On Sun, Feb 24, 2019 at 9:24 PM Jeroen Vreeken <[email protected]> wrote: > > Hi David, > > 52bits per 30ms might work out nicely. > > 30ms at 2400baud is 72 bits. If we use a 16 bit uw there are still 4 > bits to spare. > I don't think FEC is that important when using an existing FM system. > For more advanced work with TDMA that will help, but I think that would > require its own format and mode anyway, no need to mix those two up. > (Execpt use the same codec2 mode for convenience) > > I think the code changes for such a mode would be relativly simple. > fmfsk is currently fixed at 96 bits per frame, but that can be fixed > relativly easy and a new frame format should be easy to add. (Basicly > its size is between the 800XA and 2400A/B)
Given how good JM's 1600 bps codec sounds, I would be surprised if adding more bits to the codec instead of FEC were a good trade-off. Keep in mind FEC can also be used to help recover timing, train equalization, and monitor signal error levels. For digital voice for HAM FM I think the major accomplishment would be sounding better than NFM under common good conditions, while achieving usable copy in the presence of worse multipath and interference. Under that line of thinking driving bitrate out of the codec is useful specifically because it allows adding more FEC. Are there really many places where spectral efficiency is actually important for VHF+ ham communications? It was my impression that ham DV systems were pushing on spectral efficiency only because of being receiving commercial system hand-me-downs. (and even on the commercial trunked radio systems, I wonder how much of the pressure for increased spectral efficiency comes from an actual need for additional capacity vs what differentiators someone could stuff on an RFP-- certainly when I worked on a muni digital trunked radio system RFP process twenty years ago we liked the additional capacity but didn't actually have a really clear need for it, and today the trunked radio system in the municipality I live in is pretty lightly loaded) If anything, I think switching to modes that transmitted two relatively narrow signals 100KHz apart coding the same data (perhaps randomly permuted for better temporal gains) in order to get more multipath diversity would make more sense than striving for the narrowest signal. _______________________________________________ Freetel-codec2 mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/freetel-codec2
