iain.philipps wrote: > --- In [email protected] > <mailto:dstar_digital%40yahoogroups.com>, Nate Duehr <n...@...> wrote: > > > > The short answer is: Bit errors affect voice just like they > > affect data. There's no such thing as a "voice only" or a > > "data only" stream. > > Maybe I got the wrong end of the stick, but the way I read the original > posting, it looks (to me) like:- > > "We only send data, and not voice. So where is this garbled voice stuff > coming from?" > > I believe that's the reason for the "data only" assertion ...
I fully understand that, thus the statement: There's no such thing as "data only" in D-STAR DV format. You're always sending both audio and low-speed data portions of the stream, even if the rig's mic isn't keyed. Whether or not the rig bit-stuffs the audio with complete digital "all-zeros" silence, or just doesn't feed the audio into the CODEC, resulting in a similar but maybe not dead-silent result after noise, someone who's looked at the decoded streams would have to jump in and say... It's important to point out that someone feeding data into the rig and letting the rig auto-transmit on a timer, isn't just sending data. You're sending a full D-STAR stream, complete with routing headers (callsigns), an audio stream, etc. If any of it gets mangled in transit, even the "audio" you think isn't there, it'll make funny noises. The protocol is not a "smart" protocol that can use the bits for voice or data in a dynamic way by setting a bit-flag to tell the receiving radio that "this is an all-data transmission". Would have been nice, but it isn't... It's a continuously interlaced format. Nate WY0X
