Monday, February 15, 2010

Are there too many digital modes?

I have been thinking quite a lot recently about the various digital modes from the point of view of which ones are worth supporting in a program. My own software KComm developed for the Elecraft K2 and K3 supports only PSK natively, mainly due to the existence of AE4JY's PSK Core DLL which makes it easy to add PSK31/63/125 support to any Windows program. I can always use Fldigi or another program if I want to use other modes. But you get used to using a particular program and there is a certain challenge in adding support for other modes, which is why I have been spending time wondering whether it would be worthwhile to do so.

The more I think about it the more I tend to the conclusion that PSK31 is the only digital mode that is really worth bothering with. Its usage greatly exceeds everything else including, I suspect, RTTY. It has achieved 'critical mass' which means there are always new stations to work including exotic DX locations. This single point outweighs any of its disadvantages, such as the fact that the phase shift information gets messed up by disturbed paths, it's harder to set up to give a clean signal or that some MFSK modes such as Olivia that have redundancy and error correction are more reliable with weak signals. PSK31 is a narrow band mode and the benefit of that is often overlooked. I think things would be getting pretty ugly in the digital mode band segments right now if all the people who currently use PSK31 were instead trying to use much wider modes like RTTY, MFSK or Olivia.

RTTY. RTTY was the first digital mode I ever used, way back in the days before sound cards, so I have nothing personally against it. I suspect it is the second most popular digital mode after PSK31. I don't think RTTY use has died off, so much as it simply hasn't enjoyed the growth in popularity that PSK31 has. RTTY would be my main candidate for the next digital mode to support in my program. Except that in the last decade or so I have hardly ever felt any inclination to use it.

One reason is that RTTY is a grossly inefficient mode. It occupies a lot of bandwidth for its data rate, due to the fact that it was designed long before computers and DSP were invented and had to be decoded using an analog frequency discriminator. It has a limited character set geared to the Teletype devices it was meant to be used with, and requires a shift character to go between letters and numerals which adds to the likelihood of receiving gibberish if this character is not received whenever it is sent. Consequently RTTY users adopt the sledgehammer approach and run very high power to maximize their chances of good copy. RTTY really is an outmoded mode that has absolutely no advantages or benefits and instead of encouraging its continued use perhaps we should be persuading people to change to more modern alternatives.

MFSK. MFSK was developed by ZL1BPU and IZ8BLY in the early 2000's and I was an enthusiastic user and advocate of it for quite some time. MFSK16 seemed to give more reliable copy than PSK31 at the same power level. It also had a nice feature for exchanging small bitmap images, similar to SSTV. However in recent years it seems to have fallen out of favour and these days you rarely hear it.

Other multi-FSK modes have also come and (in most cases) gone, such as DominoEX, MT63 and Olivia. The benefits of all these different variations on MFSK escapes me, though I'm sure at least some of them do have benefits, however they have never caught on and would probably die out altogether if they did not live on as options in the menus of programs like HRD, Fldigi and MixW providing temptation for every new convert to digimodes to try them.

All these modes seem to achieve is confusion and a proliferation of "What's this mode?" threads on QRZ.com. There just aren't enough users to make many contacts with them, and anyone calling CQ using one of these modes is heading for disappointment unless they happen to be spotted by a PSK31 user who manages to guess what mode it is and takes pity on the poor caller. These modes should really be consigned to the bit bucket of history.

JT65A. A mode currently enjoying a surge of popularity on HF is JT65A, thanks in part to the easy to use JT65-HF software written by W6CQZ. I tried this myself recently and made a couple of contacts before starting to wonder what was the point of it? As Wikipedia explains it, JT65A was developed to allow contacts to be made over slowly varying paths where signals are expected to be weak, such as EME or long distance troposcatter contacts on VHF. Such contacts are likely to be prearranged skeds over paths where no other mode except high power CW is likely to make it.

However the use of this mode to make random contacts on HF seems pointless. People frequently seem to be making contacts when signals are strong enough that the contact could be completed more quickly, with the exchange of a lot more information, using PSK31 or one of the other digimodes. Signals on HF tend to vary quite a bit, so that even if they are very weak some of the time there are periods when they are strong enough to support "regular" modes. Do the problems that JT65A was designed to overcome actually occur on HF? I don't think so.

Conclusion. The conclusion I came to after the ponderings I have tried to summarize above is that PSK31 is really the only digital mode worth supporting if the aim is to make the maximum number of contacts. RTTY would be worth supporting for legacy reasons because it is still quite popular, although nowadays it seems to be little used outside of contests. From a technical perspective RTTY is arguably a mode that ought to be phased out.

The other modes I listed - and some that I haven't - are so esoteric that that they are not worth bothering with because there isn't a critical mass of users to ensure a good number of contacts. These modes may be interesting to try out, but as such they ought to remain within the realm of their own specialized programs - much like the original IZ8BLY Stream MFSK software of a decade ago - so that users are forced to make a conscious decision to use these modes and hopefully find out a bit about the purpose behind them, instead of just picking them from a menu because they are there.

No doubt there will be many who read this far (if anyone reads this far!) who will disagree with my conclusions. But from now on I expect to stick exclusively to the PSK modes on the HF bands and stop worrying about the other twiddly noises I occasionally hear on the airwaves.

12 comments:

  1. I think you're right.There seem to be so many modes. Personally I only use PSK31 but in the past I have used RTTY which seems to still be very popular. I have read somewhere that it is possible to use WSPR in some sort of QSO mode but I haven't looked too deeply into that.
    Certainly most of my HF operation at the moment is on PSK31, and WSPR when I am doing other things in the shack or around the house.
    73 De Kevin

    ReplyDelete
  2. Greetings Julian!

    Good question. I'm not convinced there are yet too many digital modes, but we're getting close.

    Like you, I see PSK31 as being the top dog of the bunch at the moment, but we've moved into a technology slipstream that's very much like the Internet. Things change with such rapidity that what is "hot" one day is "old" news just a few days later. These many modes are just like that. Where we used to get a new mode to play with every few decades, we now get them every few months; or so it seems.

    Ultimately I believe this will be the undoing of CW. Morse has survived into the 21st century because it has been the most cost effective and efficient mode we've had available. PSK31 and a few other soundcard modes have effectively ended that hegemony and now we're beginning to see more Digi ops than CW ops.

    The resulting struggle for bandwidth will determine who "wins".

    I'm betting on the Digi modes but I haven't sold any of my keys... :-)

    73, Jeff KE9V

    ReplyDelete
  3. I personally don't see digimodes as being any threat to Morse. CW is the digimode that can be sent and received with or without a computer (and currently it can be received better without, although I'm sure a talented programmer could close the gap if he had a mind to.) It can also be sent and received with the very simplest of equipment. People are learning morse just so that they can make contacts with little two transistor homebrew transceivers. So I think the interest in QRP alone will ensure the survival of morse.

    The digital modes that offer better weak signal performance than CW are either slower or occupy a wider bandwidth, or both, and need a higher duty cycle. I'm not ruling out that someone will invent a digital mode that is better than morse in every respect and if by then computer based SDR is almost universal CW could face a tough challenge. But it hasn't happened yet!

    ReplyDelete
  4. Hello Julian, it's a very lucid statement. I think your right. With Fldigi I could try a few digi modes, but because the most of us uses PSK31 and the decoding is pretty good with it, so this mode will survive for a long time. I try JT65A for a few months now, but I work stations I could work also with CW and phone. And JT65 is very complicated, too short, too slow to make a contact. For QRP is CW, PSK31, WSPR the best mode to work with. I would be wonderful if more amateurs would try the qso mode with WSPR (WSJT7). I think it will outperform JT65A. I lost my interest for WSPR as a beaconing station. It would be really nice making qso's with WSPR. 73 Paul

    ReplyDelete
  5. It could all seem like a big conspiracy at times, to divide and conquer. Let us tune through the bands and see, Oh yes one mode per operator and everyone else turned off because they could not hear anyone on their mode. It is not just too many digital modes but too many modes/bands and types of working. The diversity of the hobby keeps us coming back but it also allows us to loose track of friends and become bored when everyone else subscribes to the latest fad except us. We cannot stop progress but I am sure we should bury some of the legacy modes or at least only bring them out on special occasions once a year.

    ReplyDelete
  6. WSPR has a QSO mode? Where'd I miss that? Does anyone have a link to such a mode?

    ReplyDelete
  7. I agree with Julian's earlier reply and the cw point of view.

    I think cw operators employ that mode because they genuinely enjoy sending and receiving Morse. And will continue to do so for as long as possible.

    Of course we QRPers appreciate cw because it really is an effective mode of communicating when running very low power.

    I've recently read of amateurs who have reduced their stations to one or two QRP rigs. One fellow in the UK now has an MFJ Cub as his station transceiver. And, GM3OXX is known for his mighty one-watt on 20 meters. I have his QSL card.

    QRP cw lives.

    72 Dick N2UGB

    ReplyDelete
  8. Perhaps you fellows haven't been keeping up with the US QRPers plans to QSY from 7040/7030 to 7100+ due to the continued expansion of the digital ops on 40m?

    As I said originally, as the number of digital mode ops outpaces the number of CW ops, the CW boys are going to have a much harder time finding a clear spot on the dial.

    It isn't a matter of them not wanting to make dits and dahs; the spirit is willing but the bandwidth is lacking...

    73, Jeff KE9V

    ReplyDelete
  9. I have been following that discussion and I can't see the logic in it. Digital is just as popular in Europe and we have managed for decades with QRP CW at 7.030 and digital starting at 7.035. If QRPers are getting digital QRM on 7.040 that's probably because some US ops are trying to contact digital ops over here. Moving the QRP frequency up to an area used for SSB over here is not going to help either the QRM problem or the ability to make QRP - QRP contacts on 40m.

    I think what's going on in the US QRP mailing list is a storm in a teacup being stirred up by people determined to make "mode war". If there are problems, they are caused by the US having a different 40m band plan from the rest of the world.

    ReplyDelete
  10. Julian some of that is true to be certain. But given the frequent RTTY contests, and the warm-ups for those contests (as if that makes any sense whatsoever) that drag noisy sigs all the way down to 7035, and add to that the growing number of casual digital mode operators who are now camping around 7030-7040, and QRP CW is having a tough time finding a clear spot in the US.

    Poor band planning, bad operating practices, the short-sightedness of building radios without VFOs, whatever ... the result is that US QRPers are having a bit of a tough time finding a spot on the dial to call their own.

    And I don't believe this is a short-term problem, but let's see where we are in another year...

    73, Jeff KE9V

    ReplyDelete
  11. I am not stuck to the QRP frequencies. I work QRP wherever I can. I work also between the QRO guys, and I do not always mention that I am working QRP. I even look in the DX section for contacts. And I made a lot of DX, only exchanging 5NN TU, but never mind. If they hear me I will work them. But I am not defined to QRP frequencies. 73 Paul

    ReplyDelete
  12. Jeff - interesting that you cite RTTY as the main cause of the complaints. I don't think RTTY is a mode whose use is growing, it is one that mostly only gets wheeled out during contests and as I said it is an inefficient and outmoded mode. It's about time its use was deprecated, as they say in the software biz. If it didn't waste so much spectrum it wouldn't need to stray into the CW sub-bands.

    The desire to see all the activity in one waterfall (and to be seen on other people's waterfalls) works to constrain the bandwidth used by PSK31. Of course there are other digimodes like ALE and PACTOR that are a pain in the a** for everybody, but they are only there because of a desire to have reliable communications channels for "emcomms" so I guess that's OK. :)

    ReplyDelete