Loading ...
Sorry, an error occurred while loading the content.

WSJT Tutorial & Phantom signal

Expand Messages
  • Dave Miller
    Has anyone ever tried to decode the hidden signal in the DL7UAE file under the JT65B folder of the WSJT samples download. The WSJT6 User manual s tutorial
    Message 1 of 13 , Aug 26, 2008
    • 0 Attachment
      Has anyone ever tried to decode the "hidden" signal in the DL7UAE file
      under the JT65B folder of the WSJT samples download. The WSJT6 User
      manual's tutorial (page 9) says that you should be able to decode a
      second station (SP6GWB) just 22 hz away from the primary signal
      (DL7UAE). I have followed the tutorial step by step dozens of times
      with Versions 4.9, 5.9, and the 7.0 beta and all yield the same
      results; no SP6GWB signal, only the DL7UAE signal is decoded. When I
      follow the directions I do get a header line with a * or a #
      terminator, but nothing to the right but blank space. Has anyone else
      successfully decoded this second signal? I am trying to learn the ins
      and outs of this program before I get on the air and mess up a QSO.

      Dave N9SS EN50
    • Dave Miller
      Additional information on my previous post on the JT65 tutorial: I was able to decode the second signal by playing around with my rate in value by setting it
      Message 2 of 13 , Aug 26, 2008
      • 0 Attachment
        Additional information on my previous post on the JT65 tutorial:

        I was able to decode the second signal by playing around with my "rate
        in" value by setting it to a value near what my "rate out" reading on
        the WSJT display was indicating. Yes, that is opposite what is
        supposed to be entered according to the instructions.

        Why I would need to set this is beyond me since the program is just
        opening a file directly and I wouldn't think it would need the sound
        card for this mode. When I adjusted the "rate in" to about 1.005 the
        second signal from SP6GWB would decode, but not the first! When I
        changed the "rate in" back to 1.0, then I could decode the first
        signal from DL7UAE, but cud no longer decode the second signal???
        Please note: the "rate in" reading from the first window on the main
        display was 0.978 which if entered into the "rate in" under settings
        would cause the decoder to stop functioning all together! What this
        all tells me is you better have a darn good sound card for WSJT or you
        will be struggling with poor decodes and other erratic decoding
        behavior.

        Fortunately, my ham station computer has a good sound card that
        displays 1.000 and 1.000 for the rate in and rate out, but the
        computer that I was learning WSJT on has a motherboard type sound card
        that is evidently not suitable for even running a tutorial with .wav
        files off the hard drive!



        --- In wsjtgroup@yahoogroups.com, "Dave Miller" <n9ssham@...> wrote:
        >
        > Has anyone ever tried to decode the "hidden" signal in the DL7UAE file
        > under the JT65B folder of the WSJT samples download. The WSJT6 User
        > manual's tutorial (page 9) says that you should be able to decode a
        > second station (SP6GWB) just 22 hz away from the primary signal
        > (DL7UAE). I have followed the tutorial step by step dozens of times
        > with Versions 4.9, 5.9, and the 7.0 beta and all yield the same
        > results; no SP6GWB signal, only the DL7UAE signal is decoded. When I
        > follow the directions I do get a header line with a * or a #
        > terminator, but nothing to the right but blank space. Has anyone else
        > successfully decoded this second signal? I am trying to learn the ins
        > and outs of this program before I get on the air and mess up a QSO.
        >
        > Dave N9SS EN50
        >
      • Dave Miller
        At the risk of being a general nuisance I will post a reply to my own post once more..... I tried the same tests on my Ham Shack computer and had the same
        Message 3 of 13 , Aug 27, 2008
        • 0 Attachment
          At the risk of being a general nuisance I will post a reply to my own
          post once more..... I tried the same tests on my Ham Shack computer
          and had the same results. I could only decode the second signal
          (SP6GWB) when the rate in was changed from 1.0 to 1.003. However, at
          this setting I could no longer decode the first signal DL7UAE. On this
          second computer, my sound card was reading .99999999 or 1.0000000 on
          both rate in and rate out, so this evidently is not a sound card issue.

          I had to restore rate in to 1.000 in order to decode DL7AUE. However,
          at this setting I can not decode the SP6GWB signal. Has anyone else
          experienced this behavior, having to play around with the rate in
          setting to get a proper JT65 decode on a file? Perhaps the original
          file was recorded with a clock rate error and requires this offset to
          decode properly, but why then would I need two separate offsets to
          decode the same wave file? The file name I am working with is:
          DL7UAE_040308_002400.wav 571kb dated 1/4/2005 from the
          WSJT_Samples.exe installer download.

          de N9SS, Dave



          --- In wsjtgroup@yahoogroups.com, "Dave Miller" <n9ssham@...> wrote:
          >
          > Additional information on my previous post on the JT65 tutorial:
          >
          > I was able to decode the second signal by playing around with my "rate
          > in" value by setting it to a value near what my "rate out" reading on
          > the WSJT display was indicating. Yes, that is opposite what is
          > supposed to be entered according to the instructions.
          >
          > Why I would need to set this is beyond me since the program is just
          > opening a file directly and I wouldn't think it would need the sound
          > card for this mode. When I adjusted the "rate in" to about 1.005 the
          > second signal from SP6GWB would decode, but not the first! When I
          > changed the "rate in" back to 1.0, then I could decode the first
          > signal from DL7UAE, but cud no longer decode the second signal???
          > Please note: the "rate in" reading from the first window on the main
          > display was 0.978 which if entered into the "rate in" under settings
          > would cause the decoder to stop functioning all together! What this
          > all tells me is you better have a darn good sound card for WSJT or you
          > will be struggling with poor decodes and other erratic decoding
          > behavior.
          >
          > Fortunately, my ham station computer has a good sound card that
          > displays 1.000 and 1.000 for the rate in and rate out, but the
          > computer that I was learning WSJT on has a motherboard type sound card
          > that is evidently not suitable for even running a tutorial with .wav
          > files off the hard drive!
          >
          >
          >
          > --- In wsjtgroup@yahoogroups.com, "Dave Miller" <n9ssham@> wrote:
          > >
          > > Has anyone ever tried to decode the "hidden" signal in the DL7UAE file
          > > under the JT65B folder of the WSJT samples download. The WSJT6 User
          > > manual's tutorial (page 9) says that you should be able to decode a
          > > second station (SP6GWB) just 22 hz away from the primary signal
          > > (DL7UAE). I have followed the tutorial step by step dozens of times
          > > with Versions 4.9, 5.9, and the 7.0 beta and all yield the same
          > > results; no SP6GWB signal, only the DL7UAE signal is decoded. When I
          > > follow the directions I do get a header line with a * or a #
          > > terminator, but nothing to the right but blank space. Has anyone else
          > > successfully decoded this second signal? I am trying to learn the ins
          > > and outs of this program before I get on the air and mess up a QSO.
          > >
          > > Dave N9SS EN50
          > >
          >
        • Chet, N8RA
          Hi Dave, Purely guessing here: the wav file is a recording of the off the air audio. So if one of the signals present had not generated his modulation with his
          Message 4 of 13 , Aug 27, 2008
          • 0 Attachment
            Hi Dave,
             
            Purely guessing here: the wav file is a recording of the off the air audio. So if one of the signals present had not generated his modulation with his sound card being perfect, then it would be "off" in its interrelated frequencies , and you would have to set your receiving decoder to be "off" similarly to decode correctly. The modulation being "off" would not be the same as a station with the right modulation being off slightly in transmit freq.
             
            Also FWIW, my 2 hamshack computers have inexpensive mainboards with built in audio, and they are both right on, no adjustment needed.
             
            73
            Chet


            From: wsjtgroup@yahoogroups.com [mailto:wsjtgroup@yahoogroups.com] On Behalf Of Dave Miller
            Sent: Wednesday, August 27, 2008 6:31 PM
            To: wsjtgroup@yahoogroups.com
            Subject: [wsjtgroup] Re: WSJT Tutorial & Phantom signal

            At the risk of being a general nuisance I will post a reply to my own
            post once more..... I tried the same tests on my Ham Shack computer
            and had the same results. I could only decode the second signal
            (SP6GWB) when the rate in was changed from 1.0 to 1.003. However, at
            this setting I could no longer decode the first signal DL7UAE. On this
            second computer, my sound card was reading .99999999 or 1.0000000 on
            both rate in and rate out, so this evidently is not a sound card issue.

            I had to restore rate in to 1.000 in order to decode DL7AUE. However,
            at this setting I can not decode the SP6GWB signal. Has anyone else
            experienced this behavior, having to play around with the rate in
            setting to get a proper JT65 decode on a file? Perhaps the original
            file was recorded with a clock rate error and requires this offset to
            decode properly, but why then would I need two separate offsets to
            decode the same wave file? The file name I am working with is:
            DL7UAE_040308_ 002400.wav 571kb dated 1/4/2005 from the
            WSJT_Samples. exe installer download.

            de N9SS, Dave

            --- In wsjtgroup@yahoogrou ps.com, "Dave Miller" <n9ssham@... > wrote:

            >
            > Additional
            information on my previous post on the JT65 tutorial:
            >
            > I was
            able to decode the second signal by playing around with my "rate
            > in"
            value by setting it to a value near what my "rate out" reading on
            > the
            WSJT display was indicating. Yes, that is opposite what is
            > supposed to
            be entered according to the instructions.
            >
            > Why I would need to
            set this is beyond me since the program is just
            > opening a file directly
            and I wouldn't think it would need the sound
            > card for this mode. When I
            adjusted the "rate in" to about 1.005 the
            > second signal from SP6GWB
            would decode, but not the first! When I
            > changed the "rate in" back to
            1.0, then I could decode the first
            > signal from DL7UAE, but cud no longer
            decode the second signal???
            > Please note: the "rate in" reading from the
            first window on the main
            > display was 0.978 which if entered into the
            "rate in" under settings
            > would cause the decoder to stop functioning all
            together! What this
            > all tells me is you better have a darn good sound
            card for WSJT or you
            > will be struggling with poor decodes and other
            erratic decoding
            > behavior.
            >
            > Fortunately, my ham station
            computer has a good sound card that
            > displays 1.000 and 1.000 for the
            rate in and rate out, but the
            > computer that I was learning WSJT on has a
            motherboard type sound card
            > that is evidently not suitable for even
            running a tutorial with .wav
            > files off the hard drive!
            >
            >
            >
            > --- In
            href="mailto:wsjtgroup%40yahoogroups.com">wsjtgroup@yahoogrou ps.com, "Dave Miller" <n9ssham@> wrote:
            > >
            > > Has anyone ever
            tried to decode the "hidden" signal in the DL7UAE file
            > > under the
            JT65B folder of the WSJT samples download. The WSJT6 User
            > > manual's
            tutorial (page 9) says that you should be able to decode a
            > > second
            station (SP6GWB) just 22 hz away from the primary signal
            > > (DL7UAE).
            I have followed the tutorial step by step dozens of times
            > > with
            Versions 4.9, 5.9, and the 7.0 beta and all yield the same
            > > results;
            no SP6GWB signal, only the DL7UAE signal is decoded. When I
            > > follow
            the directions I do get a header line with a * or a #
            > > terminator,
            but nothing to the right but blank space. Has anyone else
            > >
            successfully decoded this second signal? I am trying to learn the ins
            > > and outs of this program before I get on the air and mess up a QSO.
            > >
            > > Dave N9SS EN50
            > >
            >

          • Dave Miller
            Thanks Chet! That is a perfectly logical explanation and it is also quite interesting because when I made such minor changes (.003) in the rate in setting
            Message 5 of 13 , Aug 27, 2008
            • 0 Attachment
              Thanks Chet! That is a perfectly logical explanation and it is also
              quite interesting because when I made such minor changes (.003) in the
              "rate in" setting it makes the difference between perfect decoding and
              no decoding! Not only that, but the entire sync display area looked
              totally different with very small adjustments to "rate in". What this
              tells me is that if the station that I am running with is not set
              perfectly I may not be able to decode him or even see his sync tone
              properly. This is definitely food for thought as I prepare for my
              first eme skeds.

              de N9SS, Dave

              --- In wsjtgroup@yahoogroups.com, "Chet, N8RA" <chetsubaccount@...> wrote:
              >
              > Hi Dave,
              >
              > Purely guessing here: the wav file is a recording of the off the air
              audio.
              > So if one of the signals present had not generated his modulation
              with his
              > sound card being perfect, then it would be "off" in its interrelated
              > frequencies , and you would have to set your receiving decoder to be
              "off"
              > similarly to decode correctly. The modulation being "off" would not
              be the
              > same as a station with the right modulation being off slightly in
              transmit
              > freq.
              >
              > Also FWIW, my 2 hamshack computers have inexpensive mainboards with
              built in
              > audio, and they are both right on, no adjustment needed.
              >
              > 73
              > Chet
              >
              > _____
              >
              > From: wsjtgroup@yahoogroups.com [mailto:wsjtgroup@yahoogroups.com]
              On Behalf
              > Of Dave Miller
              > Sent: Wednesday, August 27, 2008 6:31 PM
              > To: wsjtgroup@yahoogroups.com
              > Subject: [wsjtgroup] Re: WSJT Tutorial & Phantom signal
              >
              >
              >
              > At the risk of being a general nuisance I will post a reply to my own
              > post once more..... I tried the same tests on my Ham Shack computer
              > and had the same results. I could only decode the second signal
              > (SP6GWB) when the rate in was changed from 1.0 to 1.003. However, at
              > this setting I could no longer decode the first signal DL7UAE. On this
              > second computer, my sound card was reading .99999999 or 1.0000000 on
              > both rate in and rate out, so this evidently is not a sound card issue.
              >
              > I had to restore rate in to 1.000 in order to decode DL7AUE. However,
              > at this setting I can not decode the SP6GWB signal. Has anyone else
              > experienced this behavior, having to play around with the rate in
              > setting to get a proper JT65 decode on a file? Perhaps the original
              > file was recorded with a clock rate error and requires this offset to
              > decode properly, but why then would I need two separate offsets to
              > decode the same wave file? The file name I am working with is:
              > DL7UAE_040308_002400.wav 571kb dated 1/4/2005 from the
              > WSJT_Samples.exe installer download.
              >
              > de N9SS, Dave
              >
              > --- In wsjtgroup@yahoogrou <mailto:wsjtgroup%40yahoogroups.com> ps.com,
              > "Dave Miller" <n9ssham@> wrote:
              > >
              > > Additional information on my previous post on the JT65 tutorial:
              > >
              > > I was able to decode the second signal by playing around with my "rate
              > > in" value by setting it to a value near what my "rate out" reading on
              > > the WSJT display was indicating. Yes, that is opposite what is
              > > supposed to be entered according to the instructions.
              > >
              > > Why I would need to set this is beyond me since the program is just
              > > opening a file directly and I wouldn't think it would need the sound
              > > card for this mode. When I adjusted the "rate in" to about 1.005 the
              > > second signal from SP6GWB would decode, but not the first! When I
              > > changed the "rate in" back to 1.0, then I could decode the first
              > > signal from DL7UAE, but cud no longer decode the second signal???
              > > Please note: the "rate in" reading from the first window on the main
              > > display was 0.978 which if entered into the "rate in" under settings
              > > would cause the decoder to stop functioning all together! What this
              > > all tells me is you better have a darn good sound card for WSJT or you
              > > will be struggling with poor decodes and other erratic decoding
              > > behavior.
              > >
              > > Fortunately, my ham station computer has a good sound card that
              > > displays 1.000 and 1.000 for the rate in and rate out, but the
              > > computer that I was learning WSJT on has a motherboard type sound card
              > > that is evidently not suitable for even running a tutorial with .wav
              > > files off the hard drive!
              > >
              > >
              > >
              > > --- In wsjtgroup@yahoogrou <mailto:wsjtgroup%40yahoogroups.com>
              ps.com,
              > "Dave Miller" <n9ssham@> wrote:
              > > >
              > > > Has anyone ever tried to decode the "hidden" signal in the
              DL7UAE file
              > > > under the JT65B folder of the WSJT samples download. The WSJT6 User
              > > > manual's tutorial (page 9) says that you should be able to decode a
              > > > second station (SP6GWB) just 22 hz away from the primary signal
              > > > (DL7UAE). I have followed the tutorial step by step dozens of times
              > > > with Versions 4.9, 5.9, and the 7.0 beta and all yield the same
              > > > results; no SP6GWB signal, only the DL7UAE signal is decoded. When I
              > > > follow the directions I do get a header line with a * or a #
              > > > terminator, but nothing to the right but blank space. Has anyone
              else
              > > > successfully decoded this second signal? I am trying to learn
              the ins
              > > > and outs of this program before I get on the air and mess up a QSO.
              > > >
              > > > Dave N9SS EN50
              > > >
              > >
              >
            • Chet, N8RA
              Dave, Based on your questions, I started into learning more about using JT6B. FWIW, following the tutorial prompts, I was able to decode SP6GWB without
              Message 6 of 13 , Aug 28, 2008
              • 0 Attachment
                Dave,
                 
                Based on your questions, I started into learning more about using JT6B.
                 
                FWIW, following the tutorial prompts, I was able to decode SP6GWB without touching my sound card frequencies. But I did have to use K1JT as the call in my info window, as mentioned in the tutorial since it takes the deep search decoder to do it. So maybe you have discovered something about those signals, or about your sound card. My WSJT 7 main screen is showing 1.0000 for the freqs with an occasional +/- 1 digit flicker in the last decimal place.
                 
                73
                Chet, N8RA


                From: wsjtgroup@yahoogroups.com [mailto:wsjtgroup@yahoogroups.com] On Behalf Of Dave Miller
                Sent: Wednesday, August 27, 2008 10:55 PM
                To: wsjtgroup@yahoogroups.com
                Subject: [wsjtgroup] Re: WSJT Tutorial & Phantom signal

                Thanks Chet! That is a perfectly logical explanation and it is also
                quite interesting because when I made such minor changes (.003) in the
                "rate in" setting it makes the difference between perfect decoding and
                no decoding! Not only that, but the entire sync display area looked
                totally different with very small adjustments to "rate in". What this
                tells me is that if the station that I am running with is not set
                perfectly I may not be able to decode him or even see his sync tone
                properly. This is definitely food for thought as I prepare for my
                first eme skeds.

                de N9SS, Dave

                --- In wsjtgroup@yahoogrou ps.com, "Chet, N8RA" <chetsubaccount@ ...> wrote:

                >
                > Hi
                Dave,
                >
                > Purely guessing here: the wav file is a recording of the
                off the air
                audio.
                > So if one of the signals present had not generated
                his modulation
                with his
                > sound card being perfect, then it would be
                "off" in its interrelated
                > frequencies , and you would have to set your
                receiving decoder to be
                "off"
                > similarly to decode correctly. The
                modulation being "off" would not
                be the
                > same as a station with the
                right modulation being off slightly in
                transmit
                > freq.
                >
                > Also FWIW, my 2 hamshack computers have inexpensive mainboards
                with
                built in
                > audio, and they are both right on, no adjustment
                needed.
                >
                > 73
                > Chet
                >
                > _____
                >
                > From:
                href="mailto:wsjtgroup%40yahoogroups.com">wsjtgroup@yahoogrou ps.com [mailto:wsjtgroup@yahoogrou ps.com]
                On Behalf
                > Of Dave Miller
                > Sent: Wednesday, August 27, 2008 6:31
                PM
                > To:
                href="mailto:wsjtgroup%40yahoogroups.com">wsjtgroup@yahoogrou ps.com
                >
                Subject: [wsjtgroup] Re: WSJT Tutorial & Phantom signal
                >
                >
                >
                > At the risk of being a general nuisance I will post a reply to
                my own
                > post once more..... I tried the same tests on my Ham Shack
                computer
                > and had the same results. I could only decode the second
                signal
                > (SP6GWB) when the rate in was changed from 1.0 to 1.003. However,
                at
                > this setting I could no longer decode the first signal DL7UAE. On
                this
                > second computer, my sound card was reading .99999999 or 1.0000000
                on
                > both rate in and rate out, so this evidently is not a sound card
                issue.
                >
                > I had to restore rate in to 1.000 in order to decode
                DL7AUE. However,
                > at this setting I can not decode the SP6GWB signal. Has
                anyone else
                > experienced this behavior, having to play around with the
                rate in
                > setting to get a proper JT65 decode on a file? Perhaps the
                original
                > file was recorded with a clock rate error and requires this
                offset to
                > decode properly, but why then would I need two separate
                offsets to
                > decode the same wave file? The file name I am working with
                is:
                > DL7UAE_040308_ 002400.wav 571kb dated 1/4/2005 from the
                >
                WSJT_Samples. exe installer download.
                >
                > de N9SS, Dave
                >
                > --- In wsjtgroup@yahoogrou
                <mailto:wsjtgroup% 40yahoogroups. com> ps.com,
                > "Dave
                Miller" <n9ssham@> wrote:
                > >
                > > Additional information
                on my previous post on the JT65 tutorial:
                > >
                > > I was able
                to decode the second signal by playing around with my "rate
                > > in"
                value by setting it to a value near what my "rate out" reading on
                > >
                the WSJT display was indicating. Yes, that is opposite what is
                > >
                supposed to be entered according to the instructions.
                > >
                > > Why I would need to set this is beyond me since the program is just
                > > opening a file directly and I wouldn't think it would need the
                sound
                > > card for this mode. When I adjusted the "rate in" to about
                1.005 the
                > > second signal from SP6GWB would decode, but not the
                first! When I
                > > changed the "rate in" back to 1.0, then I could
                decode the first
                > > signal from DL7UAE, but cud no longer decode the
                second signal???
                > > Please note: the "rate in" reading from the first
                window on the main
                > > display was 0.978 which if entered into the
                "rate in" under settings
                > > would cause the decoder to stop
                functioning all together! What this
                > > all tells me is you better have
                a darn good sound card for WSJT or you
                > > will be struggling with poor
                decodes and other erratic decoding
                > > behavior.
                > >
                > > Fortunately, my ham station computer has a good sound card that
                > > displays 1.000 and 1.000 for the rate in and rate out, but the
                > >
                computer that I was learning WSJT on has a motherboard type sound card
                > > that is evidently not suitable for even running a tutorial with
                .wav
                > > files off the hard drive!
                > >
                > >
                > >
                > > --- In wsjtgroup@yahoogrou
                <mailto:wsjtgroup% 40yahoogroups. com>
                ps.com,
                > "Dave
                Miller" <n9ssham@> wrote:
                > > >
                > > > Has anyone
                ever tried to decode the "hidden" signal in the
                DL7UAE file
                > > >
                under the JT65B folder of the WSJT samples download. The WSJT6 User
                > > > manual's tutorial (page 9) says that you should be able to decode a
                > > > second station (SP6GWB) just 22 hz away from the primary
                signal
                > > > (DL7UAE). I have followed the tutorial step by step
                dozens of times
                > > > with Versions 4.9, 5.9, and the 7.0 beta and
                all yield the same
                > > > results; no SP6GWB signal, only the DL7UAE
                signal is decoded. When I
                > > > follow the directions I do get a
                header line with a * or a #
                > > > terminator, but nothing to the
                right but blank space. Has anyone
                else
                > > > successfully decoded
                this second signal? I am trying to learn
                the ins
                > > > and outs
                of this program before I get on the air and mess up a QSO.
                > > >
                > > > Dave N9SS EN50
                > > >
                > >
                >

              • Russ K2TXB
                Gentlemen: Interesting discussion and results. The deep search algorithm uses a process similar to what you do when you copy a call partially, and you think
                Message 7 of 13 , Aug 28, 2008
                • 0 Attachment
                  Gentlemen:  Interesting discussion and results.  The deep search algorithm uses a process similar to what you do when you copy a call partially, and you think you might know that call.  You hear K2??B and think "I wonder if that is Russ, K2TXB?".  So you listen carefully to the middle two letters.  Now that you think you 'know' what they are, sure enough you copy the T and X weakly.  Now you are sure.  Conversely, you think the first letter is an A and not sure on the next letter.  So you think "No I don't think that is him", and so you consider all the other calls you know to see if any of them fit the pattern.  If one does, you repeat the process of trying to be sure the missing letter fits.
                   
                  Deep search works a lot like that.  It is astoundingly good at pulling weak, known, calls out of the noise.  (And a lot of people cannot get their heads around the fact that it works just like people do it in their heads, and try to say it is not sufficient to identify a call because it is known.  But that is another subject.)
                   
                  But when you discover that a slight tweak to the sound card frequency can decode that same signal without using deep search, that is something indeed!  I will be interested to hear what Joe has to say about that, but it seems to me that it could lead to another way of improving decodes with or without using deep search.
                   
                  73, Russ K2TXB


                  From: wsjtgroup@yahoogroups.com [mailto:wsjtgroup@yahoogroups.com] On Behalf Of Chet, N8RA
                  Sent: Thursday, August 28, 2008 8:40 AM
                  To: 'Dave Miller'; wsjtgroup@yahoogroups.com
                  Subject: RE: [wsjtgroup] Re: WSJT Tutorial & Phantom signal

                  Dave,
                   
                  Based on your questions, I started into learning more about using JT6B.
                   
                  FWIW, following the tutorial prompts, I was able to decode SP6GWB without touching my sound card frequencies. But I did have to use K1JT as the call in my info window, as mentioned in the tutorial since it takes the deep search decoder to do it. So maybe you have discovered something about those signals, or about your sound card. My WSJT 7 main screen is showing 1.0000 for the freqs with an occasional +/- 1 digit flicker in the last decimal place.
                   
                  73
                  Chet, N8RA


                  From: wsjtgroup@yahoogrou ps.com [mailto:wsjtgroup@ yahoogroups. com] On Behalf Of Dave Miller
                  Sent: Wednesday, August 27, 2008 10:55 PM
                  To: wsjtgroup@yahoogrou ps.com
                  Subject: [wsjtgroup] Re: WSJT Tutorial & Phantom signal

                  Thanks Chet! That is a perfectly logical explanation and it is also
                  quite interesting because when I made such minor changes (.003) in the
                  "rate in" setting it makes the difference between perfect decoding and
                  no decoding! Not only that, but the entire sync display area looked
                  totally different with very small adjustments to "rate in". What this
                  tells me is that if the station that I am running with is not set
                  perfectly I may not be able to decode him or even see his sync tone
                  properly. This is definitely food for thought as I prepare for my
                  first eme skeds.

                  de N9SS, Dave

                  --- In wsjtgroup@yahoogrou ps.com, "Chet, N8RA" <chetsubaccount@ ...> wrote:
                  >
                  > Hi Dave,
                  >
                  > Purely guessing here: the wav file is a recording of the off the air
                  audio.
                  > So if one of the signals present had not generated his modulation
                  with his
                  > sound card being perfect, then it would be "off" in its interrelated
                  > frequencies , and you would have to set your receiving decoder to be
                  "off"
                  > similarly to decode correctly. The modulation being "off" would not
                  be the
                  > same as a station with the right modulation being off slightly in
                  transmit
                  > freq.
                  >
                  > Also FWIW, my 2 hamshack computers have inexpensive mainboards with
                  built in
                  > audio, and they are both right on, no adjustment needed.
                  >
                  > 73
                  > Chet
                  >
                  > _____
                  >
                  > From: wsjtgroup@yahoogrou ps.com [mailto:wsjtgroup@yahoogrou ps.com]
                  On Behalf
                  > Of Dave Miller
                  > Sent: Wednesday, August 27, 2008 6:31 PM
                  > To: wsjtgroup@yahoogrou ps.com
                  > Subject: [wsjtgroup] Re: WSJT Tutorial & Phantom signal
                  >
                  >
                  >
                  > At the risk of being a general nuisance I will post a reply to my own
                  > post once more..... I tried the same tests on my Ham Shack computer
                  > and had the same results. I could only decode the second signal
                  > (SP6GWB) when the rate in was changed from 1.0 to 1.003. However, at
                  > this setting I could no longer decode the first signal DL7UAE. On this
                  > second computer, my sound card was reading .99999999 or 1.0000000 on
                  > both rate in and rate out, so this evidently is not a sound card issue.
                  >
                  > I had to restore rate in to 1.000 in order to decode DL7AUE. However,
                  > at this setting I can not decode the SP6GWB signal. Has anyone else
                  > experienced this behavior, having to play around with the rate in
                  > setting to get a proper JT65 decode on a file? Perhaps the original
                  > file was recorded with a clock rate error and requires this offset to
                  > decode properly, but why then would I need two separate offsets to
                  > decode the same wave file? The file name I am working with is:
                  > DL7UAE_040308_ 002400.wav 571kb dated 1/4/2005 from the
                  > WSJT_Samples. exe installer download.
                  >
                  > de N9SS, Dave
                  >
                  > --- In wsjtgroup@yahoogrou <mailto:wsjtgroup% 40yahoogroups. com> ps.com,
                  > "Dave Miller" <n9ssham@> wrote:
                  > >
                  > > Additional information on my previous post on the JT65 tutorial:
                  > >
                  > > I was able to decode the second signal by playing around with my "rate
                  > > in" value by setting it to a value near what my "rate out" reading on
                  > > the WSJT display was indicating. Yes, that is opposite what is
                  > > supposed to be entered according to the instructions.
                  > >
                  > > Why I would need to set this is beyond me since the program is just
                  > > opening a file directly and I wouldn't think it would need the sound
                  > > card for this mode. When I adjusted the "rate in" to about 1.005 the
                  > > second signal from SP6GWB would decode, but not the first! When I
                  > > changed the "rate in" back to 1.0, then I could decode the first
                  > > signal from DL7UAE, but cud no longer decode the second signal???
                  > > Please note: the "rate in" reading from the first window on the main
                  > > display was 0.978 which if entered into the "rate in" under settings
                  > > would cause the decoder to stop functioning all together! What this
                  > > all tells me is you better have a darn good sound card for WSJT or you
                  > > will be struggling with poor decodes and other erratic decoding
                  > > behavior.
                  > >
                  > > Fortunately, my ham station computer has a good sound card that
                  > > displays 1.000 and 1.000 for the rate in and rate out, but the
                  > > computer that I was learning WSJT on has a motherboard type sound card
                  > > that is evidently not suitable for even running a tutorial with .wav
                  > > files off the hard drive!
                  > >
                  > >
                  > >
                  > > --- In wsjtgroup@yahoogrou <mailto:wsjtgroup% 40yahoogroups. com>
                  ps.com,
                  > "Dave Miller" <n9ssham@> wrote:
                  > > >
                  > > > Has anyone ever tried to decode the "hidden" signal in the
                  DL7UAE file
                  > > > under the JT65B folder of the WSJT samples download. The WSJT6 User
                  > > > manual's tutorial (page 9) says that you should be able to decode a
                  > > > second station (SP6GWB) just 22 hz away from the primary signal
                  > > > (DL7UAE). I have followed the tutorial step by step dozens of times
                  > > > with Versions 4.9, 5.9, and the 7.0 beta and all yield the same
                  > > > results; no SP6GWB signal, only the DL7UAE signal is decoded. When I
                  > > > follow the directions I do get a header line with a * or a #
                  > > > terminator, but nothing to the right but blank space. Has anyone
                  else
                  > > > successfully decoded this second signal? I am trying to learn
                  the ins
                  > > > and outs of this program before I get on the air and mess up a QSO.
                  > > >
                  > > > Dave N9SS EN50
                  > > >
                  > >
                  >

                • Barry Garratt
                  I decode it just fine without changing any of my setup. The signal is 22 Hz away so setting TOL to 10 will allow it to be decoded just fine. The decoder will
                  Message 8 of 13 , Aug 28, 2008
                  • 0 Attachment
                    I decode it just fine without changing any of my setup. The signal is 22 Hz away so setting TOL to 10 will allow it to be decoded just fine.
                     
                    The decoder will always decode the strongest signal by default so any setting of TOL greater than 10 will only decode the DL7.
                     
                    73, Barry VE3CDX/W7


                    From: wsjtgroup@yahoogroups.com [mailto:wsjtgroup@yahoogroups.com] On Behalf Of Russ K2TXB
                    Sent: Thursday, August 28, 2008 6:49 AM
                    To: wsjtgroup@yahoogroups.com
                    Subject: RE: [wsjtgroup] Re: WSJT Tutorial & Phantom signal

                    Gentlemen:  Interesting discussion and results.  The deep search algorithm uses a process similar to what you do when you copy a call partially, and you think you might know that call.  You hear K2??B and think "I wonder if that is Russ, K2TXB?".  So you listen carefully to the middle two letters.  Now that you think you 'know' what they are, sure enough you copy the T and X weakly.  Now you are sure.  Conversely, you think the first letter is an A and not sure on the next letter.  So you think "No I don't think that is him", and so you consider all the other calls you know to see if any of them fit the pattern.  If one does, you repeat the process of trying to be sure the missing letter fits.
                     
                    Deep search works a lot like that.  It is astoundingly good at pulling weak, known, calls out of the noise.  (And a lot of people cannot get their heads around the fact that it works just like people do it in their heads, and try to say it is not sufficient to identify a call because it is known.  But that is another subject.)
                     
                    But when you discover that a slight tweak to the sound card frequency can decode that same signal without using deep search, that is something indeed!  I will be interested to hear what Joe has to say about that, but it seems to me that it could lead to another way of improving decodes with or without using deep search.
                     
                    73, Russ K2TXB


                    From: wsjtgroup@yahoogrou ps.com [mailto:wsjtgroup@ yahoogroups. com] On Behalf Of Chet, N8RA
                    Sent: Thursday, August 28, 2008 8:40 AM
                    To: 'Dave Miller'; wsjtgroup@yahoogrou ps.com
                    Subject: RE: [wsjtgroup] Re: WSJT Tutorial & Phantom signal

                    Dave,
                     
                    Based on your questions, I started into learning more about using JT6B.
                     
                    FWIW, following the tutorial prompts, I was able to decode SP6GWB without touching my sound card frequencies. But I did have to use K1JT as the call in my info window, as mentioned in the tutorial since it takes the deep search decoder to do it. So maybe you have discovered something about those signals, or about your sound card. My WSJT 7 main screen is showing 1.0000 for the freqs with an occasional +/- 1 digit flicker in the last decimal place.
                     
                    73
                    Chet, N8RA


                    From: wsjtgroup@yahoogrou ps.com [mailto:wsjtgroup@ yahoogroups. com] On Behalf Of Dave Miller
                    Sent: Wednesday, August 27, 2008 10:55 PM
                    To: wsjtgroup@yahoogrou ps.com
                    Subject: [wsjtgroup] Re: WSJT Tutorial & Phantom signal

                    Thanks Chet! That is a perfectly logical explanation and it is also
                    quite interesting because when I made such minor changes (.003) in the
                    "rate in" setting it makes the difference between perfect decoding and
                    no decoding! Not only that, but the entire sync display area looked
                    totally different with very small adjustments to "rate in". What this
                    tells me is that if the station that I am running with is not set
                    perfectly I may not be able to decode him or even see his sync tone
                    properly. This is definitely food for thought as I prepare for my
                    first eme skeds.

                    de N9SS, Dave

                    --- In wsjtgroup@yahoogrou ps.com, "Chet, N8RA" <chetsubaccount@ ...> wrote:
                    >
                    > Hi Dave,
                    >
                    > Purely guessing here: the wav file is a recording of the off the air
                    audio.
                    > So if one of the signals present had not generated his modulation
                    with his
                    > sound card being perfect, then it would be "off" in its interrelated
                    > frequencies , and you would have to set your receiving decoder to be
                    "off"
                    > similarly to decode correctly. The modulation being "off" would not
                    be the
                    > same as a station with the right modulation being off slightly in
                    transmit
                    > freq.
                    >
                    > Also FWIW, my 2 hamshack computers have inexpensive mainboards with
                    built in
                    > audio, and they are both right on, no adjustment needed.
                    >
                    > 73
                    > Chet
                    >
                    > _____
                    >
                    > From: wsjtgroup@yahoogrou ps.com [mailto:wsjtgroup@yahoogrou ps.com]
                    On Behalf
                    > Of Dave Miller
                    > Sent: Wednesday, August 27, 2008 6:31 PM
                    > To: wsjtgroup@yahoogrou ps.com
                    > Subject: [wsjtgroup] Re: WSJT Tutorial & Phantom signal
                    >
                    >
                    >
                    > At the risk of being a general nuisance I will post a reply to my own
                    > post once more..... I tried the same tests on my Ham Shack computer
                    > and had the same results. I could only decode the second signal
                    > (SP6GWB) when the rate in was changed from 1.0 to 1.003. However, at
                    > this setting I could no longer decode the first signal DL7UAE. On this
                    > second computer, my sound card was reading .99999999 or 1.0000000 on
                    > both rate in and rate out, so this evidently is not a sound card issue.
                    >
                    > I had to restore rate in to 1.000 in order to decode DL7AUE. However,
                    > at this setting I can not decode the SP6GWB signal. Has anyone else
                    > experienced this behavior, having to play around with the rate in
                    > setting to get a proper JT65 decode on a file? Perhaps the original
                    > file was recorded with a clock rate error and requires this offset to
                    > decode properly, but why then would I need two separate offsets to
                    > decode the same wave file? The file name I am working with is:
                    > DL7UAE_040308_ 002400.wav 571kb dated 1/4/2005 from the
                    > WSJT_Samples. exe installer download.
                    >
                    > de N9SS, Dave
                    >
                    > --- In wsjtgroup@yahoogrou <mailto:wsjtgroup% 40yahoogroups. com> ps.com,
                    > "Dave Miller" <n9ssham@> wrote:
                    > >
                    > > Additional information on my previous post on the JT65 tutorial:
                    > >
                    > > I was able to decode the second signal by playing around with my "rate
                    > > in" value by setting it to a value near what my "rate out" reading on
                    > > the WSJT display was indicating. Yes, that is opposite what is
                    > > supposed to be entered according to the instructions.
                    > >
                    > > Why I would need to set this is beyond me since the program is just
                    > > opening a file directly and I wouldn't think it would need the sound
                    > > card for this mode. When I adjusted the "rate in" to about 1.005 the
                    > > second signal from SP6GWB would decode, but not the first! When I
                    > > changed the "rate in" back to 1.0, then I could decode the first
                    > > signal from DL7UAE, but cud no longer decode the second signal???
                    > > Please note: the "rate in" reading from the first window on the main
                    > > display was 0.978 which if entered into the "rate in" under settings
                    > > would cause the decoder to stop functioning all together! What this
                    > > all tells me is you better have a darn good sound card for WSJT or you
                    > > will be struggling with poor decodes and other erratic decoding
                    > > behavior.
                    > >
                    > > Fortunately, my ham station computer has a good sound card that
                    > > displays 1.000 and 1.000 for the rate in and rate out, but the
                    > > computer that I was learning WSJT on has a motherboard type sound card
                    > > that is evidently not suitable for even running a tutorial with .wav
                    > > files off the hard drive!
                    > >
                    > >
                    > >
                    > > --- In wsjtgroup@yahoogrou <mailto:wsjtgroup% 40yahoogroups. com>
                    ps.com,
                    > "Dave Miller" <n9ssham@> wrote:
                    > > >
                    > > > Has anyone ever tried to decode the "hidden" signal in the
                    DL7UAE file
                    > > > under the JT65B folder of the WSJT samples download. The WSJT6 User
                    > > > manual's tutorial (page 9) says that you should be able to decode a
                    > > > second station (SP6GWB) just 22 hz away from the primary signal
                    > > > (DL7UAE). I have followed the tutorial step by step dozens of times
                    > > > with Versions 4.9, 5.9, and the 7.0 beta and all yield the same
                    > > > results; no SP6GWB signal, only the DL7UAE signal is decoded. When I
                    > > > follow the directions I do get a header line with a * or a #
                    > > > terminator, but nothing to the right but blank space. Has anyone
                    else
                    > > > successfully decoded this second signal? I am trying to learn
                    the ins
                    > > > and outs of this program before I get on the air and mess up a QSO.
                    > > >
                    > > > Dave N9SS EN50
                    > > >
                    > >
                    >

                  • Russ K2TXB
                    ... Hz ... Hi Barry. Yes, well I don t have the file as I have not downloaded the newer WSJT programs (as I am not currently on the air). Of course I have
                    Message 9 of 13 , Aug 28, 2008
                    • 0 Attachment
                      Barry, VE3CDX/W7 wrote:

                      > I decode it just fine without changing any of my setup. The signal is 22
                      Hz
                      > away so setting TOL to 10 will allow it to be decoded just fine.

                      Hi Barry. Yes, well I don't have the file as I have not downloaded the newer
                      WSJT programs (as I am not currently on the air). Of course I have often
                      used freeze and tolerance to copy weak signals close to strong ones.
                      Perhaps the gentlemen who were discussing changing the sound card frequency
                      settings to copy that signal were not aware that there was a more direct
                      way. They seemed knowledgeable to me, so I did not consider that they may
                      not have tried tolerance and freeze. If the signal can easily be decoded
                      without sound card frequency change, then my speculation about an improved
                      algorithm is null and void (programming terms), hi.

                      73, Russ K2TXB


                      --------------- original message ------------------------------
                      From: wsjtgroup@yahoogroups.com [mailto:wsjtgroup@yahoogroups.com] On Behalf
                      Of Russ K2TXB
                      Sent: Thursday, August 28, 2008 6:49 AM
                      To: wsjtgroup@yahoogroups.com
                      Subject: RE: [wsjtgroup] Re: WSJT Tutorial & Phantom signal


                      Gentlemen: Interesting discussion and results. The deep search algorithm
                      uses a process similar to what you do when you copy a call partially, and
                      you think you might know that call. You hear K2??B and think "I wonder if
                      that is Russ, K2TXB?". So you listen carefully to the middle two letters.
                      Now that you think you 'know' what they are, sure enough you copy the T and
                      X weakly. Now you are sure. Conversely, you think the first letter is an A
                      and not sure on the next letter. So you think "No I don't think that is
                      him", and so you consider all the other calls you know to see if any of them
                      fit the pattern. If one does, you repeat the process of trying to be sure
                      the missing letter fits.

                      Deep search works a lot like that. It is astoundingly good at pulling weak,
                      known, calls out of the noise. (And a lot of people cannot get their heads
                      around the fact that it works just like people do it in their heads, and try
                      to say it is not sufficient to identify a call because it is known. But
                      that is another subject.)

                      But when you discover that a slight tweak to the sound card frequency can
                      decode that same signal without using deep search, that is something indeed!
                      I will be interested to hear what Joe has to say about that, but it seems to
                      me that it could lead to another way of improving decodes with or without
                      using deep search.

                      73, Russ K2TXB
                    • Dave Miller
                      Ok thanks to all for all the great responses. You have all kept me very busy testing out various permutations and suggestions. Yes, I know that you have to
                      Message 10 of 13 , Aug 28, 2008
                      • 0 Attachment
                        Ok thanks to all for all the great responses. You have all kept me
                        very busy testing out various permutations and suggestions.

                        Yes, I know that you have to select the sync sig 22hz from the main
                        one, Have freeze selected and set TOL to 10hz. I have done this so
                        many times now that I can do it blind-folded.

                        What I missed was that you are supposed to run the tutorial with W1JT
                        in the My Call setup. When I did that AND turned on the Deep Search
                        option I could then successfully decode the second station without
                        changing my RATE IN to 1.005.

                        And yes, I can decode the second signal, WITHOUT DEEP SEARCH turned on
                        AND with MY CALL set to my actual call, IF I change RATE IN to
                        somewhere around 1.003 to 1.005 depending on the computer I am using.
                        I can ONLY copy the first (louder) DL7 signal at this RATE IN setting
                        IF and only IF I enter K1JT in the MY CALL setup AND turn on DEEP SEARCH.

                        Faced with the options here I think I will investigate a way to
                        extract the correct offset so I can tune the station in, so to speak,
                        when the station is not at the perfect clock rate. I much prefer that
                        to telling the WSJT decoder half of the exchange so that it can tell
                        me who is talking to me!

                        What still puzzles me is why some of you indicated that you were able
                        to copy the second SP6 signal WITHOUT changing any of your normal
                        settings. Does this include the My Call setting?

                        Thanks all again for your thorough responses!

                        de N9SS, Dave EN50

                        --- In wsjtgroup@yahoogroups.com, "Russ K2TXB" <k2txb@...> wrote:
                        >
                        > Barry, VE3CDX/W7 wrote:
                        >
                        > > I decode it just fine without changing any of my setup. The signal
                        is 22
                        > Hz
                        > > away so setting TOL to 10 will allow it to be decoded just fine.
                        >
                        > Hi Barry. Yes, well I don't have the file as I have not downloaded
                        the newer
                        > WSJT programs (as I am not currently on the air). Of course I have
                        often
                        > used freeze and tolerance to copy weak signals close to strong ones.
                        > Perhaps the gentlemen who were discussing changing the sound card
                        frequency
                        > settings to copy that signal were not aware that there was a more direct
                        > way. They seemed knowledgeable to me, so I did not consider that
                        they may
                        > not have tried tolerance and freeze. If the signal can easily be
                        decoded
                        > without sound card frequency change, then my speculation about an
                        improved
                        > algorithm is null and void (programming terms), hi.
                        >
                        > 73, Russ K2TXB
                        >
                        >
                        > --------------- original message ------------------------------
                        > From: wsjtgroup@yahoogroups.com [mailto:wsjtgroup@yahoogroups.com]
                        On Behalf
                        > Of Russ K2TXB
                        > Sent: Thursday, August 28, 2008 6:49 AM
                        > To: wsjtgroup@yahoogroups.com
                        > Subject: RE: [wsjtgroup] Re: WSJT Tutorial & Phantom signal
                        >
                        >
                        > Gentlemen: Interesting discussion and results. The deep search
                        algorithm
                        > uses a process similar to what you do when you copy a call
                        partially, and
                        > you think you might know that call. You hear K2??B and think "I
                        wonder if
                        > that is Russ, K2TXB?". So you listen carefully to the middle two
                        letters.
                        > Now that you think you 'know' what they are, sure enough you copy
                        the T and
                        > X weakly. Now you are sure. Conversely, you think the first letter
                        is an A
                        > and not sure on the next letter. So you think "No I don't think that is
                        > him", and so you consider all the other calls you know to see if any
                        of them
                        > fit the pattern. If one does, you repeat the process of trying to
                        be sure
                        > the missing letter fits.
                        >
                        > Deep search works a lot like that. It is astoundingly good at
                        pulling weak,
                        > known, calls out of the noise. (And a lot of people cannot get
                        their heads
                        > around the fact that it works just like people do it in their heads,
                        and try
                        > to say it is not sufficient to identify a call because it is known. But
                        > that is another subject.)
                        >
                        > But when you discover that a slight tweak to the sound card
                        frequency can
                        > decode that same signal without using deep search, that is something
                        indeed!
                        > I will be interested to hear what Joe has to say about that, but it
                        seems to
                        > me that it could lead to another way of improving decodes with or
                        without
                        > using deep search.
                        >
                        > 73, Russ K2TXB
                        >
                      • Chet, N8RA
                        Dave, You mentioned a combo that I don t remember trying when I went thru the tutorial hints at decoding the second signal. So I tried decoding the SP6 with my
                        Message 11 of 13 , Aug 29, 2008
                        • 0 Attachment
                          Dave,
                           
                          You mentioned a combo that I don't remember trying when I went thru the tutorial hints at decoding the second signal.
                          So I tried decoding the SP6 with my normal settings, specifically with my call left in the options screen. I
                           could not decode that signal either with or without the aggressive deep search selected.
                           
                          What you have found remains quite interesting.
                           
                          73
                          Chet


                          From: wsjtgroup@yahoogroups.com [mailto:wsjtgroup@yahoogroups.com] On Behalf Of Dave Miller
                          Sent: Thursday, August 28, 2008 9:09 PM
                          To: wsjtgroup@yahoogroups.com
                          Subject: [wsjtgroup] Re: WSJT Tutorial & Phantom signal

                          Ok thanks to all for all the great responses. You have all kept me
                          very busy testing out various permutations and suggestions.

                          Yes, I know that you have to select the sync sig 22hz from the main
                          one, Have freeze selected and set TOL to 10hz. I have done this so
                          many times now that I can do it blind-folded.

                          What I missed was that you are supposed to run the tutorial with W1JT
                          in the My Call setup. When I did that AND turned on the Deep Search
                          option I could then successfully decode the second station without
                          changing my RATE IN to 1.005.

                          And yes, I can decode the second signal, WITHOUT DEEP SEARCH turned on
                          AND with MY CALL set to my actual call, IF I change RATE IN to
                          somewhere around 1.003 to 1.005 depending on the computer I am using.
                          I can ONLY copy the first (louder) DL7 signal at this RATE IN setting
                          IF and only IF I enter K1JT in the MY CALL setup AND turn on DEEP SEARCH.

                          Faced with the options here I think I will investigate a way to
                          extract the correct offset so I can tune the station in, so to speak,
                          when the station is not at the perfect clock rate. I much prefer that
                          to telling the WSJT decoder half of the exchange so that it can tell
                          me who is talking to me!

                          What still puzzles me is why some of you indicated that you were able
                          to copy the second SP6 signal WITHOUT changing any of your normal
                          settings. Does this include the My Call setting?

                          Thanks all again for your thorough responses!

                          de N9SS, Dave EN50

                          --- In wsjtgroup@yahoogrou ps.com, "Russ K2TXB" <k2txb@...> wrote:

                          >
                          > Barry, VE3CDX/W7
                          wrote:
                          >
                          > > I decode it just fine without changing any of my
                          setup. The signal
                          is 22
                          > Hz
                          > > away so setting TOL to 10
                          will allow it to be decoded just fine.
                          >
                          > Hi Barry. Yes, well I
                          don't have the file as I have not downloaded
                          the newer
                          > WSJT programs
                          (as I am not currently on the air). Of course I have
                          often
                          > used
                          freeze and tolerance to copy weak signals close to strong ones.
                          > Perhaps
                          the gentlemen who were discussing changing the sound card
                          frequency
                          >
                          settings to copy that signal were not aware that there was a more direct
                          >
                          way. They seemed knowledgeable to me, so I did not consider that
                          they may
                          > not have tried tolerance and freeze. If the signal can easily
                          be
                          decoded
                          > without sound card frequency change, then my speculation
                          about an
                          improved
                          > algorithm is null and void (programming terms),
                          hi.
                          >
                          > 73, Russ K2TXB
                          >
                          >
                          >
                          ------------ --- original message ------------ --------- ---------
                          > From:
                          href="mailto:wsjtgroup%40yahoogroups.com">wsjtgroup@yahoogrou ps.com [mailto:wsjtgroup@yahoogrou ps.com]
                          On Behalf
                          > Of Russ K2TXB
                          > Sent: Thursday, August 28, 2008 6:49
                          AM
                          > To:
                          href="mailto:wsjtgroup%40yahoogroups.com">wsjtgroup@yahoogrou ps.com
                          >
                          Subject: RE: [wsjtgroup] Re: WSJT Tutorial & Phantom signal
                          >
                          >
                          > Gentlemen: Interesting discussion and results. The deep
                          search
                          algorithm
                          > uses a process similar to what you do when you copy
                          a call
                          partially, and
                          > you think you might know that call. You hear
                          K2??B and think "I
                          wonder if
                          > that is Russ, K2TXB?". So you listen
                          carefully to the middle two
                          letters.
                          > Now that you think you 'know'
                          what they are, sure enough you copy
                          the T and
                          > X weakly. Now you are
                          sure. Conversely, you think the first letter
                          is an A
                          > and not sure on
                          the next letter. So you think "No I don't think that is
                          > him", and so you
                          consider all the other calls you know to see if any
                          of them
                          > fit the
                          pattern. If one does, you repeat the process of trying to
                          be sure
                          > the
                          missing letter fits.
                          >
                          > Deep search works a lot like that. It is
                          astoundingly good at
                          pulling weak,
                          > known, calls out of the noise.
                          (And a lot of people cannot get
                          their heads
                          > around the fact that it
                          works just like people do it in their heads,
                          and try
                          > to say it is not
                          sufficient to identify a call because it is known. But
                          > that is another
                          subject.)
                          >
                          > But when you discover that a slight tweak to the
                          sound card
                          frequency can
                          > decode that same signal without using deep
                          search, that is something
                          indeed!
                          > I will be interested to hear what
                          Joe has to say about that, but it
                          seems to
                          > me that it could lead to
                          another way of improving decodes with or
                          without
                          > using deep
                          search.
                          >
                          > 73, Russ K2TXB
                          >

                        • Dave Miller
                          And.... before any gets upset with me cause they think I am bashing WSJT and deep search... please let me modify my earlier statement, I much prefer that to
                          Message 12 of 13 , Aug 29, 2008
                          • 0 Attachment
                            And.... before any gets upset with me 'cause they think I am bashing
                            WSJT and deep search... please let me modify my earlier statement,

                            "I much prefer that to telling the WSJT decoder half of the exchange
                            so that it can tell me who is talking to me!"

                            I should have left off the last two words of that sentence, "to me",
                            because many times I may be just listening in on a QSO and my call
                            will not be one of the calls that I need to decode. In that instance,
                            using only deep search, I would not be able to decode the signal
                            unless I knew at least one of the calls and entered it into My Call in
                            the setup options. I much prefer to be able to "tune in" any signal
                            and decode it without first knowing what I want to decode.

                            Oh boy, when I say it like that... I will probably get in more trouble
                            that if I would have let my previous post stand.

                            But this is an awesome program and I am only trying to learn how to
                            maximize its potential toward digging out the unknown week sigs. So
                            please be nice.

                            de N9SS, Dave

                            --- In wsjtgroup@yahoogroups.com, "Chet, N8RA" <chetsubaccount@...> wrote:
                            >
                            > Dave,
                            >
                            > You mentioned a combo that I don't remember trying when I went thru the
                            > tutorial hints at decoding the second signal.
                            > So I tried decoding the SP6 with my normal settings, specifically
                            with my
                            > call left in the options screen. I
                            > could not decode that signal either with or without the aggressive deep
                            > search selected.
                            >
                            > What you have found remains quite interesting.
                            >
                            > 73
                            > Chet
                            >
                            > _____
                            >
                            > From: wsjtgroup@yahoogroups.com [mailto:wsjtgroup@yahoogroups.com]
                            On Behalf
                            > Of Dave Miller
                            > Sent: Thursday, August 28, 2008 9:09 PM
                            > To: wsjtgroup@yahoogroups.com
                            > Subject: [wsjtgroup] Re: WSJT Tutorial & Phantom signal
                            >
                            >
                            >
                            > Ok thanks to all for all the great responses. You have all kept me
                            > very busy testing out various permutations and suggestions.
                            >
                            > Yes, I know that you have to select the sync sig 22hz from the main
                            > one, Have freeze selected and set TOL to 10hz. I have done this so
                            > many times now that I can do it blind-folded.
                            >
                            > What I missed was that you are supposed to run the tutorial with W1JT
                            > in the My Call setup. When I did that AND turned on the Deep Search
                            > option I could then successfully decode the second station without
                            > changing my RATE IN to 1.005.
                            >
                            > And yes, I can decode the second signal, WITHOUT DEEP SEARCH turned on
                            > AND with MY CALL set to my actual call, IF I change RATE IN to
                            > somewhere around 1.003 to 1.005 depending on the computer I am using.
                            > I can ONLY copy the first (louder) DL7 signal at this RATE IN setting
                            > IF and only IF I enter K1JT in the MY CALL setup AND turn on DEEP
                            SEARCH.
                            >
                            > Faced with the options here I think I will investigate a way to
                            > extract the correct offset so I can tune the station in, so to speak,
                            > when the station is not at the perfect clock rate. I much prefer that
                            > to telling the WSJT decoder half of the exchange so that it can tell
                            > me who is talking to me!
                            >
                            > What still puzzles me is why some of you indicated that you were able
                            > to copy the second SP6 signal WITHOUT changing any of your normal
                            > settings. Does this include the My Call setting?
                            >
                            > Thanks all again for your thorough responses!
                            >
                            > de N9SS, Dave EN50
                            >
                            > --- In wsjtgroup@yahoogrou <mailto:wsjtgroup%40yahoogroups.com> ps.com,
                            > "Russ K2TXB" <k2txb@> wrote:
                            > >
                            > > Barry, VE3CDX/W7 wrote:
                            > >
                            > > > I decode it just fine without changing any of my setup. The signal
                            > is 22
                            > > Hz
                            > > > away so setting TOL to 10 will allow it to be decoded just fine.
                            > >
                            > > Hi Barry. Yes, well I don't have the file as I have not downloaded
                            > the newer
                            > > WSJT programs (as I am not currently on the air). Of course I have
                            > often
                            > > used freeze and tolerance to copy weak signals close to strong ones.
                            > > Perhaps the gentlemen who were discussing changing the sound card
                            > frequency
                            > > settings to copy that signal were not aware that there was a more
                            direct
                            > > way. They seemed knowledgeable to me, so I did not consider that
                            > they may
                            > > not have tried tolerance and freeze. If the signal can easily be
                            > decoded
                            > > without sound card frequency change, then my speculation about an
                            > improved
                            > > algorithm is null and void (programming terms), hi.
                            > >
                            > > 73, Russ K2TXB
                            > >
                            > >
                            > > --------------- original message ------------------------------
                            > > From: wsjtgroup@yahoogrou <mailto:wsjtgroup%40yahoogroups.com> ps.com
                            > [mailto:wsjtgroup@yahoogrou <mailto:wsjtgroup%40yahoogroups.com> ps.com]
                            > On Behalf
                            > > Of Russ K2TXB
                            > > Sent: Thursday, August 28, 2008 6:49 AM
                            > > To: wsjtgroup@yahoogrou <mailto:wsjtgroup%40yahoogroups.com> ps.com
                            > > Subject: RE: [wsjtgroup] Re: WSJT Tutorial & Phantom signal
                            > >
                            > >
                            > > Gentlemen: Interesting discussion and results. The deep search
                            > algorithm
                            > > uses a process similar to what you do when you copy a call
                            > partially, and
                            > > you think you might know that call. You hear K2??B and think "I
                            > wonder if
                            > > that is Russ, K2TXB?". So you listen carefully to the middle two
                            > letters.
                            > > Now that you think you 'know' what they are, sure enough you copy
                            > the T and
                            > > X weakly. Now you are sure. Conversely, you think the first letter
                            > is an A
                            > > and not sure on the next letter. So you think "No I don't think
                            that is
                            > > him", and so you consider all the other calls you know to see if any
                            > of them
                            > > fit the pattern. If one does, you repeat the process of trying to
                            > be sure
                            > > the missing letter fits.
                            > >
                            > > Deep search works a lot like that. It is astoundingly good at
                            > pulling weak,
                            > > known, calls out of the noise. (And a lot of people cannot get
                            > their heads
                            > > around the fact that it works just like people do it in their heads,
                            > and try
                            > > to say it is not sufficient to identify a call because it is
                            known. But
                            > > that is another subject.)
                            > >
                            > > But when you discover that a slight tweak to the sound card
                            > frequency can
                            > > decode that same signal without using deep search, that is something
                            > indeed!
                            > > I will be interested to hear what Joe has to say about that, but it
                            > seems to
                            > > me that it could lead to another way of improving decodes with or
                            > without
                            > > using deep search.
                            > >
                            > > 73, Russ K2TXB
                            > >
                            >
                          • Joe Taylor
                            Hi Dave and all, I m just back from vacation, trying to get caught up on email. I ve read your series of messages on trying to decode both signals in the
                            Message 13 of 13 , Sep 3, 2008
                            • 0 Attachment
                              Hi Dave and all,

                              I'm just back from vacation, trying to get caught up on
                              email. I've read your series of messages on trying to
                              decode both signals in the DL7UAE file, and the several
                              responses. It seems that you have it all figured out by now
                              ... but just in case, the essential points are these:

                              1. Since it's going to be a Deep Search decode, you need to
                              have Deep Search enabled. In earlier versions of WSJT this
                              was true with the default setup; more recently (since
                              v5.9.5, as I recall) it has been OFF by default.

                              2. Again, since it's going to be a Deep Search decode, of
                              course you must have "My Call" set to K1JT. (If you don't
                              understand how DS works, you should read
                              http://physics.princeton.edu/pulsar/K1JT/JT65.pdf .) This
                              is precisely why Step #1 of the Tutorial says:

                              "1. If you have already installed and used a version of
                              WSJT6, delete your WSJT.INI file or temporarily rename it to
                              something else. This will ensure that you start the tutorial
                              with everything in the default configuration."

                              As for your results of playing with the sample-rate
                              correction factor: as you understood, the factor should be
                              1.0000 when doing the Tutorial, because the sample files
                              were recorded on a system without any sample-rate error.

                              The SP6GWB signal happens to be very close to the threshold
                              of decodability with the KV encoder. In such a situation,
                              small tweaks of anything that affects the sync process can
                              sometimes make the difference between decode and no decode.
                              You found a tweak that works in this particular case, but
                              it won't be useful in general because it depends on details
                              of the random noise in this particular file.

                              Hope this helps!

                              -- 73, Joe, K1JT

                              Dave Miller wrote:
                              > Has anyone ever tried to decode the "hidden" signal in the DL7UAE file
                              > under the JT65B folder of the WSJT samples download. The WSJT6 User
                              > manual's tutorial (page 9) says that you should be able to decode a
                              > second station (SP6GWB) just 22 hz away from the primary signal
                              > (DL7UAE). I have followed the tutorial step by step dozens of times
                              > with Versions 4.9, 5.9, and the 7.0 beta and all yield the same
                              > results; no SP6GWB signal, only the DL7UAE signal is decoded. When I
                              > follow the directions I do get a header line with a * or a #
                              > terminator, but nothing to the right but blank space. Has anyone else
                              > successfully decoded this second signal? I am trying to learn the ins
                              > and outs of this program before I get on the air and mess up a QSO.
                              >
                              > Dave N9SS EN50
                            Your message has been successfully submitted and would be delivered to recipients shortly.