4 ms·

Always nice to see people mention radio stuff on HN. Just a quick remark, I believe Hellschreiber is transmitted as a single line, just displayed as multiple i

• genewitch 5y ago
  You're correct, although it's not technically "a line" it's a column of "on" or "off" - the receiver is, per the protocol, supposed to output at least one copy of each character. The wiki mentions that having three possible rows allows for deviations in timing, however, three rows also allows for deviation of frequency. Today's radios, combined with FLOSS (or at least 'source available') software can get timing synchronized within 20ms, including the delay in actually getting the RF out of the 'antenna' - pci latency, soundcard buffer latency, radio soundcard latency (if you're using a radio that connects via USB, see ic-7100 for example). When I use WSJT-X, which is "weak signal" software, i routinely notice less than 1s timing deviation median. WSJT-X relies on the system clock that the software uses to be accurate, as transmissions start at published times - like exactly on the minute boundary, :00. all reception reports include the time delta. As another data point, slow hell mode uses 30hz of bandwidth, and i have used kiwisdr websites to 'hear myself' nearly everywhere in the world using slow hell mode. No one else is using this mode, so i generally have to use kiwisdr to test different modes, antennas, and radios. Another open source project, iirc.