| |
PA2AGA > HDDIG 06.07.00 02:09l 209 Lines 7243 Bytes #-9409 (0) @ EU
BID : HD_2000_180E
Read: GUEST
Subj: HamDigitalDigest 2000/180E
Path: DB0AAB<DB0SL<DB0RGB<DB0MRW<DB0ERF<DB0BRI<DB0SM<PI8DAZ<PI8GCB<PI8HGL
Sent: 000705/2114Z @:PI8HGL.#ZH1.NLD.EU #:58937 [Den Haag] FBB $:HD_2000_180E
From: PA2AGA@PI8HGL.#ZH1.NLD.EU
To : HDDIG@EU
Date: Wed, 05 Jul 00 21:28:14 MET
Message-Id: <hd_2000_180E>
From: pa2aga@pe1mvx.ampr.org
To: hd_broadcast@pa2aga.ampr.org
X-BBS-Msg-Type: B
Gary human operators running CW with "matched filters" can perform just as
well as PSK-31! No doubt about that. The only problem is for most hams is
that they don't have sufficient CW skills to do it themselves. That's just
a fact of skill training not an indictment of CW as a digital mode. I
maintain that CW is just as efficient as PSK-31 when received by machine in
the same bandwidths and superior when recevied by a skilled human operator.
Now if we can only learn what the human operator does instinctively while
copying under those conditions maybe we could find more economical and
robust machine algorithm decoders for digital modes!
What say you? Is it not worth study? What does a skilled human CW operator
do that is superior to the receiver implementations of PSK-31?
Now I am not putting down PSK-31, it is indeed a fine mode. In fact because
it approaches, but does not match human operators using CW, it gives many
the opportunity to experience the thrill and effectiveness of operating a
highly efficient digital mode on the HF channel without having to learn the
skills of copying Morse effectively by ear.
[snip]
>
> Matched filter technique cannot be applied to CW because the waveform
> has no a priori known spectrum when unknown text is sent. (If you know the
> message that is to be sent, you can form a matched filter for the
transmission,
[snip]
Gary, in practice, as you know there is not real "matched" filter, since the
"matched" filter would be anti-causal. All so-called matched filters are
merely approximations to an ideal matched filter. Gary not very many are
even good approximations. In fact in most modern "digital" receivers the
so-called matched filter is just a duplicate of the transmitter filter! Not
matched in the ideal sense but "close". For instance the "easy" way that
most digital designers use is to put a so-called "raised cosine" filter in
the transmitter, with alpha around 0.1 - 0.12 or so... and then they put the
same damm filter in the receiver! That's not "matched" and you know it! I
know it because I have designed many of them....
When it comes to "matched filters" Gary, "close" is what is wanted, and
"close" is all that's needed.
Gary a the 500 Hz, 250 Hz and 100 Hz CW filters on my HF receivers are very
very close to an ideal matched filter for CW at hand sending speeds of 18 -
25 wpm. In fact they are a much closer approximation to a "matched" filter
than many of the so-called matched filters in the modern digital mode
decoders! Lie the "raised cosine" approximations I mentioned above.
BTW... have you ever used the narrow CW filters on your HF receiver, if so
how often? etc... etc... They are pretty damm good approximations to
matched filters.
[snip]
erformance. CW just isn't as good as some of the more modern
> techniques.
>[snip]
Yes it is...
[snip]
> we may
> continue to use CW if we like. But it is a mistake to try to claim for
> it a superiority that it does not enjoy in fact.
[snip]
Wrong! Gary CW when run at comparable data rates in the same bandwidths
with appropriate filters and highly skilled operators is almost as efficient
as most of the so-called new digital modes...
And what is more, we stand to learn something from observing human operators
and studying how they can beat some of the best modern decoding algorithms.
Perhaps we could learn to implement similar algorithms to those found in the
brains of the best human digital operators.
Peter [CW is digital] K1PO
------------------------------
Date: Tue, 4 Jul 2000 14:05:31 -0500
From: "Peter O. Brackett" <ab4bc@ix.netcom.com>
Subject: CW versus hi speed digital etc.
Steve;
Does that include Florida?
Heh, heh!
Peter [Not a yankee] K1PO
Madison, AL [near Rocket City, USA]
"Steve Sampson" <ssampson@usa-site.net> wrote in message
news:sm4a14g1op326@corp.supernews.com...
> > Peter ["CW is digital"] K1PO
>
> There's nothing worse than Yankee's in Alabama...
>
>
>
------------------------------
Date: Tue, 4 Jul 2000 15:45:34 -0400
From: "Bob Lewis" <aa4pb@erols.com>
Subject: CW versus hi speed digital etc.
Gary,
You might as well give up. No amount of logic is ever going to
convince him that CW isn't the most reliable, most efficient mode ever
to be developed. The developers of network systems made a huge mistake
in not using CW for all computer networking applications :-)
By the way, after years in government communications (code school and
copying 5 letter groups), this NG is the first time I ever heard
anyone say that 5 letter code groups contained FEC. If its true I
wonder why they never taught us that.
------------------------------
Date: Tue, 4 Jul 2000 19:29:35 -0400
From: "Bob Lewis" <aa4pb@erols.com>
Subject: CW versus hi speed digital etc.
> I just said that we all have a lot to learn from examining CW and
why it
> performs so well given that it seems [at least superficially] to be
a very
> simple digital system in comparison to the great complexity inherent
> in the new modes...
>
> OK??
Ok, I can agree with that, Peter. Simplicity for the given performance
is CW's strongest point.
------------------------------
Date: Tue, 4 Jul 2000 18:14:22 -0500
From: "Peter O. Brackett" <ab4bc@ix.netcom.com>
Subject: CW versus hi speed digital etc.
Bob:
Hey man, I never said that CW is the greatest of all the digital modes!
You'r giving me a bum rap here...
I just said that we all have a lot to learn from examining CW and why it
performs so well given that it seems [at least superficially] to be a very
simple digital system in comparison to the great complexity inherent in the
new modes...
OK??
Peter ["The digital guy"] K1PO
"Bob Lewis" <aa4pb@erols.com> wrote in message
news:8jtess$p52$1@bob.news.rcn.net...
> Gary,
>
> You might as well give up. No amount of logic is ever going to
> convince him that CW isn't the most reliable, most efficient mode ever
> to be developed. The developers of network systems made a huge mistake
> in not using CW for all computer networking applications :-)
>
> By the way, after years in government communications (code school and
> copying 5 letter groups), this NG is the first time I ever heard
> anyone say that 5 letter code groups contained FEC. If its true I
> wonder why they never taught us that.
>
>
>
>
------------------------------
Date: Wed, 05 Jul 2000 09:51:47 GMT
From: hamish@cloud.net.au (Hamish Moffatt VK3SB)
Subject: CW versus hi speed digital etc.
Peter O. Brackett <ab4bc@ix.netcom.com> wrote:
> Often the mode comparators, opportunistically take the "modes" apart in
> their comparisons and compare a whole "mode" to a modulation. Say in this
> case comparing OOK to PacTor. OOK is just a modulation, while PacTor is a
> whole mode repleted with several stages of encoding, embedded modualtion and
> channel coding, etc...
OK, but that's exactly what YOU are doing. When someone says CW uses
on-off keying and is inefficient, you say CW could easily be applied to
To be continued in digest: hd_2000_180F
Read previous mail | Read next mail
| |
