Goodbye FT8, Hello Olivia, The MAGIC Digital Mode For HF!

Discussion in 'Amateur Radio News' started by KJ4YZI, Oct 23, 2017.

ad: L-HROutlet
ad: l-rl
ad: L-MFJ
ad: Subscribe
ad: Left-2
ad: abrind-2
ad: Left-3
ad: QSOToday-1
  1. KK5JY

    KK5JY Ham Member QRZ Page

    Then refute it, technically, with facts. I read K1JT's early papers on the subject, and he specifically calls out that exact design decision in a footnote. You may treat things you don't want to believe as "fake news," but the truth is that the JT modes were designed to only recover half the encoded data from the radio channel, just as I described. The "clock" half of the data must be conveyed by other means. It's there in black and white for people who read the facts before the cry "fake news." :rolleyes:

    If you have any doubt as to what I am saying, disconnect your computer from the internet, set the clock to a random time, and try to make JT or FT QSOs. It won't work because the mode has no clock recovery. Other modes can handle that situation just fine, because they do full clock recovery on the received analog data from the radio.
     
  2. W0PV

    W0PV Ham Member QRZ Page

    Too much "randomness" doesn't work in either case.

    Other classic modes won't work as well if a sked to meet DX at 2000Z for a 60 second QSO window and the offset difference relative between each stations clock is > 1 minute. Two ships passing in the night ...

    Or said another way, if my PC clock is deliberately, or, coincidentally, set to equal another PC clock, both the same offset from any universal standard, FT / JT modes work fine for QSO's between those PC's.

    I am not disputing "clock recovery" aspects between the modes. Only that the method that is incorporated in FT / JT makes communicating that clock data OTA irrelevant, not necessary for useful information xfr.
     
    Last edited: Oct 24, 2017
    KD6SM likes this.
  3. KK5JY

    KK5JY Ham Member QRZ Page

    People keep trying to compare clock recovery with a QSO schedule. There is no comparison. Clock recovery for a scheduled QSO has to operate just like it does for an impromptu QSO, regardless of mode.
    Sure it will, because you synchronized the data stream clocks, a priori, using a communication means other than the HF channel. If you used another mode on that same HF channel to synchronize the clocks, you are getting slightly closer to what other modes do internally, but then doing so requires a mode other than JT modes, because they have to have clocks sync'ed before transmission. So you could use voice, another data mode, WWV, whatever. But doing so means that the clock portion of the JT QSO was communicated using that mode, and not using JT itself, and the result is the same -- those modes do no clock recovery, and thus their claimed SNR values cannot be compared to modes that do their own clock recovery.
    Respectfully, this is completely incorrect. JT clock information and synchronization are critical, and the software manual will tell you so. All data modems have to receive both the data states and the clock used to encode them. JT only communicates the data, but the clock part of the transmission is ignored by the receiver, in favor of using the system clock, which is synchronized with the transmitter clock using other means than the JT data channel. It says so in the manual, and it says so in the whitepapers that K1JT did on his work. For other modes, the minimum SNR possible is determined both by the fidelity of the data transmission, and the ability of the recovery algorithm to extract the clock from that same transmission. JT modes use other communication channels to align the data bits with the clock used to encode them, and so the JT modes seem to be able to decode at a lower SNR that other modes, when the truth is that a big chunk of the information was transferred by means other than the HF channel, and so the SNR values are not comparable with those of other modes.

    On a somewhat related note: people think that this critique of the JT modes means that I'm calling the JT modes "bad," or "useless," or "invalid," or whatever. That is simply not the case. I have used the recent popularity of JT65 and FT8 to do extensive testing of antenna designs and propagation. If I thought the modes were useless, I would not have used them for those purposes, and I certainly would not have written Internet articles about doing so. One place that JT65 SNR numbers can be compared is with themselves. You can compare the SNR of one JT QSO to the SNR of another JT QSO, because the "clock handicap" is the same, assuming that the clocks are well-synchronized with NTP, etc. You cannot compare JT65 to FT8, but you can compare FT8 to itself across QSOs, for the same reason.

    Such comparisons and operations have value, but the value is limited to like-mode comparisons, because the demodulator design chosen is not complete in the same way that all other data modes are complete. So comparison between JT/FT and Olivia is simply invalid, and until the JT demodulators add clock recovery, that's how things will continue to be.
     
    K2NCC likes this.
  4. W4NNF

    W4NNF XML Subscriber QRZ Page

    The problem is...nobody uses Olivia. What are people using? FT8 to make contacts. PSK31 to chew the rag. Sorry.
     
    KN6Q likes this.
  5. OE3NHW

    OE3NHW Ham Member QRZ Page

    Hi, I looked all that interesting Video on Olivia, never tried that mode so far...I just encountered FT8. I saw the call of my friend (OA4TT) in the cluster for 6m, he worked Europe, that was in july...Ok, it took some time for me to install it easily after returning from my home country Austria (OE3NHW) to Peru (OA6Q), where I'm staying at least for 6 months a year for being a resident. You can imagine a lot hams worldwide are interesting in a contact with Peru. BUT with these poor conditions it got rather difficult..In 2 months since end of july I had only 900 qsos (inc. 533 in the cq ww rtty contest) ...that means zero for many days, in better years it has been >1000/month and now the world changed a lot, since oct. 1st I had already 1360 qsos in only little bit more than 3 weeks. FT8 is not that much different than a pileup in phone or other modes. And FT8 offers big advantages, you get all calls who are looking for you into the waterfall and of course decoding is accurate, error free, decodes even 2 calls on the same frequency. What bothers me is that a lot operators are not familiar enough or not fast enough to send the proper message, some are familiar working split as they could see your call on any frequency and waist a lot of my time and let others wait. And remember, there are many waiting for a qso with Peru, FT8 change the conditions, its like getting 100 Bonus-Sunspots... 5 bands are open often over many hours during daytime, never before here rather near the equator (16 degrees South). And some directions at the same time like never before. I have a lot experience as I'm running the station from Peru since 10 years. But ANYWAY I WILL TRY OLIVIA to do also something else.
    Thanks for the great video on Olivia and Good Luck
    73 Hans
     
    KK5R likes this.
  6. KK5JY

    KK5JY Ham Member QRZ Page

    I see Olivia QSOs all the time. It's not the spaghetti-ball that you find on the frequencies used by FT8, but so what? FT8 is new and its "fad" value is still high. It will eventually equalize with the other data modes, just like PSK did and just like JT65 did.
     
  7. W4NNF

    W4NNF XML Subscriber QRZ Page

    You clearly explained why you THINK people would prefer Olivia. Are you surprised thatFT8 users disagree with you? What did you expect?
     
    WU8Y and WI3U like this.
  8. K3GAU

    K3GAU Ham Member QRZ Page

    I agree with W0PV. It really doesn't make any difference what time the two clocks involved say as long as the second hands agree as to when the time period begins.

    The reason JT, etc have to use an external clock reference of some sort is that they were designed to receive signals that have traveled over very long distances with varying path lengths and are very weak. Neither condition is conducive to being able to recover an accurate clock signal. Consider that the EME path delay is something over two seconds and varies somewhat.

    Years ago some folks devised something they called CCW (I think). Regular Morse cw letters were sent but timed by master clocks to begin at designated times and with prescribed element lengths. They could copy cw signals than were much weaker than they were normally able to copy without sync. It sounded just like regular cw at about 10 wpm.

    Dave K3GAU
     
  9. W4NNF

    W4NNF XML Subscriber QRZ Page

    There's ample PSK31 activity on 20 depending on conditions, but you're correct, there's more FT8...people are just flocking to it. We'll see how long that lasts, but for now I know I am having a ball.
     
  10. W4NNF

    W4NNF XML Subscriber QRZ Page

    PSK31 highly automated? Depends on the ops. Sure, you'll see some QSOs where all the stations do is exchange macros, but there are also many, many traditional QSOs going on, everything from a reasonably brief exchange of pleasantries to full blown rag chews.
     

Share This Page