Well, without measuring anything on a path simulator, P III works very well.
Uses most of the good tricks on P II, like convolutional encoding (sort of FEC) Viterbi (maximum likelyhood....i.e. minimum repeats) decoding, and some clever speed change strategies. P III improves in modulation, keeping PSK at the maximum distance between constellation points, which is about the same with DBPSK or DQPSK. It is better to keep the most robust constellation (P II did not) and add more carriers as propagation allows. Only very bad propagation stops P III. And it is FAST, works when WinDRM does not. Seemingly, P III is strong enough at 100 bauds with the rest of the tricks it uses, and seemingly does not NEED to hold the horses down to 50 baud. I don't know for sure, either. To me, it seems that testing something on a simulator with just gaussian white noise does not tell the whole story. Doppler and spread delay should be mixed in the propagation cocktail. Why? So far, I have been unable to get MT63 to work on 40 meters, NVIS. Not so on long distance 20 meters QSO's, where MT63 works very well. I have been doing some measurements on doppler with Spectran on 6 and 7 MHz broadcast stations, and I have seen up to 5 Hz doppler one evening, with very a complex "weaved" QSB patterns. Possibly, that may kill MT63 at 10 baud.... So far, the most robust HF mode to me is Olivia. Only once it has failed against PSK31 in a QSO with Argentina on 40, when the phasemeter seemed to go nuts on MixW, and garbled PSK31 a lot, Olivia did not work. I have had good results in tests with PAX / PAX-2, RTTYM, Contestia, Olivia, and a voice mode Patrick included in MultiPSK, used as keyboard to keyboard mode. To me, so far, DominoEX at 4 baud with FEC is not as robust as Olivia. And Chip64/128 does NOT work on 40 meters. People say that MultiPSK's 100 baud packet mode works very well....I have not been able to witness it so far. So, here are my 2 cents....any other comments to amount up to a dime? Jose, CO2JA DuBose Walt Civ AETC CONS/LGCA wrote: > Please allow my 2 cents worth. > > KN6KB in his presentation of SCAMP to the DCC a couple of years had a > slide that showed where he measured P III with a channel simulator > from KC7WW. I showed that around -5 dB SNR there was still something > in the area of 200 WPM throughput. > > I believe that you might be able to see some incidental throughput > perhaps at less than 50 WPM. > > The thing that will absolutely kill Pactor III is its 100 baud > physical signaling rate. There are conditions on HF when this high of > a baud rate simply will NOT propagate. > > IMHO, if they were to drop the baud rate to 50, they might even see a > higher throughput at a -5 dB SNR. > > Its interesting to note that KC7WW measured MT63 on his channel > simulator and MT63 showed 200 WPM at or near -5 dB SNR. > > I discussed this at leangth with Vic Poor when we had breakfast at > the DCC and he agreed that the above was probably true for Pactor III > and MT63-2K between -2 and -7 dB. > > If anyone would like to verify KN6KB's measurements, I recommend you > purchase a copy of the DCC proceedings from TAPR that contain KN6KB's > SCAMP presentation...or purchase a KC7WW channel simulator and check > it yourself. Note, you might also try duplicating the test using Moe > Wheatly's PathSimulator. It runs on a PC and is free to download. > > 73, > > Walt/K5YFW > > -----Original Message----- From: digitalradio@yahoogroups.com > <mailto:digitalradio%40yahoogroups.com> > [mailto:digitalradio@yahoogroups.com > <mailto:digitalradio%40yahoogroups.com>]On Behalf Of KV9U Sent: > Wednesday, January 10, 2007 3:12 PM To: digitalradio@yahoogroups.com > <mailto:digitalradio%40yahoogroups.com> Subject: Re: [digitalradio] > Pactor versus Olivia > > Rich and group, > > I have read from the SCS website that there is still throughput with > P2 and especially P3 when down to maybe as low as -16 db below noise. > Others have claimed that Pactor 2 and 3 drastically drop off by the > time you reach -5 db S/N such as KN6KB's RFfootprints powerpoint on > comparing various digital modes. > > Can you quantify what some throughputs might be at the low S/N > ratios? There isn't any miracle modulation scheme with pactor 2 and 3 > from what I can see. What they do is optimize many little things > which gives them improved throughput. > > I question whether P3 is going to work a lot better than the wider > Olivia under the most difficult conditions, although I hear the claim > made that it does extremely well. But I am not sure of the actual > throughput under real world conditions. The curious thing is that if > it can not operate below 100 baud, there should be times of doppler, > bit smearing, multipath, etc., that would make pactor modes > completely unusable even though low baud rate modes would work quite > well. > > 73, > > Rick, KV9U > > Rich Mulvey wrote: > > > What BW/tones were you using for the Olivia comparison? > > > > The thing about Pactor 1 is that it's adequate for relatively good > > conditions, but is noticably worse than Pactor 2 and 3 when things > > get marginal. Pactor 3 absolutely shines under the absolute worst > > conditions, when all of the sound card mode operators have given up > > and turned off their rigs, complaining that the band is dead. ;-) > > > > - Rich
