Elecraft › [K3]

K3 - RX audio degradation with ANF

Classic

List

Threaded

11 messages Options

Bill Clarke

K3 - RX audio degradation with ANF

I am reposting this question, as I received no responses the first time:

I generally only use the ANF when it is needed. I do not run it all the
time. The reason I do not run the ANF all the time is a slight
degradation of the quality of the SSB received audio when it is engaged.
Although it does not interfere with the intelligibility of the audio -
it does introduce a fuzzy aspect to otherwise armchair copy.

Of note: On my Kenwood TS-480, I leave the ANF (Beat Cancel in Kenwood
speak) on at all times. It does its job as well as the K3 - and is
without the degradation factor.

Any suggestions regarding the use of the ANF that I would benefit from?

Thanks,

Bill W2BLC

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

Don Wilhelm-4

Re: K3 - RX audio degradation with ANF

Bill,

Please enlighten me - I have searched through the K3 manual, and do not
find anything labeled "ANF"
If you would please give whatever you are trying to run the same label
used in the K3 manual perhaps we can provide some answer.

73,
Don W3FPR

On 5/18/2012 7:52 AM, Bill wrote:
> I am reposting this question, as I received no responses the first time:
>
> I generally only use the ANF when it is needed. I do not run it all the
> time. The reason I do not run the ANF all the time is a slight
> degradation of the quality of the SSB received audio when it is engaged.
> Although it does not interfere with the intelligibility of the audio -
> it does introduce a fuzzy aspect to otherwise armchair copy.
>
>
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

Arthur Burke

Re: K3 - RX audio degradation with ANF

My reaction was the same - I think he's referring to Automatic Notch Filter.

Art - N4PJ

On Fri, May 18, 2012 at 8:12 AM, Don Wilhelm <[hidden email]> wrote:

> Bill,
>
> Please enlighten me - I have searched through the K3 manual, and do not
> find anything labeled "ANF"
> If you would please give whatever you are trying to run the same label
> used in the K3 manual perhaps we can provide some answer.
>
> 73,
> Don W3FPR
>
> On 5/18/2012 7:52 AM, Bill wrote:
> > I am reposting this question, as I received no responses the first time:
> >
> > I generally only use the ANF when it is needed. I do not run it all the
> > time. The reason I do not run the ANF all the time is a slight
> > degradation of the quality of the SSB received audio when it is engaged.
> > Although it does not interfere with the intelligibility of the audio -
> > it does introduce a fuzzy aspect to otherwise armchair copy.
> >
> >
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:[hidden email]
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

Bill Clarke

Re: K3 - RX audio degradation with ANF

In reply to this post by Don Wilhelm-4

That would be the button labeled as NTCH. It is described on page 25 of
the owner's manual as Auto Notch.

Bill W2BLC
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

Gary Gregory

Re: K3 - RX audio degradation with ANF

In reply to this post by Don Wilhelm-4

*Don,

your not alone but I guessed it was Auto-Notch Filter..:-)

Bill is correct though, there is distortion when I have AGC=12 and SLP=2
but if I go to AGC=8 it appears to be almost gone.

Not sure why this is?

73's
Gary
*
On 18 May 2012 22:12, Don Wilhelm <[hidden email]> wrote:

--
Gary
VK4FD - Motorhome Mobile
Elecraft Equipment
K3 #679, KPA-500 #018
Living the dream!!!
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

Tim Tucker

Re: K3 - RX audio degradation with ANF

ANF is the common use abbreviation for Auto Notch Filter. A lot of rigs
use that nomenclature.

The K3's ANF on SSB has not worked correctly IMO for several firmware
releases. Someone posted a fix a while back that involved resetting back
to factory settings, rolling back to a previous firmware, reloading all
your settings and then loading the current firmware. That's a lot of work
and I was hoping a more elegant fix could be implemented in a firmware
update.

In my experience, the issue isn't so much extreme distortion on the signal,
it's that it doesn't notch the offending signal properly. To reproduce,
watch your RX signal with a strong carrier on a waterfall with a computer
application like MixW. Turn on the ANF and you'll see that the K3 notches
on either side of the offending carrier, but not actually right on the
carrier. If you have good ears, you can hear this yourself. The manual
notch works fine, of course.

I haven't tried the latest beta firmware, so I don't know if this issue has
been addressed yet. It needs to be addressed, though.

Tim

On Fri, May 18, 2012 at 5:35 AM, Gary Gregory <[hidden email]> wrote:

> *Don,
>
> your not alone but I guessed it was Auto-Notch Filter..:-)
>
>
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

Bill K9YEQ

Re: K3 - RX audio degradation with ANF

Tim,

I use the "ANF" regularly on 75 meters to block heterodynes and have no
issues or distortion and it works extremely well.

I use AGC settings:

Nor decay
.30 HLDA
Nor PLS
004 SLP
007 THR
030 S

I use 2.8 8 pole filter and on occasion a 1.8 8 pole filter. I have my PRE
amp off and ATT on, generally as the noise floor is around 1-2 S units using
my ~ 350' horizontal loop. My inverted V Double Bazooka has a slightly
higher noise floor. Again, I see no issues and am using the latest
firmware. The notch removes the offending notch, but not the signal itself,
of course. I also have a P3. I think the Notch is working very well as
currently designed.

73,
Bill
K9YEQ

-----Original Message-----
From: [hidden email]
[mailto:[hidden email]] On Behalf Of Tim Tucker
Sent: Friday, May 18, 2012 10:57 AM
To: Elecraft Reflector
Subject: Re: [Elecraft] K3 - RX audio degradation with ANF

ANF is the common use abbreviation for Auto Notch Filter. A lot of rigs use
that nomenclature.

The K3's ANF on SSB has not worked correctly IMO for several firmware
releases. Someone posted a fix a while back that involved resetting back to
factory settings, rolling back to a previous firmware, reloading all your
settings and then loading the current firmware. That's a lot of work and I
was hoping a more elegant fix could be implemented in a firmware update.

In my experience, the issue isn't so much extreme distortion on the signal,
it's that it doesn't notch the offending signal properly. To reproduce,
watch your RX signal with a strong carrier on a waterfall with a computer
application like MixW. Turn on the ANF and you'll see that the K3 notches
on either side of the offending carrier, but not actually right on the
carrier. If you have good ears, you can hear this yourself. The manual
notch works fine, of course.

I haven't tried the latest beta firmware, so I don't know if this issue has
been addressed yet. It needs to be addressed, though.

Tim

On Fri, May 18, 2012 at 5:35 AM, Gary Gregory <[hidden email]> wrote:

> *Don,
>
> your not alone but I guessed it was Auto-Notch Filter..:-)
>

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

Lu Romero - W4LT

Re: K3 - RX audio degradation with ANF

In reply to this post by Bill Clarke

I hear it. I have always heard it. Its always been there.

I think the "distortion" you are hearing and seeing in the
waterfall you looked at is the notch in the audio itself and
the ANF routine's "hunting around" for the frequency to
notch as it samples the audio its being fed.

Within the audio spectrum of a voice there are many
frequencies, and they are constantly changing as speech is
being produced. The tone (of the carrier) you are trying to
notch also falls within those frequencies. The notch
produced by the ANF has a finite "width" and "depth" plus or
minus several cycles, as you noticed in the waterfall, which
has a rise time and a fall time. When that is superimposed
over a voice, there is a "comb filter" like effect in some
frequencies adjecent to the tone you are notching as the
automatic notch routine samples and then attempts to blank
the interfering tone.

We are talking about notching audio here, ANF cant magically
get rid of just the tone and leave what is being covered up
by the tone alone, it has to get rid of ALL the energy in
the spectrum occupied by the tone, plus or minus the notch
filter width and depth, so there literally is a "hole"
there, and it "moves" as the voice and the tone mix. You
hear the hole, and notice that "something is missing". The
hole "moves around" as the interference is blanked and the
ANF refreshes its decision at whatever frequency it samples
of what it needs to blank. That's why you dont see it in
the manual notch, that one does not move around on its own.

At least this is my long winded "guess" to what one hears
here. K9YC may have a better way of explaining this than me
(Im a duffer, he's a pro at this game).

Ive always heard this to a greater or lesser extent in any
radio or device that notches audio frequencies
automatically. So called "feedback destroyers" in stage and
studio audio have the same function as ANF, I hear this
effect on them as well to a greater or lesser extent. They
have gotten so good lately that it is barely preceptible,
but its there, you can see it on a spectrum analyser.
(Google Sabine Adaptive Audio to see a representative
device, their older devices sound quite similar to what you
hear in the K3 ANF; the newer ones are truly magical!).

I dont think that it is *THAT* objectionable... It is there,
I hear it, but I know why it is happening. Is it normal?
Maybe. Could it be better? Possibly. I find it acceptable.
Its a great aural reminder that you have the auto notch
turned on!

-lu-w4lt-

-----------------------

-----Original Message-----
From: [hidden email]
[[hidden email]] On Behalf Of Tim Tucker
Sent: Friday, May 18, 2012 10:57 AM
To: Elecraft Reflector
Subject: Re: [Elecraft] K3 - RX audio degradation with ANF

ANF is the common use abbreviation for Auto Notch Filter. A
lot of rigs use
that nomenclature.

The K3's ANF on SSB has not worked correctly IMO for several
firmware
releases. Someone posted a fix a while back that involved
resetting back to
factory settings, rolling back to a previous firmware,
reloading all your
settings and then loading the current firmware. That's a
lot of work and I
was hoping a more elegant fix could be implemented in a
firmware update.

In my experience, the issue isn't so much extreme distortion
on the signal,
it's that it doesn't notch the offending signal properly.
To reproduce,
watch your RX signal with a strong carrier on a waterfall
with a computer
application like MixW. Turn on the ANF and you'll see that
the K3 notches
on either side of the offending carrier, but not actually
right on the
carrier. If you have good ears, you can hear this yourself.
The manual
notch works fine, of course.

I haven't tried the latest beta firmware, so I don't know if
this issue has
been addressed yet. It needs to be addressed, though.

Tim

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

Bill Clarke

Re: K3 - RX audio degradation with ANF

You said, "They have gotten so good lately that it is barely
preceptible, but its there, you can see it on a spectrum analyser.
(Google Sabine Adaptive Audio to see a representative device, their
older devices sound quite similar to what you hear in the K3 ANF; the
newer ones are truly magical!)."

Professional musicians are the pickiest sound folks in the world. What
the K3 does would be most unacceptable to them and I can say that, for
sure, I have never heard similar at any professional music venue. The
musicians would strangle the audio tech if it were so.

All professional automatic notch filtering used in the music field is
done at the "what we call the audio level." It is their ONLY level, as
they are not working with IFs etc. Hence, I can only assume that the
audio level automatic notch filtering found in the K3 is not yet up to
par. I say "yet" as that is the advantage of updates. At some point in
the future, the ANF can be brought up to higher standards.

In the interim, there are AGC settings that can be varied - resulting in
more acceptable listening. Specifically, using AGC SLP level at 2 makes
a major improvement to my ears.

--
IN GOD I TRUST (but, NOT a single politician)
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

Jim Rogers, W4ATK

Re: K3 - RX audio degradation with ANF

I think we may be comparing apples to oranges here.

In the K3 digital ANF, the notch filter is looking for coherent signals. The premise is that audio is less coherent than a CW carrier. Thus the ANF "finds" the coherent CW carrier and notches it out, leaving the remaining audio.

On May 19, 2012, at 4:27 PM, Bill wrote:

> You said, "They have gotten so good lately that it is barely
> preceptible, but its there, you can see it on a spectrum analyser.
> (Google Sabine Adaptive Audio to see a representative device, their
> older devices sound quite similar to what you hear in the K3 ANF; the
> newer ones are truly magical!)."
>
> Professional musicians are the pickiest sound folks in the world. What
> the K3 does would be most unacceptable to them and I can say that, for
> sure, I have never heard similar at any professional music venue. The
> musicians would strangle the audio tech if it were so.
>
> All professional automatic notch filtering used in the music field is
> done at the "what we call the audio level." It is their ONLY level, as
> they are not working with IFs etc. Hence, I can only assume that the
> audio level automatic notch filtering found in the K3 is not yet up to
> par. I say "yet" as that is the advantage of updates. At some point in
> the future, the ANF can be brought up to higher standards.
>
> In the interim, there are AGC settings that can be varied - resulting in
> more acceptable listening. Specifically, using AGC SLP level at 2 makes
> a major improvement to my ears.
>
> --
> IN GOD I TRUST (but, NOT a single politician)
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:[hidden email]
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html

Lu Romero - W4LT

Re: K3 - RX audio degradation with ANF

In reply to this post by Bill Clarke

I agree, in more ways than one, that we are talking apples
and oranges here. The code between the devices has similar
applications, but a Sabine Feedback Destroyer has multiple
instances operating in a dedicated processor with nothing
else to do but find the "carriers" while the K3 ANF is
sharing its runtime with lots of other "priorities".

Another issue is that the ANF cannot "remove" the audio
covered up by the interfering carrier, NOTHING can do this.
All ANF can do is throw a notch over the interfering
frequency... That is what it does, it cant actually REMOVE
the coherent interferance from the "incoherent"
intelligence.

I was just at the NAB show where I saw a demonstration of a
technology that does extract the underlying audio, which was
truly magic. Also saw a demo of a video editing system
(Adobe Premiere) that has a similar feature built in. Lots
of non real time processing is needed to do this magic,
however.

If the ANF can be made better, great. I find it acceptable.
Not perfect, but acceptable. If making it better doubles
the price of the radio, then Im not for it, for what its
worth :)

-lu-w4lt-

Message: 24
Date: Sat, 19 May 2012 16:50:24 -0500
From: W4ATK <[hidden email]>
Subject: Re: [Elecraft] K3 - RX audio degradation with ANF
To: Bill <[hidden email]>
Cc: [hidden email]
Message-ID:
<[hidden email]>
Content-Type: text/plain; charset=us-ascii

I think we may be comparing apples to oranges here.

In the K3 digital ANF, the notch filter is looking for
coherent signals. The premise
is that audio is less coherent than a CW carrier. Thus the
ANF "finds" the
coherent CW carrier and notches it out, leaving the
remaining audio.

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html