of less importance (that is 2nd or later priority). Anonymity is of more importance (that is, first priority). User voice has biological marks/fingerprints. A user's own voice should not be broadcasted/sent to other side. Both/all side MUST do similar to this (when using Tor): VoIP software -> voice -> convert voice2text -> text -> sent to other side, over encrypted tunnel -> text2voice(use_profile_male_John) -> voice. Such VoIP client need to have practice+correction+train MODE. So users can SPEAK various things and REPLAY it, by choosing one out of different common audio-profiles, like John, Jane, etc, to hear and test what other side will hear. Then use it for conversation. And it will also be nice to have speech variation even for single profile, like John. Then more than one person can use such for GROUP DISCUSSION purpose, instead of (two-party) TALK/DISCUSSION. This idea came from practical use case description from another user: that user has sent his/her TYPED-IN text in a Text2Speech free software, that software SPEAKED/PLAYED those typed text out load, and software was configured to use specific output audio channel, that audio channel was used as input for VoIP software to travel over Tor, for that communication, (initially VPN was used, but later) SSH ethernet tunnel was used and worked. - -- Bright Star. Oct 23, 2013. Received from Mirimir, on 2013-10-22 11:08 PM: > On 10/22/2013 11:42 PM, Matt wrote: > >> On Tue, 22 Oct 2013 21:15:46 -0600 >> Mirimir <[email protected]> wrote: >> >>> Was there much break-up? >>> >> >> The sound quality was surprisingly clear in my tests, and I didn't have >> any dropped calls, even over extended periods of time. >> >>> I never got around to directly comparing UDP mode via VPN via Tor, >>> versus TCP mode via just Tor. My initial experience convinced me that >>> VoIP and anonymity are entirely incompatible :( >> >> I don't think this true at all. If you want to use torsocks and mumble >> together, you may want to try it with dgoulet's new torsocks rewrite. > > I'm referring to voice recognition. Once a resourceful adversary has a > sample of your voice from Mumble, they can match it with intercepts from > POTS, cellphone systems, Skype, Zoom, other VoIP, etc. And it's very > hard to disguise voice effectively. > -----BEGIN PGP SIGNATURE----- iQIcBAEBCgAGBQJSaK7jAAoJEID2ikYfWSP6HO4P/2gIzmY26LR0hX2huEYWCGrc o1jhJcXt5SSomi2m2KKbv/vls+L+GKTFzPaRZtZhOQ+DeShiDOQzuXj1Dvmbrzep 8UDRYL/TmyTGP2ZGcEap9GpKdJCSYiaKxIxcRh6O0UcKpUK2xc5udF675bT68mOs 7zRS2oHGrlDMlwAltCB9J0qohLVfCgHtDTojvIx085cvXM/FFrC4sFON+CISL8oc 94fx4tOdSVzHx9ZT9FY4MXYvXUy7mPmHf+vmKeaa9dc8q+ulpVf+66BmCxYZZnT3 Nzu3hF8ymxTwwDsmIgF8G15lk1E504r0fXm9XmZkYN8nzYBT/6iaKeMz/lMktXSy ZJoD/6hZI8EfXumScin7g1M/c2bMoopj8DD4AyBhvyd/D6m1EypLkKCKkxdI0KnV iUEWxPyXiqNQKoeEPtd8LrQ85SX7htMT6KY0gqV7S563v2cJp6luiYRg03j3kq7v OClZyODgrCcTPnbXHiie1SMmMTlBVaaYXncyNHde6PoeRGJpw8MT8aAZg6xLhxR3 mheszV8S9vXqvZ3kdQ9JpBZUfM5sucC+6kcFd47KsD20l85v4SNMnN3KnHiyaKd3 O0tUK+oP44Bq7wh0KdlkoXORWbDY+sJiwC1/ayzwdIDxAM/QQuPOWawH++dVZfDQ FvmJyGVnpWxgO5oUBLST =ZQZa -----END PGP SIGNATURE----- -- tor-talk mailing list - [email protected] To unsubscribe or change other settings go to https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-talk
