This seems like a lot of unnecessary steps just to have simulated CW
with the backend being some new digital scheme which uses the
robustness of (the components of) FT8.
To what end is unclear.
JS8Call has actually taken the WSJT modes and turned it into a viable
messaging system using ASCII text. I guess the next logical step is to
have the characters read out on CW... which... to me seems kind of
pointless.
I do CW because I like simplicity. I can build a transmitter from less
than 10 components and it can get on the air. Or I can use a vintage
rig or even a modern rig and operate a very basic radio signal. That
is the beauty of it, it strips away layers of unnecessary complexity.
BTW I am sure if your conceived new digital weak signal mode CW system
becomes reality, you'll have "real radio men" saying that "it's not
real CW" and "it's not real radio" and the reason cited will be "it's
computers communicating and not radio." I can very much guarantee
that.
Ria
N2RJ
On Fri, Feb 18, 2022 at 4:10 PM David Gilbert <ab7echo@gmail.com> wrote:
>
>
> When I say it is being "held hostage", I don't mean legally or
> physically. I mean that just about everyone ... including you, it seems
> ... thinks that the protocol used in WSJT-X is the only or best way to
> do things. Techniques such as LPDC, FEC, and Costas Arrays could be
> used to send encoded CW messages (i.e., macros from a logging program)
> in digital format that use a different combination of rate, message
> length, bandwidth, and number of tones more appropriate for a typical CW
> exchange, decode those short message bursts on the receiving end, and
> convert them to audible CW for interpretation by the operator. The only
> differences apparent to the user would be that the S/N is maybe 8 dB
> better and the audible CW (being software generated) is essentially
> noise free. I've dug into this enough to know it is possible, but I'm
> not proficient enough in either the techniques or software coding to be
> able to do it myself. My oldest son is an expert in this stuff (his
> career) and has helped me understand the principles, but he has zero
> interest in ham radio and doesn't want to bother with it.
>
> Yes, the source code for QSJT-X is open, but it's a large collection of
> different routines using different languages ... including legacy
> Fortan. I think it would probably be better and more efficient to take
> the same techniques and write them from scratch in a common, more modern
> language. The science is readily available.
>
> I will say again ... FT8 is merely a rigid implementation of some clever
> but standard techniques for signal-to-noise enhancement. FT8 is NOT ...
> repeat, NOT ... the definition of how to do it in general. There
> wouldn't even be a need to make DSP-enhanced CW channelized ... it could
> require tuning around just like RTTY does and it could function in a
> bandwidth no wider than 200 Hz ... roughly similar to normal CW ... with
> partial capability to function with overlapping signals just like FT8
> does now.
>
> I've corresponded with Joe on this a few times. He is simply fully
> committed to the various WSJT-X implementations and I understand his
> position, but he will readily admit that the same techniques could be
> used differently to serve other ham radio forms of operation. My point
> is that the FORM of FT8 has become so ingrained in the minds of almost
> every ham that nobody with the capability is willing to extend the
> precepts of FT8 to other forms of amateur radio communication. The
> success of FT8 has effectively constrained the use of what could and
> should be advances in the performance of at least CW. The constraint is
> perception ... nothing else.
>
> 73,
> Dave AB7E
>
>
>
> On 2/18/2022 11:26 AM, rjairam@gmail.com wrote:
> > The code for WSJT-X is open source under GPLv3 and absolutely not
> > proprietary. Joe has written in great detail about the protocol and
> > the software in QEX and that article is freely available on the WSJT
> > website.
> >
> > Nothing is being held hostage. If someone wants to adapt anything from
> > it, they are more than welcome to.
> >
> > A lot of what makes FT8 so robust is structured into the protocol
> > itself - small payloads, use of LDPC and FEC (Low density parity check
> > and forward error correction) as well as a priori on the decoded side
> > where cumulative decodes are used to guess the content of partial
> > decodes. Certainly, repeating CW with different variations in speed
> > and remembering messages from before and re-creating that is sort of
> > equivalent but adapting the techniques of FT8 to CW where you have
> > manual telegraphy probably isn't going to happen.
> >
> > But if you mean that nobody wants to pick up the source code and do
> > stuff with it, well maybe that's true. However there have been other
> > software being developed such as JTDX, MSHV and even WSJT-z which has
> > used the source code from WSJT-X. Just not for CW because of the
> > aforementioned challenges in the way FT8 is fundamentally defined.
> >
> > 73
> > Ria, N2RJ
> >
> > On Fri, Feb 18, 2022 at 12:59 PM David Gilbert <ab7echo@gmail.com> wrote:
> >>
> >> The signal processing techniques utilized in FT8 could have been used to
> >> enhance the signal-to-noise of normal CW in a way that would have been
> >> virtually transparent to the operator. It's truly a shame that WSJT-X
> >> has landlocked those techniques within FT8 in the minds of just about
> >> everyone in our hobby, as well as giving those techniques such a bad
> >> image that nobody seems willing to take up the mantle to fix it.
> >>
> >> The signal processing technology behind FT8 is awesome, but it is
> >> neither unique nor proprietary. It should have enhanced the hobby for
> >> all of us, especially for us contesters since almost all of our
> >> transmissions come from keypresses in a logging program anyway. Instead
> >> it is being held hostage to WSJT-X for no real reason other than inertia.
> >>
> >> 73,
> >> Dave AB7E
> >>
> >>
> >>
> >> On 2/18/2022 8:13 AM, James Cain wrote:
> >>> I agree with HA3LN that FT-8 is a terrible blow to what an ARRL director
> >>> snidely referred to -- on this reflector -- as "our sandbox." Yes, I play
> >>> only unassisted and you call it what you want.
> >>>
> >>> A neighbor guy dropped by and I had just set up my Superstation for this
> >>> weekend -- a TS590 on a card table and a 11-foot wire dropped out the 2nd
> >>> floor window. "I can talk to Europe, South America, maybe even Japan with
> >>> this" said I. "On Morse code."
> >>>
> >>> cain K1TN
> >> _______________________________________________
> >> CQ-Contest mailing list
> >> CQ-Contest@contesting.com
> >> http://lists.contesting.com/mailman/listinfo/cq-contest
>
_______________________________________________
CQ-Contest mailing list
CQ-Contest@contesting.com
http://lists.contesting.com/mailman/listinfo/cq-contest
|