AaronJ wrote:
John Smith I wrote:

Minor, inconsequential and random errors are easily programmed into the
computer generated model, but will give the morse that "unique
signature" of the "imitated keyers style."

IMO the perfect fist sounds like computer generated CW. And it's the easiest to
copy. All those so called 'unique fists' can be copied but it's like trying to
understand someone from Brooklyn (or Texas)...

Yanno, after reading enough of these posts, you are likely to come to
the same conclusion I have; these guys think a ham ticket is the
equivalent to a "Doctorate In Everything."

Once you catch on, it is quite funny to watch, and indicative of what
kind of educational level is predominate here.

Regards,
JS

John Smith I wrote:

AaronJ wrote:
they are using a computer. I've even read posts here of hams that said if they
suspect someone is copying CW by computer they switch to excessive weight and
poor sending just to mess up the copy. Seems kind of snobbish to me.

They (those guys with the brass taps on, doing a tap dance here) truly
show their ignorance if that is their claim. They get away with that
chit because few have professional programming experience...

The programming experience of most hams has little to do with their ability to
work computer CW. Virtually everybody I work (who admits to) using a computer on
CW is using either a *commercial* multimode decoding box or *commercial*
computer software and an interface.

If the ear can tell the difference between a di and a dah, the computer
sure as heck will not be fooled!

IMO the ear still beats computer copy in *real life*. I've played with a lot of
multimodes and software over the years and none has ever come close to ear copy
under poor signal conditions and/or someone with a poor fist.

This would only require that polymorphic coding be used in the software...

My computer screen might read NNTTA when the op was really sending CQ. The
computer saw five letters when there were really two because the op had poor
letter spacing. My ear picked up the CQ easily, but the computer read gibberish.
When I contact him I will still be able to understand him while the computer
continues to spit out gibberish. Your program is going to need to understand
English grammar and Q signals along with timing to solve this
type of problem...

Any software engineer who possesses a bachelors degree will support
this, even if they know NOTHING about amateur radio and ONLY that there
will ALWAYS be a difference in length between a di and a dah and these
length differences are (or may be) of a constantly variable nature.

It may be possible to write such a *perfect* CW program, but so far I haven't
found one. Maybe someone else here has. So for our victim who is using currently
available software it's certainly possible for the jerks to screw up his copy
with poor sending. But lets hope they were just spouting off in the post.

BTW Some advantages of computer CW:
It's fun to have a 80 or 100+ WPM QSO.
It's a great way to get your code speed up. Watch the screen while you listen to
the code at just above your speed. After awhile turn the screen away and you
will be copying in your head at the new speed.

A disadvantage of using computer CW:
During a QSO you can't turn up the speaker and leave the room
to take a leak...

AaronJ wrote:
...

I don't use commercial ware. I do have some open source code of others
I have used and incorporated into C/C++ programs I have created myself.

A search of the net will turn up many freeware and open source programs
capable of fooling the best, I can hardly believe there is no
commercial-ware which can't, perhaps it is in its use and configuration
by the user.

And trust me, long after your ear fails to copy solid, a good software
utility can ...

Regards,
JS

AaronJ wrote:
I was simply referring to those who refuse to work someone on CW *only* because
they are using a computer.

Anyone using an IC-756PRO for CW is using a computer. :-)
--
73, Cecil http://www.w5dxp.com

AaronJ wrote:
Cecil Moore wrote:
That's the point I was trying to make.

My apologies, I thought you were serious.

Seriousness on r.r.a.p???? Shirley you jest. :-)
--
73, Cecil http://www.w5dxp.com

John Smith I wrote:
Cecil Moore wrote:
Straight keys are relatively easy to recognize.
Sometimes it is impossible to program a machine
to be that bad on purpose. :-) I heard a guy on
straight key night where his dits were 75% the
length of his dahs - amazingly hard to copy.

You find me a hacker (i.e. "Professional Software Engineer") who says
that would be even above childs play and I will give up my anonymity.
On a complexity scale of 1-to-10 that does NOT even register ...

Methinks you misunderstood. My point is that a computer
is absolutely necessary to be able to copy that guy's
bad fist. I doubt that any human ears could do it.
--
73, Cecil http://www.w5dxp.com

AaronJ wrote:
John Smith I wrote:

Minor, inconsequential and random errors are easily programmed into the
computer generated model, but will give the morse that "unique
signature" of the "imitated keyers style."

IMO the perfect fist sounds like computer generated CW. And it's the easiest to
copy. All those so called 'unique fists' can be copied but it's like trying to
understand someone from Brooklyn (or Texas)...

Well, there was some debate about this a few years back. It centered
around a couple of things; 1) being a lack of an actual definition of
Morse Code in Title 47, and 2) the desire of several of the Pro-Code
Test folks to claim that a method of TEACHING Morse Code should be used
as a Morse Code Exam, i.e., the Farnsworth Code.

Morse Code had previously been defined with specific dot, dash, and
space interval ratios. Exams were then defined as Morse Code sent at
rates of 20, 13, and 5 WPM. Today, they use the Teaching Method of
Farnsworth Code, where the dot, dash, and interval can be anything
desired, and character speeds of 13 to 15 WPM for a 5 WPM exam. That's
fine for learning the code as Part 97 doesn't address any particular
method, nor does it advocate any particular vendor such as W5YI or
ARRL.

Yet Part 97 still, even to this day, requires a Morse Code Exam
(Farnsworth isn't mentioned) at a Morse Code Rate of 5 WPM.
Lengthening the space interval isn't addressed as a way to get 13 to 15
WPM character speeds down to 5 WPM word rate. But hey, Part 97 is only
a suggestion, right?

And none of that matters now, anyway.

Anyhow, the Pro-Code Exam folks were all over the notion that code was
an individual thing and that each person's code sounded like "speech"
to them because of all of the little and big imperfections, and
sometimes the big imperfections were deliberate. I was chided for
suggesting that manually sent code should be formed as precisely as one
could make it, which sparked another debate. Apparently, humans trying
to send perfect code shouldn't be a goal in amateur radio, even if
unachievable. Which took us full circle to the humans emulating modems
of the original invention of Sammy Morse, the code paper tape with
dashes and longer dashes scribed on them.

Oh, well, it was a nice walk in the park. The PCTA folks arguments
were as imperfect as the code they send.

John Smith I wrote:

I don't use commercial ware.

I once wrote a homebrew CW receive program in Atari Basic that actually did
pretty well on my old 800XL (if the conditions and sending fist were pretty
good).

I have used and incorporated into C/C++ programs I have created myself.

But I think you will find that most hams are like me. They are not
professionally in electronics or computers, and thus buy mostly commercial
radios and software.

A search of the net will turn up many freeware and open source programs
capable of fooling the best,

Do you mean a CW program that prints perfectly under all conditions found on the
ham bands? I assure you that no such animal exists.

I can hardly believe there is no commercial-ware which can't,

I doubt that there's much of a market for CW software. The vast majority of CW
ops are over 60 and dying off fast. Then the FCC won't force newcomers to learn
the code anymore so that market is gone. (I sure wouldn't have learned it if
they hadn't forced me to.) Since there's not much money to be made on a dying
market there's not much incentive to invest a lot of time writing the perfect
program. And I suspect most CW ops are like me and probably wouldn't buy it
anyway cause we would prefer to do it the old fashioned way anyway...

And trust me, long after your ear fails to copy solid, a good software utility can ...

We'll just have to agree to disagree on this one. IMO the old computer between
the ears still does the best all around job, especially under poor conditions.

Cecil Moore wrote:

Anyone using an IC-756PRO for CW is using a computer. :-)

And it can be handy. Sometimes when I'm playing with the computer CW program and
it's printing well, I can have a short conversation with the XYL and then catch
up on the QSO a minute later. Some oldtimers used to claim to be able to head
copy on the radio and chat with a person in the room at the same time. But I
could never do that...

wrote:

AaronJ wrote:

IMO the perfect fist sounds like computer generated CW.

Well, there was some debate about this a few years back.

There's always been a debate about what is perfect code. Generally the argument
is over weighting, but sometimes spacing too. That's why I covered my assertion
with an IMO...

Test folks to claim that a method of TEACHING Morse Code should be used
as a Morse Code Exam, i.e., the Farnsworth Code.

I have no opinion on the best way to learn the code. I learned my code the old
fashioned way, one letter at a time back in 1957. And I passed my test decoding
and writing one letter at a time. But in practice I don't it one letter at a
time.

folks were all over the notion that code was
an individual thing and that each person's code sounded like "speech"

When you learn to head copy (no paper), eventually you will not decode the
individual letters but rather the words. The words have their distinct sounds.
And even if someone uses a word you don't know, the code sequence sounds like
phonics in your head and you still 'hear' the word. For this reason some say
that CW is a pseudo-language. That may be a poor word for it but you get the
idea. And if you talk to the same guy or group of guys regularly you will learn
to recognize their 'voices' on CW.

I was chided for suggesting that manually sent code should be formed
as precisely as one could make it,

IMO you should try to send like a computer...

Which took us full circle to the humans emulating modems

I like CW only because it's fun *to me*. The analogy might be that going
someplace by horse is not efficient, but it's sure fun...