[Elecraft] DSP value?

[Andrew Catanzaro]

Dale,

I understand what you are saying below.  I have never run into a
DSP unit, however, that ever left a CW element untainted.  It takes
a finite length of processing time for the signal to be identified as such
and passed through the output of the filter.  This tiny fraction of a dit
may make no difference at 20 WPM, but at higher speeds that rolloff
of the leading edge becomes a larger part of the same dit..
I find the effect to be a deterrent to comfortable copy.  At really high
speeds, it subtracts from intelligibility regardless of how high or low
the unprocessed S/N ratio is.  With all the traffic on this reflector in
the past couple weeks about the waveform of the transmitted K2
signal, I would think most ops would be more concerned about the
shape of the waveform hitting their ears on the receiving end.  I do not
own a storage oscilloscope, but I'm pretty sure my point could be
effectively displayed in a measurable format.  I have found that relatively
tight analogue filtering to be superior at increasing the S/N ratio compared
to digital noise subtraction or SCAF filtering.  If there are too many poles
of Butterworth, Chebyshev, or whatever filtering, there can be time domain
smearing of the CW elements, and the benefit of even this filtering goes
away.
The KAF2 is just right!

BUT, I know the guys at Elecraft make superior products and it would
be stupid of me judge anything they make before hearing it.  And of course,
to really analyze the KDSP2, I'll have to buy one.

Besides, I need something to build, and more important than that, I have to
support Elecraft so they keep producing wonderful radios.

73,
Andy W9NJY
Milwaukee WI


----- Original Message ----- 
From: "Dale Boresz" <dmb@lightstream.net>
To: "Andrew Catanzaro" <acatan@wi.rr.com>
Sent: Thursday, May 08, 2003 8:13 AM
Subject: Re: [Elecraft] DSP value?


> Andrew Catanzaro wrote:
>
> >Headphones are absolutely necessary for good CW copy because putting the
source of the CW at the ear eliminates multipart distortion that causes the
brain to work overtime in copying even in-the-clear signals.  So when I hear
something on the speaker I want to hear better, I just put the cans on my
head.  I leave the filtering wide-open on the TimeWave and use the KAF2 when
needed.
> >
> I agree completely. However, when you put on the headphones, you are
> also able to hear the noise surrounding the signal that much clearer as
> well. This is where the noise-reduction of the KDSP2 really helps; it
> reduces that surrounding noise to a level that makes it much less
> distracting, without changing the level of the desired signal. The
> result is that your ear/brain doesn't have to work as hard to separate
> the signal from the noise, and copy is far more enjoyable.
>
> >It's a great filter that doesn't ring, and there's no digital bandwidth
> >that's narrower that's more useful!
> >
> I have been a long-time fan of the KAF2 (being primarily a cw op), and
> being one of the KDSP2 field testers, I have also had the opportunity to
> work with the KDSP2 on the air for some weeks now. While I've always
> liked the character of the audio with the KAF2 filter engaged, it
> accomplishes this at the expense of reduced bandwidth - whether the
> reduction is needed or not. Many times I have engaged the KAF2's filter
> only to help lift a weak signal out of the noise, even when there were
> no other signals around. Being one who prefers a wider bandwidth for
> general operation, I find the ability of the KDSP2 to reduce noise
> independently of bandwidth, to be one of it's most appreciated features.
>
> In short, the KAF2 is a wonderfully elegant and effective product, but
> the versatility and effectiveness of the KDSP2 is what keeps it firmly
> attached to my K2.   :-)
>
> 73, Dale - WA8SRA
>