Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

   Home   Help Search  
Pages: Prev 1 [2] 3 Next   Go Down
  Print  
Author Topic: Sidetone Frequency  (Read 11403 times)
NI0C
Member

Posts: 2403




Ignore
« Reply #15 on: November 19, 2010, 06:05:50 AM »

Quote
I prefer, if at all possible, to run my bandwidth as wide as possible in CW. If I can hear a couple stations above me and below me in a contest, then I start to feel that I have a broad awareness.

There are other ways to achieve this "broad awareness." For example, as I write this I'm watching a pileup on ZL8X on my new P3 panadaptor while listening to ZL8X in a 300 Hz DSP filter bandwidth.  Pretty cool.

Quote
I also have a very strong distaste for brick wall filters because they have abysmal group delay properties.

I suppose you as a CW operator are concerned about filter ringing.  The Finite Impulse Response (FIR) DSP filters have linear phase characteristics (constant delay) as well as "brick wall" amplitude response.  You'll find FIR filters in the K3 and other radios with IF DSP, as well as in some of the outboard audio DSP units (I used to use the MFJ-784B).  When augmented with a good analog "roofing" filter with a somewhat wider bandwidth than the DSP, you have an unbeatable combination for contesting.

Quote
I check my logs for WB2WIK and NI0C in the past couple years and I don't see either of you guys in there. I suspect my lax tolerances and non-insistence that the other guy zero-beat me, excludes us from QSO'ing.

Tim, that's just a function of our operating times, bands, and preferences.  I spend most of my time listening for rare DX on 160m these days.  (During the winter season, that's the only transmitting antenna I have deployed).  I do have a QSL card from Steve, WB2WIK, for a 17m SSB contact in 2004.  If I hear you CQ'ing in a contest, I'll give you a call. 

73,
Chuck  NI0C



   
Logged
DJ1YFK
Member

Posts: 188


WWW

Ignore
« Reply #16 on: November 19, 2010, 12:42:25 PM »

Have any studies been done on the optimum sidetone frequency for CW reception?

Peter, SM7CMY wrote his PhD thesis (title: "Signal Detection in Noise, with special reference to telegraphy") on this and related topics. I'm fortunate enough to have a copy of this thesis.

In  a nutshell, his experiments (with a number of test persons) concluded that the copy rate at a constant SNR increased as the frequency was decreased until about 500Hz. Below that, only a few test subjects showed further enhancement. The difference between the two extremes (2kHz vs. 500Hz) was somewhere around 3dB.

This matches very well with the purely theoretic conclusions from psycho-acoustics. A while ago (and before knowing of Peter's work), I produced a set of sound files to illustrate the effect of different pitches but most importantly of different filter bandwidths: http://fkurz.net/ham/stuff.html#noise

It pretty clearly shows that narrow filters are not of any real use, unless there's actually an interferer in the passband of the filter.

The thesis of Peter also covers other interesting topics, such as monaural vs. binaural copy and the effects of partial hearing loss on Morse abilities. One interesting conclusion (proven by experiments) is that copying signals (+ noise) which are 180° out of phase between both ears works significantly worse than in-phase signals. If only the Morse signal (but not the noise) is 180° out of phase, however, copy rates increase significantly.

By the way, although my radio is set to a pitch of 600Hz, my preferences are quite different: For RufzXP (120+wpm) I prefer a pitch of 800Hz (+/- the usual variation) and I experimented with even higher pitches with very good results, while I prefer 500Hz or 550Hz for QRQ ebooks with ebook2cw.

73
Fabian
Logged

WB2WIK
Member

Posts: 20595




Ignore
« Reply #17 on: November 19, 2010, 02:00:35 PM »



It pretty clearly shows that narrow filters are not of any real use, unless there's actually an interferer in the passband of the filter.

It didn't show that clearly to me at all.  The SNR is obviously greatly enhanced, as it should be, at the narrower bandwidths; this has nothing to do with interference and everything to do with spectral noise density.  I can always copy a weak signal at any pitch on CW far better with a 100 Hz filter than with a 2000 Hz one; as far as I know, so can anybody.  The only time it wouldn't matter is with a strong signal, so I can set AGC to "slow" and never hear any noise at all.  But with a weak signal and fast AGC, it's much easier to copy in a narrower bandwidth -- no?
Logged
DJ1YFK
Member

Posts: 188


WWW

Ignore
« Reply #18 on: November 19, 2010, 03:06:57 PM »

As far as the theory goes (e.g. in Peter's PhD thesis), and it is in good agreement with my own experience, due to the frequency bands in which the human ear is functioning (critical bands, http://en.wikipedia.org/wiki/Critical_bands), it doesn't help to use an IF filter with a bandwidth narrower than the critical bandwidth of the ear. At low frequencies (below 600Hz, http://en.wikipedia.org/wiki/Bark_scale) the critical bandwidth is around 100Hz, so any filter with a bandwidth of >100Hz doesn't help much, except in the case of in-band interference.

The sound files linked above should illustrate the effect: The (calculated) SNR significantly improves with reduced bandwidth, but the subjective impression hardly changes. The human ear/brain combination is doing a great job at filtering.

A little unrelated, but here's a nice weak signal challenge that I have posted in numerous forums:
http://dl0tud.tu-dresden.de/~dj1yfk/signal.wav

The only two guys who were able to name the correct callsign sent in this sound sample were a) an accomplished 10GHz DXer and b) a world-famous 160m DXer. If anyone can decode the callsign, let me know directly: mail@fkurz.net

73
Fabian DJ1YFK
Logged

N2EY
Member

Posts: 3879




Ignore
« Reply #19 on: November 19, 2010, 05:40:13 PM »

Quote
I've been studying it for 43 years. 900 Hz is too high, 700 to 750 Hz is about right.

What constitutes "studying it" ? 

Using Morse Code and trying out different pitches. Also observing what others do and how well it works for them.


I think you're just giving us your personal preference.  The optimum choice depends on the individual and what they are trying to accomplish. 

Right on both counts!

My main point is that each operator has to determine what works best for them. Usually it's a trial-and-error process. Some rigs are better suited for the process than others.

----

Many of the questions in these forums seem to me to ask "what's the best/quickest/standard way to do X". And often X represents something where the best/quickest varies enormously with the individual, and can only be found by direct experience.

IMHO it's better to be trying different things and finding what works for you.

For example, I've long been a fan of using a good sharp (narrow) filter on CW. Of course not all sharp filters are good; some will ring or have terrible shape factor, or have so much loss they're a liability.

But good filters do exist, both in hardware and software; one just has to find them. Not only do they knock down the QRM and boost the S/N, sharp filters make zeroing easy: If you have the other guy tuned to the center of your passband, and your rig is set up correctly, you'll be dead-on his QRG without much fuss. 

73 de Jim, N2EY





[/quote]
Logged
N3QE
Member

Posts: 2202




Ignore
« Reply #20 on: November 21, 2010, 05:13:48 AM »

I can always copy a weak signal at any pitch on CW far better with a 100 Hz filter than with a 2000 Hz one; as far as I know, so can anybody.  The only time it wouldn't matter is with a strong signal, so I can set AGC to "slow" and never hear any noise at all.  But with a weak signal and fast AGC, it's much easier to copy in a narrower bandwidth -- no?

See, here's where I differ with you fundamentally.

In my opinion, especially on say 80M or 160M, setting the bandwidth narrower, all it does is make the noise sound more like the signal.

If there's no interfering carriers, I will always choose a wide bandwidth for receive. Especially when there are static crash type QRN going around on 80M or 160M it is way more pleasant to use the wider bandwidth.

My brain is a much more powerful signal processor than anything I've ever bought.

IMHO: All a super narrow filter does is make everything coming out of the headphones sound like the same frequency, whether it be noise or signal.

And you mention AGC? I hate AGC. AGC exists to make everything the same loudness. Same principle as a narrow filter, I guess.

Tim N3QE
Logged
WB2WIK
Member

Posts: 20595




Ignore
« Reply #21 on: November 21, 2010, 11:04:28 AM »


IMHO: All a super narrow filter does is make everything coming out of the headphones sound like the same frequency, whether it be noise or signal.


Not with good I.F. DSP filtering, it doesn't.  I can listen in a 150 Hz BW and set the "pitch" to be anything I want it to be.

Meanwhile, going from 2.1 kHz to 150 Hz IF BW I can watch the noise level go from S7 to S3 or S4 (or from S3-4 to S0) without losing 1 dB of sensitivity.  The resulting S/N enhancement is enormous.

Need the right kind of receiver.
Logged
NI0C
Member

Posts: 2403




Ignore
« Reply #22 on: November 21, 2010, 04:11:30 PM »

N2EY wrote:
Quote
Many of the questions in these forums seem to me to ask "what's the best/quickest/standard way to do X". And often X represents something where the best/quickest varies enormously with the individual, and can only be found by direct experience.

That's what I was getting at in my earlier post to you, Jim.  I though were prescribing a standard pitch for everyone; I was wrong about that, and am glad you clarified.

73,
Chuck  NI0C
Logged
NI0C
Member

Posts: 2403




Ignore
« Reply #23 on: November 21, 2010, 04:29:54 PM »

DJ1YFK:

Many thanks, Fabian, for the useful and provocative information that you have shared here and also on your website.  I've listened to your recordings several times, and intend do check them again.  I still prefer the narrowest bandwidth portions of the recordings, especially on the signals you have as "weak," and "very weak."

Also with regard to your statement (quoted from one of your postings above:
Quote
it doesn't help to use an IF filter with a bandwidth narrower than the critical bandwidth of the ear. At low frequencies (below 600Hz, http://en.wikipedia.org/wiki/Bark_scale) the critical bandwidth is around 100Hz, so any filter with a bandwidth of >100Hz doesn't help much, except in the case of in-band interference.

This statement seems to contradict what you said on your website, namely:
"That also explains, why it hardly matters whether you listen to a weak signal in a 100Hz filter or a 2kHz wide filter (with, mathematically, very different SNRs, as can be seen in the table). Your ear/brain does all the work. The filter bandwidth only matters in the presence of other signals, QRM etc. "

Like WB2WIK and N2EY, I prefer narrow filters (100-300 Hz) for copying very weak signals.  In fact, I'm having good results on 160 meters using Elecraft's latest APF (audio peaking filter) that is currently being beta tested in a firmware upgrade for the K3.  I measured the frequency response of the APF this morning: 6 dB BW = 29 Hz, and 1 dB BW = 8 Hz.  For me (and many others) this filter pulls very weak signals out of the mud and makes them readable.   

73,
Chuck  NI0C
Logged
WB2WIK
Member

Posts: 20595




Ignore
« Reply #24 on: November 21, 2010, 06:47:11 PM »

DJ1YFK:

Many thanks, Fabian, for the useful and provocative information that you have shared here and also on your website.  I've listened to your recordings several times, and intend do check them again.  I still prefer the narrowest bandwidth portions of the recordings, especially on the signals you have as "weak," and "very weak."

Also with regard to your statement (quoted from one of your postings above:
Quote
it doesn't help to use an IF filter with a bandwidth narrower than the critical bandwidth of the ear. At low frequencies (below 600Hz, http://en.wikipedia.org/wiki/Bark_scale) the critical bandwidth is around 100Hz, so any filter with a bandwidth of >100Hz doesn't help much, except in the case of in-band interference.

This statement seems to contradict what you said on your website, namely:
"That also explains, why it hardly matters whether you listen to a weak signal in a 100Hz filter or a 2kHz wide filter (with, mathematically, very different SNRs, as can be seen in the table). Your ear/brain does all the work. The filter bandwidth only matters in the presence of other signals, QRM etc. "

Like WB2WIK and N2EY, I prefer narrow filters (100-300 Hz) for copying very weak signals.  In fact, I'm having good results on 160 meters using Elecraft's latest APF (audio peaking filter) that is currently being beta tested in a firmware upgrade for the K3.  I measured the frequency response of the APF this morning: 6 dB BW = 29 Hz, and 1 dB BW = 8 Hz.  For me (and many others) this filter pulls very weak signals out of the mud and makes them readable.   

73,
Chuck  NI0C

I couldn't agree more.  Of course it does, and it should.  I find the same thing.
Logged
DJ1YFK
Member

Posts: 188


WWW

Ignore
« Reply #25 on: November 21, 2010, 10:38:10 PM »

Also with regard to your statement (quoted from one of your postings above:
Quote
it doesn't help to use an IF filter with a bandwidth narrower than the critical bandwidth of the ear. At low frequencies (below 600Hz, http://en.wikipedia.org/wiki/Bark_scale) the critical bandwidth is around 100Hz, so any filter with a bandwidth of >100Hz doesn't help much, except in the case of in-band interference.
Oops. It should be "wider", not "narrower" in the first sentence, then things should make sense again.

I guess in the end the optimum performance can be achieved with whatever filter one prefers and the right amount of practice to get acquainted with its sound and characteristics.

73
Logged

K7MH
Member

Posts: 339




Ignore
« Reply #26 on: November 21, 2010, 11:25:15 PM »

Quote
so any filter with a bandwidth of >100Hz doesn't help much, except in the case of in-band interference.

Just to get it clarified, did you error also with the > symbol for greater than?

I don't go narrower than 100 Hz but it can make a huge difference. It can bring up a CW signal to a copyable signal above the noise from only knowing that a signal is there but not being able to tell what is going on with it, who he is working at the moment etc. when at a wider passband of even 500 Hz sometimes.
I went to 100 Hz early yesterday (5AM local) on 80 CW which brought up ZL8X enough that I could work them. 500 Hz wasn't narrow enough to make it all that copyable.
I have experienced that on 80 CW working several dxpeditions and other DX as well the last few years. For me it is a brutal band to work stuff on!
The DSP and variable filters are the greatest thing since peanut butter.
Logged
DJ1YFK
Member

Posts: 188


WWW

Ignore
« Reply #27 on: November 22, 2010, 12:10:01 AM »

Quote
so any filter with a bandwidth of >100Hz doesn't help much, except in the case of in-band interference.

Just to get it clarified, did you error also with the > symbol for greater than?

For me the sentence translates to "... any filter with a bandwidth of greater than 100Hz doesn't help much, ...", which is what I intended to write.

Again, to avoid more confusion, the conclusions from the cited work were:

Filters with less than 100Hz bandwidth do improve the SNR within the critical band.
Filters with more than 100Hz bandwidth do not improve the SNR within the critical band.

And as I said, this is only true for signals that are only masked by pure noise (i.e. band-filtered white noise). In the presence of interferers, QRN etc., it will not hold true.

Edit: The value of "100Hz" is only valid for the lower critical bands of hearing (below 500Hz); they are getting wider towards higher frequencies. Another thing that is mentioned in the study is the effect of the level at which a signal (and the noise) is presented; it can make a few dBs difference. Finally, the effect of hearing impairments need to be considered, as they degrade the sharpness of the auditory filters.

73
« Last Edit: November 22, 2010, 12:41:26 AM by Fabian Kurz » Logged

N2EY
Member

Posts: 3879




Ignore
« Reply #28 on: November 22, 2010, 03:24:39 AM »

N2EY wrote:
Quote
Many of the questions in these forums seem to me to ask "what's the best/quickest/standard way to do X". And often X represents something where the best/quickest varies enormously with the individual, and can only be found by direct experience.

That's what I was getting at in my earlier post to you, Jim.  I though[t you]  were prescribing a standard pitch for everyone;

Sorry I wasn't clear, Chuck. MY BAD!

I think the larger point needs to be emphasized: For some things, "the best" or "what works" is different for different people, and can only be found by trying.

All too often I see questions here and elsewhere online asking "what's the best HF rig for a newcomer?" or "what's the best way to learn Morse Code" or "what's the best simple antenna for all HF bands" and similar things. Usually there's little or no information about what the person likes to do or wants to do, what's already been tried, respources available, etc.

I can tell you what worked for me but that doesn't mean it will always work for you.

With sidetone frequency and the narrowness of filters, a lot depends on the situation. A few days ago I resurrected a one-tube regenerative receiver I had built back in the 1990s and listened to 80 CW with it. Amazingly clear and quiet reception, heard some good DX and had some fun. Really a pretty decent little rx; will probably make some good QSOs with it in the future. But using it in a DX contest would be a very different thing!

73 de Jim, N2EY
Logged
W0NHH
Member

Posts: 11




Ignore
« Reply #29 on: November 23, 2010, 05:58:04 PM »

My ears are pretty well shot but my eyes are ok so the tuning offset indicator on my FT950 puts me right down the slot.  If I depended on my ears I'd be lucky to stay in the right band. 
Logged
Pages: Prev 1 [2] 3 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!