Re: CW-to-RTTY with message buffers: are spaces there?

[Tom Hammond-2]

OK... I was gonna hold off commenting here, but NOW Bill, Rick and
Vic have weighed in I've gotta do the same... <G>

>K2VCO wrote:
>I had quite a bit of a problem *recording* messages because of a
>similar phenomenon. For example, the K3 understands my call as K2
>VCO. The extra space that I send between the 2 and the V is more
>than a letter space (apparently) but less than a word space. The K3
>records it as a word space, so when I play it back it sounds like
>that. When I tried to correct this, I ended up sending less than a
>normal letter space, the K3 did not understand my sending and recorded nothing!

I had SIGNIFICANTLY more problems recording using the original K3
algorithms than I do now. Now, unless I make a keying error (e.g.
wrong character, etc.) I can always record a 'clean' message...
before, I had to make MANY(!) attempts.

Obviously I must be doing something wrong... <G>

Not sure if a 'real time load' feature could be added without
significant complication, but that might be the only way to make some
folks happy.  Of course, we'll NEVER be able to satisfy everyone...

>One way to handle this would be to use "1/4 letter space" as the
>basic spacing element. Then recording, decoding, and playback would
>be closer to real-time.

Can you 'splain this a bit better Vic? Never heard of this method of
timing... always worked on the baud method.

Tom



_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
 http://mailman.qth.net/mailman/listinfo/elecraft   

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

[Vic K2VCO]

Tom Hammond wrote:

>> One way to handle this would be to use "1/4 letter space" as the basic
>> spacing element. Then recording, decoding, and playback would be
>> closer to real-time.
>
> Can you 'splain this a bit better Vic? Never heard of this method of
> timing... always worked on the baud method.

First I have to reverse-engineer Wayne's code!

The firmware now listens to the spaces between characters as you send.
If the space is less than some fixed threshold it assumes that it is a
letter space; if more, a word space.

When it records what you have sent, it puts in a token that indicates
either a word space or a letter space is needed. During playback, when
it comes to a letter-space token it waits for the appropriate time
(whatever that is);  when it comes to a word-space token it waits longer.

This means that while you are recording, if you wait a little bit past
the threshold that it uses to decide what size space it is, it will put
in a word space token. So when I send my call K 2VCO, I get K   2VCO.

My suggestion is that instead of just two possibilities for spacing,
there should be more. This could be implemented by creating tokens for
smaller spaces, like 1/4 of a letter space. That would sound almost like
a real-time recording, in terms of preserving the spacing as sent. But
even using units of half of a letter space would avoid the difficulty
that those of us who do not send like tape machines have when recording
messages!
--
73,
Vic, K2VCO
Fresno CA
http://www.qsl.net/k2vco
_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
 http://mailman.qth.net/mailman/listinfo/elecraft   

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

[w7aqk]

I missed some of this discussion, so I may be off track
here.  But I think I know what Vic means.  Some message
memories record in real time--you leave a long space, and it
shows up in your memory the same way.  But the K3 seems to
be tied to a rather precise gap between letters or words.

Actually, this kind of method will help point out flaws in
the way a person sends.  Some time back I was trying to
record a "CQ" message in a keyer memory, and I discovered I
had developed a tendency to leave a bit too much space
between my next-to-last call letter and the last letter.  I
was sending "W7AQK" as "W7AQ K".  I probably did it for
years!  It was also possibly one of the reasons so many
people came back to me as "W7AQ", thinking I had sent the
"K" as an "invitation to transmit".   In contests, etc.,
they weren't even waiting for the second "K" at the end.
Now that I've been paying more intention (and it hasn't been
easy getting out of that habit), I don't seem to get as many
incorrect responses.

The message buffers in the K3 are pretty large, so they make
a pretty good recording device for recording yourself and
checking on your sending technique.  I know this method is
kind of a pain in the butt for a lot of us, but I really
like it if my messages come out like machine code.  I'm not
sure other methods would do this as effectively.

Another thing I've been playing with a bit is sending CW and
letting the K3 convert it to RTTY.  If you don't move right
along, with proper spacing, you get a bunch of extra garbage
between characters.  Being able to send RTTY without a
computer is sort of a "gee whiz" feature, but it's been fun
playing with it.

Dave W7AQK

----- Original Message -----
From: "Vic K2VCO" <[hidden email]>
To: "Tom Hammond" <[hidden email]>
Cc: <[hidden email]>
Sent: Friday, January 11, 2008 3:12 PM
Subject: [Elecraft] Re: CW-to-RTTY with message buffers: are
spaces there?

> Tom Hammond wrote:
>
>>> One way to handle this would be to use "1/4 letter
>>> space" as the basic spacing element. Then recording,
>>> decoding, and playback would be closer to real-time.
>>
>> Can you 'splain this a bit better Vic? Never heard of
>> this method of timing... always worked on the baud
>> method.
>
> First I have to reverse-engineer Wayne's code!
>
> The firmware now listens to the spaces between characters
> as you send. If the space is less than some fixed
> threshold it assumes that it is a letter space; if more, a
> word space.
>
> When it records what you have sent, it puts in a token
> that indicates either a word space or a letter space is
> needed. During playback, when it comes to a letter-space
> token it waits for the appropriate time (whatever that
> is);  when it comes to a word-space token it waits longer.
>
> This means that while you are recording, if you wait a
> little bit past the threshold that it uses to decide what
> size space it is, it will put in a word space token. So
> when I send my call K 2VCO, I get K   2VCO.
>
> My suggestion is that instead of just two possibilities
> for spacing, there should be more. This could be
> implemented by creating tokens for smaller spaces, like
> 1/4 of a letter space. That would sound almost like a
> real-time recording, in terms of preserving the spacing as
> sent. But even using units of half of a letter space would
> avoid the difficulty that those of us who do not send like
> tape machines have when recording messages!
> --
> 73,
> Vic, K2VCO
> Fresno CA
> http://www.qsl.net/k2vco
> _______________________________________________
> Elecraft mailing list
> Post to: [hidden email]
> You must be a subscriber to post to the list.
> Subscriber Info (Addr. Change, sub, unsub etc.):
> http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/subscribers.htm
> Elecraft web page: http://www.elecraft.com
>

_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
 http://mailman.qth.net/mailman/listinfo/elecraft   

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com