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

Echo timing error WSJT 9

Expand Messages
  • orcityus
    Anyone else having this problem? Can you change the xmit keying timing? This is happening on 2 different computers. Keyed Xmit come on for 2.4 seconds tone
    Message 1 of 8 , Sep 20 4:11 AM
    • 0 Attachment
      Anyone else having this problem? Can you change the xmit keying timing? This is happening on 2 different computers.


      Keyed Xmit come on for 2.4 seconds tone start 1 seconds later after keyed starts and is on for 2.2 Seconds. Tone is on for 2.2 seconds total. tone is on 1 seconds into TX mode and 1.1 seconds in RX mode.

      TX tone overlaps RX mode.
      [ --- tone 2.2 sec +- ---]
      [ keyed xmit on 2.4 sec +- ][ RX 3.6 sec +----------------]
      [ --------------------------total cycle 6 seconds-----------]

      73 Dan NN7J@...
    • Marco IK1ODO
      ... I agree with Dan, the timing of audio in echo mode is... strange. The TX audio signal continues for about 1.1 s after PTT is off. In my radio (IC-910H),
      Message 2 of 8 , Sep 21 1:23 PM
      • 0 Attachment
        At 00:16 21/09/2010, you wrote:
        >
        >
        >Anyone else having this problem? Can you change the xmit keying
        >timing? This is happening on 2 different computers.

        I agree with Dan, the timing of audio in echo mode is... strange.
        The TX audio signal continues for about 1.1 s after PTT is off. In my
        radio (IC-910H), when in RX, some audio signal leaks from the rear
        audio input to the speaker, and appears on the waterfall. So I see
        echoes (with wrong timing) even without antenna... :-) and more, this
        prevents using VOX for TX.

        May I suggest that the audio tone be stopped when turning off the PTT signal?

        73 - Marco IK1ODO
      • Joe Taylor
        Hi Marco and all, Please see below for a copy of a message sent to a few others about the timing in WSJT 9.0 Echo mode. ... Timing measurements inside WSJT 9.0
        Message 3 of 8 , Sep 21 1:32 PM
        • 0 Attachment
          Hi Marco and all,

          Please see below for a copy of a message sent to a few others about the
          timing in WSJT 9.0 Echo mode.

          Marco IK1ODO wrote:
          > I agree with Dan, the timing of audio in echo mode is... strange.
          > The TX audio signal continues for about 1.1 s after PTT is off. In my
          > radio (IC-910H), when in RX, some audio signal leaks from the rear
          > audio input to the speaker, and appears on the waterfall. So I see
          > echoes (with wrong timing) even without antenna... :-) and more, this
          > prevents using VOX for TX.
          >
          > May I suggest that the audio tone be stopped when turning off the PTT signal?
          >
          > 73 - Marco IK1ODO

          Timing measurements inside WSJT 9.0 r2136 show the following:

          State Time Program Action Taken
          --------------------------------------
          t1 0.0 Raise PTT (start of cycle)
          t2 0.2 Start Tx Audio
          t3 2.6 Stop Tx Audio
          t4 2.8 Lower PTT
          t5 3.0 Start saving Rx samples
          t6 5.2 Stop Rx

          (Times are in seconds, rounded to the nearest 0.1 s.)

          There is a system-dependent latency between time t2 -- the time when the
          program starts sending Tx audio data to the operating system's output
          sound driver -- and the time when the audio signal appears on the
          computer's "Line out" jack. Similarly, there's a latency between the
          time when an Rx sample is produced by the A/D converter and the time
          (t5) when it can first be picked up by the program.

          For sake of discussion, let's say these latencies are both 0.3 s. The
          augmented timing table then looks like this:

          State Time Program Action Taken
          --------------------------------------------
          t1 0.0 Raise PTT (start of cycle)
          t2 0.2 Start Tx Audio
          0.5 Actual transmission begins
          t3 2.6 Stop Tx Audio
          2.7 First Rx audio sample taken
          t4 2.8 Lower PTT
          2.9 Transmission ends
          t5 3.0 Start saving Rx samples
          t6 5.2 Stop Rx

          So in this example, yes: the first sample of received audio is taken
          before the transmission has ended.

          In my station, with WSJT 9.0 running under my Windows 7, this scheme
          seems to work fine. I haven't measured the latencies, but apparently
          they are small enough.

          To some extent it's possible to control the latencies by means of
          parameters set at program startup. I can adjust these downward, to see
          what happens. I can also adjust the programmed times at which things
          happen in the 6-second Echo cycle. Maybe I'll get to it next week, or
          the week after.

          -- 73, Joe, K1JT
        • Marco IK1ODO
          ... Hello Joe, all very clear. Probably my audio chain adds a long delay, I measure approx 1.1 s from PTT off to audio stop. I use Win XP Pro and an external
          Message 4 of 8 , Sep 21 1:44 PM
          • 0 Attachment
            At 22:32 21/09/2010, Joe Taylor wrote:
            >
            >
            >Hi Marco and all,
            >
            >Please see below for a copy of a message sent to a few others about the
            >timing in WSJT 9.0 Echo mode.

            Hello Joe,

            all very clear. Probably my audio chain adds a long delay, I measure
            approx 1.1 s from PTT off to audio stop.
            I use Win XP Pro and an external USB (Creative) sound interface.
            Tomorrow I will look if there is some way to reduce the buffer lenght.
            I wanted to keep that interface for my radio uses, to free the main
            audio channel (on motherboard) for other uses... but of course may
            try to drive the radio from the other output.

            73 - Marco IK1ODO
          • Bill VanAlstyne W5WVO
            I m not sure exactly how much latency to expect from a USB sound device, Marco, but it seems to me that it has to be a lot greater than an internal bus-mounted
            Message 5 of 8 , Sep 21 5:59 PM
            • 0 Attachment
              I'm not sure exactly how much latency to expect from a USB sound device, Marco, but it seems to me that it has to be a lot greater than an internal bus-mounted sound device. My Windows 7 desktop machine uses the PCI-X bus, and this interface is quite fast. When I first set up this computer, I (like you) installed an inexpensive USB sound device for use with my radio-related software. But the cheap hardware in it failed within a month or so, I think from being overdriven while operating with the AGC turned off... :-)  Instead of replacing it, I decided to spend the additional money and buy a  second identical PCI-X bus internal sound card. It's unfortunate that PCI-X bus peripherals are as expensive as they are right now, but that's the reality. Hopefully the prices will come down over the next couple years as PCI-X becomes ubiquitous.
               
              Perhaps WSJT could incorporate a few different latency models, selectable based on the bus architecture through which the sound device is accessed... ? Don't know if that would solve the problem, but it sounds plausible to me.
               
              If you do end up using the motherboard (default) sound channels, Marco, remember to disable all the Windows O/S sounds so they don't get transmitted while you're on the air.
               
              Bill W5WVO
               

              Sent: Tuesday, September 21, 2010 20:44
              Subject: Re: [wsjtgroup] Echo timing error WSJT 9

               

              At 22:32 21/09/2010, Joe Taylor wrote:

              >
              >
              >Hi Marco and
              all,
              >
              >Please see below for a copy of a message sent to a few
              others about the
              >timing in WSJT 9.0 Echo mode.

              Hello Joe,

              all very clear. Probably my audio chain adds a long delay, I measure
              approx 1.1 s from PTT off to audio stop.
              I use Win XP Pro and an external USB (Creative) sound interface.
              Tomorrow I will look if there is some way to reduce the buffer lenght.
              I wanted to keep that interface for my radio uses, to free the main
              audio channel (on motherboard) for other uses... but of course may
              try to drive the radio from the other output.

              73 - Marco IK1ODO

            • Barry Garratt
              Check this out. It may help determine where your latency problems are. It s actually quite good. http://www.thesycon.de/deu/latency_check.shtml Barry KS7DX
              Message 6 of 8 , Sep 21 6:12 PM
              • 0 Attachment

                Check this out. It may help determine where your latency problems are. It’s actually quite good.

                 

                http://www.thesycon.de/deu/latency_check.shtml

                 

                Barry KS7DX

                 

                 

                From: wsjtgroup@yahoogroups.com [mailto:wsjtgroup@yahoogroups.com] On Behalf Of Bill VanAlstyne W5WVO
                Sent: Tuesday, September 21, 2010 5:59 PM
                To: [WSJTGROUP]
                Subject: Re: [wsjtgroup] Echo timing error WSJT 9

                 

                 

                I'm not sure exactly how much latency to expect from a USB sound device, Marco, but it seems to me that it has to be a lot greater than an internal bus-mounted sound device. My Windows 7 desktop machine uses the PCI-X bus, and this interface is quite fast. When I first set up this computer, I (like you) installed an inexpensive USB sound device for use with my radio-related software. But the cheap hardware in it failed within a month or so, I think from being overdriven while operating with the AGC turned off... :-)  Instead of replacing it, I decided to spend the additional money and buy a  second identical PCI-X bus internal sound card. It's unfortunate that PCI-X bus peripherals are as expensive as they are right now, but that's the reality. Hopefully the prices will come down over the next couple years as PCI-X becomes ubiquitous.

                 

                Perhaps WSJT could incorporate a few different latency models, selectable based on the bus architecture through which the sound device is accessed... ? Don't know if that would solve the problem, but it sounds plausible to me.

                 

                If you do end up using the motherboard (default) sound channels, Marco, remember to disable all the Windows O/S sounds so they don't get transmitted while you're on the air.

                 

                Bill W5WVO

                 

                 

                Sent: Tuesday, September 21, 2010 20:44

                Subject: Re: [wsjtgroup] Echo timing error WSJT 9

                 

                 

                At 22:32 21/09/2010, Joe Taylor wrote:

                >
                >
                >Hi Marco and all,
                >
                >Please see below for a copy of a message sent to a few others about the
                >timing in WSJT 9.0 Echo mode.

                Hello Joe,

                all very clear. Probably my audio chain adds a long delay, I measure
                approx 1.1 s from PTT off to audio stop.
                I use Win XP Pro and an external USB (Creative) sound interface.
                Tomorrow I will look if there is some way to reduce the buffer lenght.
                I wanted to keep that interface for my radio uses, to free the main
                audio channel (on motherboard) for other uses... but of course may
                try to drive the radio from the other output.

                73 - Marco IK1ODO

              • Marco IK1ODO
                Still about the bad timing of echo mode when using USB sond cards. Would it be possible in the next version of WSJT to optionally use ASIO drivers? Those are
                Message 7 of 8 , Sep 24 12:22 PM
                • 0 Attachment
                  Still about the bad timing of echo mode when using USB sond cards.
                  Would it be possible in the next version of WSJT to optionally use
                  ASIO drivers? Those are free on the web, and have very low latency
                  (40 ms is a typical value). They work with my different audio
                  interfaces, but the program must address them instead of the Windows
                  standard drivers. Just as an example, Winrad by I2PHD uses ASIO
                  drivers (if available) and the latency is minimal.

                  73 - Marco IK1ODO
                • Bill VanAlstyne W5WVO
                  Marco, I m not the programmer of WSJT. I m forwarding your email on to the WSJT Group mailing list and to Joe K1JT. 73, Bill W5WVO ... From: Marco IK1ODO
                  Message 8 of 8 , Sep 24 12:23 PM
                  • 0 Attachment
                    Marco, I'm not the programmer of WSJT. I'm forwarding your email on to the
                    WSJT Group mailing list and to Joe K1JT.

                    73,
                    Bill W5WVO

                    --------------------------------------------------
                    From: "Marco IK1ODO" <ik1odo@...>
                    Sent: Friday, September 24, 2010 19:18
                    To: "Bill VanAlstyne W5WVO" <w5wvo@...>
                    Subject: Re: [wsjtgroup] Echo timing error WSJT 9

                    > Still about the bad timing of echo mode when using USB sond cards. Would
                    > it be possible in the next version of WSJT to optionally use ASIO drivers?
                    > Those are free on the web, and have very low latency (40 ms is a typical
                    > value). They work with my different audio interfaces, but the program must
                    > address them instead of the Windows standard drivers. Just as an example,
                    > Winrad by I2PHD uses ASIO drivers (if available) and the latency is
                    > minimal.
                    >
                    > 73 - Marco IK1ODO
                    >
                    >
                    >
                  Your message has been successfully submitted and would be delivered to recipients shortly.