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

28230 Print

Expand Messages
  • Don Poaps
    DE VA7DGP CN89ME WWW.QRZ/DB/VA7DGP 73 Don DE VA+Sa�9�9ME
    Message 1 of 25 , Oct 26, 2013
    • 0 Attachment
      DE VA7DGP CN89ME WWW.QRZ/DB/VA7DGP 73 Don
      DE VA+Sa�9�9ME WvY2Gxh�/VA7DKB������L�����L�����Fگ��������0n
      Arcadia, FL.This is SHORT skip!!!CHAT 0.1.6.0 TEST - K1DOW - Arcadia,
      Florida
      |T�������HoYЂ�ie�hp�uW�Q+�EY���U#�YR���z:�b�i�F�Jt�����wb&E�Q�܅UZ`AYjee&f��YYM&0B����j�*ͺYV���]KZ������ZT,&��#��Fw{y�O��v�w�y}4"�Y�Z`��U���Y8�A�
      - H%JE^ H�"`�f��&]Na�&����VA%FUTy��������Sa TEST TEST DE KK8G
      KK8G r
      ����$I��P\�(����D� *** Beacon sent @ 2013/10/26 19:25:52 UTC

      DE VA7DGP CN89ME WWW.QRZ/DB/VA7DGP 73 Don
      DE VAwE��9ME YUW.QRZ/�Sl�u�~
      'S�R̪¬��(��%�/Y#��F!�.*5T@CH�
      \���hWKRY�p�P,�R%4,XlECKz�ʍt�7��~��+JZ��A�AGEt�b
      %f���V�V�{Ɵj�V�Qf��VY�U��wi���eTG�i����W���,"-


      --
      Don Poaps
      New Westminster, BC
      VA7DGP
      Satern Member
      Vector Member
      Surrey Amateur Radio
      Metrotown Salvation Army Community Church Member
    • Rick Muething
      All, Here in LA. No radio equipment so can’t participate in any over the air testing. I have investigated some of the decoding problems observed showing a
      Message 2 of 25 , Oct 26, 2013
      • 0 Attachment
         
         
        All,
         
        Here in LA.  No radio equipment so can’t participate in any over the air testing.
         
        I have investigated some of the decoding problems observed showing a cycling of the decode quality  which look something like this:
         
        image
         
        I am now pretty convinced this is due to some significant errors in the sound card capture or Playback rate.  The current beta software should have enough correction capability to handle up to +/- 1000 ppm (.1% ) error but the above type of distortion seen in some cases is something more like 10,000 ppm or 1% .Pretty bad for even a cheap sound card
        I have uploaded a utility program found in the MixW software package called CheckSR.exe   to the V4Protocol FTP section.  If you download this simple utility you can check the accuracy of your sound card. 
        I would suggest if you are seeing results like the above or others have reported seeing your signal like this then run the simple CheckSR.exe program.
        To run select 12000 Hz sample rate (that is what is used on the H4Chat program.  Start it and let it run at least a minute or more. Stop and look at the errors reported in ppm.   If they are over 1000 the sound card is suspect.  If you see over 1000 on the capture rate try another test at 48000 and see how it compares to the first test. (normally 48000 and 12000 are more “cardinal” rates which have been shown in many tests to have the lowest errors on most sound cards).
        There are some ways to compensate for this error in the software (the “calibration” is really just a compensation in the H4TNC DSP code) and I am evaluating some different ways.  If  you have a sound card or sound card interface that is over 1000 ppm please report it to the reflector along with the sound card type so we have some idea of what percentage of those with problems we have.
        Just for reference: The SCS PTC II and DR7xxx DSP TNC have a very stable temperature compensated Xtal oscillator that is better than 10 ppm.  This is what is required for really high performance DSP.  Some tests I have done on the built in sound cards on the newer radios (Kenwood TS-590, Icom 7100, 7200, 7600) have been very good too...< 100 ppm.
         
        73,
         
        Rick KN6KB
         
      • n6kzb
        Rick in answer to your query, ICOM 7100 sound card rate after 25 minutes with ratechecker is: Input diff in PPM -2 Output diff in PPM -1 Mike ... All, Here
        Message 3 of 25 , Oct 26, 2013
        • 0 Attachment

          Rick in answer to your query, ICOM 7100 sound card rate after 25 minutes with ratechecker is: 

          Input diff in PPM   -2  Output diff in PPM -1


          Mike

           



          ---In V4Protocol@yahoogroups.com, <v4protocol@yahoogroups.com> wrote:

           
           
          All,
           
          Here in LA.  No radio equipment so can’t participate in any over the air testing.
           
          I have investigated some of the decoding problems observed showing a cycling of the decode quality  which look something like this:
           
          image
           
          I am now pretty convinced this is due to some significant errors in the sound card capture or Playback rate.  The current beta software should have enough correction capability to handle up to +/- 1000 ppm (.1% ) error but the above type of distortion seen in some cases is something more like 10,000 ppm or 1% .Pretty bad for even a cheap sound card
          I have uploaded a utility program found in the MixW software package called CheckSR.exe   to the V4Protocol FTP section.  If you download this simple utility you can check the accuracy of your sound card. 
          I would suggest if you are seeing results like the above or others have reported seeing your signal like this then run the simple CheckSR.exe program.
          To run select 12000 Hz sample rate (that is what is used on the H4Chat program.  Start it and let it run at least a minute or more. Stop and look at the errors reported in ppm.   If they are over 1000 the sound card is suspect.  If you see over 1000 on the capture rate try another test at 48000 and see how it compares to the first test. (normally 48000 and 12000 are more “cardinal” rates which have been shown in many tests to have the lowest errors on most sound cards).
          There are some ways to compensate for this error in the software (the “calibration” is really just a compensation in the H4TNC DSP code) and I am evaluating some different ways.  If  you have a sound card or sound card interface that is over 1000 ppm please report it to the reflector along with the sound card type so we have some idea of what percentage of those with problems we have.
          Just for reference: The SCS PTC II and DR7xxx DSP TNC have a very stable temperature compensated Xtal oscillator that is better than 10 ppm.  This is what is required for really high performance DSP.  Some tests I have done on the built in sound cards on the newer radios (Kenwood TS-590, Icom 7100, 7200, 7600) have been very good too...< 100 ppm.
           
          73,
           
          Rick KN6KB
           
        • Kirk
          Rick, That’s an interesting and worthwhile find. I’ve used MixW many times but never noticed that ap. I’m seeing around -50 both in and out on my sound
          Message 4 of 25 , Oct 26, 2013
          • 0 Attachment

            Rick,

             

            That’s an interesting and worthwhile find.  I’ve used MixW many times but never noticed that ap.  I’m seeing around -50 both in and out on my sound card. 

             

            Thanks,

             

            Kirk, K6KAR

             

            From: V4Protocol@yahoogroups.com [mailto:V4Protocol@yahoogroups.com] On Behalf Of Rick Muething
            Sent: Saturday, October 26, 2013 3:13 PM
            To: V4Protocol@yahoogroups.com
            Subject: [V4Protocol] Sound Card Sample rate checker

             

             

             

             

            All,

             

            Here in LA.  No radio equipment so can’t participate in any over the air testing.

             

            I have investigated some of the decoding problems observed showing a cycling of the decode quality  which look something like this:

             

            image

             

            I am now pretty convinced this is due to some significant errors in the sound card capture or Playback rate.  The current beta software should have enough correction capability to handle up to +/- 1000 ppm (.1% ) error but the above type of distortion seen in some cases is something more like 10,000 ppm or 1% .Pretty bad for even a cheap sound card

            I have uploaded a utility program found in the MixW software package called CheckSR.exe   to the V4Protocol FTP section.  If you download this simple utility you can check the accuracy of your sound card. 

            I would suggest if you are seeing results like the above or others have reported seeing your signal like this then run the simple CheckSR.exe program.

            To run select 12000 Hz sample rate (that is what is used on the H4Chat program.  Start it and let it run at least a minute or more. Stop and look at the errors reported in ppm.   If they are over 1000 the sound card is suspect.  If you see over 1000 on the capture rate try another test at 48000 and see how it compares to the first test. (normally 48000 and 12000 are more “cardinal” rates which have been shown in many tests to have the lowest errors on most sound cards).

            There are some ways to compensate for this error in the software (the “calibration” is really just a compensation in the H4TNC DSP code) and I am evaluating some different ways.  If  you have a sound card or sound card interface that is over 1000 ppm please report it to the reflector along with the sound card type so we have some idea of what percentage of those with problems we have.

            Just for reference: The SCS PTC II and DR7xxx DSP TNC have a very stable temperature compensated Xtal oscillator that is better than 10 ppm.  This is what is required for really high performance DSP.  Some tests I have done on the built in sound cards on the newer radios (Kenwood TS-590, Icom 7100, 7200, 7600) have been very good too...< 100 ppm.

             

            73,

             

            Rick KN6KB

             

          • bob_emerson2
            At least in my case, CheckSR gives very inaccurate results when used with a SignaLink USB. With a rate of 48000 selected: CheckSR reports my SignaLink has an
            Message 5 of 25 , Oct 26, 2013
            • 0 Attachment

              At least in my case, CheckSR gives very inaccurate results when used with a SignaLink USB.


              With a rate of 48000 selected: CheckSR reports my SignaLink has an input rate of 46044.45 and an output of 46010.02.  Off by almost 2000 Hz.


              Using Flidgi's WWV sound card calibration, the SignaLink is only off from 48000 by 47 ppm. And the 47 ppm adjustment works in Flidgi.


              Irv W4IWK





              ---In V4Protocol@yahoogroups.com, <v4protocol@yahoogroups.com> wrote:

               
               
              All,
               
              Here in LA.  No radio equipment so can’t participate in any over the air testing.
               
              I have investigated some of the decoding problems observed showing a cycling of the decode quality  which look something like this:
               
              image
               
              I am now pretty convinced this is due to some significant errors in the sound card capture or Playback rate.  The current beta software should have enough correction capability to handle up to +/- 1000 ppm (.1% ) error but the above type of distortion seen in some cases is something more like 10,000 ppm or 1% .Pretty bad for even a cheap sound card
              I have uploaded a utility program found in the MixW software package called CheckSR.exe   to the V4Protocol FTP section.  If you download this simple utility you can check the accuracy of your sound card. 
              I would suggest if you are seeing results like the above or others have reported seeing your signal like this then run the simple CheckSR.exe program.
              To run select 12000 Hz sample rate (that is what is used on the H4Chat program.  Start it and let it run at least a minute or more. Stop and look at the errors reported in ppm.   If they are over 1000 the sound card is suspect.  If you see over 1000 on the capture rate try another test at 48000 and see how it compares to the first test. (normally 48000 and 12000 are more “cardinal” rates which have been shown in many tests to have the lowest errors on most sound cards).
              There are some ways to compensate for this error in the software (the “calibration” is really just a compensation in the H4TNC DSP code) and I am evaluating some different ways.  If  you have a sound card or sound card interface that is over 1000 ppm please report it to the reflector along with the sound card type so we have some idea of what percentage of those with problems we have.
              Just for reference: The SCS PTC II and DR7xxx DSP TNC have a very stable temperature compensated Xtal oscillator that is better than 10 ppm.  This is what is required for really high performance DSP.  Some tests I have done on the built in sound cards on the newer radios (Kenwood TS-590, Icom 7100, 7200, 7600) have been very good too...< 100 ppm.
               
              73,
               
              Rick KN6KB
               
            • Steve Michaels
              de N4AY checksr.exe Soundblaster Go-Fi Pro external 12000hz Input 12127.91 Output 12002.18 48000 Input 47997.58 Output 47997.58 de N4AY ... de
              Message 6 of 25 , Oct 26, 2013
              • 0 Attachment
                de N4AY  checksr.exe

                Soundblaster Go-Fi Pro external

                12000hz

                Input 12127.91     Output   12002.18

                48000

                Input  47997.58     Output   47997.58

                de N4AY




                On Sat, Oct 26, 2013 at 7:47 PM, <w4iwk@...> wrote:
                 

                At least in my case, CheckSR gives very inaccurate results when used with a SignaLink USB.


                With a rate of 48000 selected: CheckSR reports my SignaLink has an input rate of 46044.45 and an output of 46010.02.  Off by almost 2000 Hz.


                Using Flidgi's WWV sound card calibration, the SignaLink is only off from 48000 by 47 ppm. And the 47 ppm adjustment works in Flidgi.


                Irv W4IWK





                ---In V4Protocol@yahoogroups.com, <v4protocol@yahoogroups.com> wrote:

                 
                 
                All,
                 
                Here in LA.  No radio equipment so can’t participate in any over the air testing.
                 
                I have investigated some of the decoding problems observed showing a cycling of the decode quality  which look something like this:
                 
                image
                 
                I am now pretty convinced this is due to some significant errors in the sound card capture or Playback rate.  The current beta software should have enough correction capability to handle up to +/- 1000 ppm (.1% ) error but the above type of distortion seen in some cases is something more like 10,000 ppm or 1% .Pretty bad for even a cheap sound card
                I have uploaded a utility program found in the MixW software package called CheckSR.exe   to the V4Protocol FTP section.  If you download this simple utility you can check the accuracy of your sound card. 
                I would suggest if you are seeing results like the above or others have reported seeing your signal like this then run the simple CheckSR.exe program.
                To run select 12000 Hz sample rate (that is what is used on the H4Chat program.  Start it and let it run at least a minute or more. Stop and look at the errors reported in ppm.   If they are over 1000 the sound card is suspect.  If you see over 1000 on the capture rate try another test at 48000 and see how it compares to the first test. (normally 48000 and 12000 are more “cardinal” rates which have been shown in many tests to have the lowest errors on most sound cards).
                There are some ways to compensate for this error in the software (the “calibration” is really just a compensation in the H4TNC DSP code) and I am evaluating some different ways.  If  you have a sound card or sound card interface that is over 1000 ppm please report it to the reflector along with the sound card type so we have some idea of what percentage of those with problems we have.
                Just for reference: The SCS PTC II and DR7xxx DSP TNC have a very stable temperature compensated Xtal oscillator that is better than 10 ppm.  This is what is required for really high performance DSP.  Some tests I have done on the built in sound cards on the newer radios (Kenwood TS-590, Icom 7100, 7200, 7600) have been very good too...< 100 ppm.
                 
                73,
                 
                Rick KN6KB
                 


              • Ron Wenig
                Hi Rick, I ran the CheckSR program on my Kenwood TS-590 with its built in sound card. The difference PPM was 0, both input and output. When I see this kind of
                Message 7 of 25 , Oct 27, 2013
                • 0 Attachment
                  Hi Rick,

                  I ran the CheckSR program on my Kenwood TS-590 with its built in sound card.  The difference PPM was 0, both input and output.

                  When I see this kind of decoding problems I just take it for granted that it's just poor propagation.

                  73, Ron ny3j

                  On 10/26/2013 4:13 PM, Rick Muething wrote:
                   

                   
                   
                  All,
                   
                  Here in LA.  No radio equipment so can’t participate in any over the air testing.
                   
                  I have investigated some of the decoding problems observed showing a cycling of the decode quality  which look something like this:


                • k1dow_4
                  Rick; Great idea on suggesting the group check their sound cards for error! Here is my report for my Creative Soundblaster PCI sound card tested at 12000Hz:
                  Message 8 of 25 , Oct 27, 2013
                  • 0 Attachment
                    Rick;

                    Great idea on suggesting the group check their sound cards for error!

                    Here is my report for my Creative Soundblaster PCI sound card tested at 12000Hz:
                       
                                                                          INPUT            OUTPUT
                                                                          ----------            ------------
                    Difference in ppm:                              184                   7


                    73/Russ
                    K1DOW

                                       
                       


                    ---In V4Protocol@yahoogroups.com, <v4protocol@yahoogroups.com> wrote:

                     
                     
                    All,
                     
                    Here in LA.  No radio equipment so can’t participate in any over the air testing.
                     
                    I have investigated some of the decoding problems observed showing a cycling of the decode quality  which look something like this:
                     
                    image
                     
                    I am now pretty convinced this is due to some significant errors in the sound card capture or Playback rate.  The current beta software should have enough correction capability to handle up to +/- 1000 ppm (.1% ) error but the above type of distortion seen in some cases is something more like 10,000 ppm or 1% .Pretty bad for even a cheap sound card
                    I have uploaded a utility program found in the MixW software package called CheckSR.exe   to the V4Protocol FTP section.  If you download this simple utility you can check the accuracy of your sound card. 
                    I would suggest if you are seeing results like the above or others have reported seeing your signal like this then run the simple CheckSR.exe program.
                    To run select 12000 Hz sample rate (that is what is used on the H4Chat program.  Start it and let it run at least a minute or more. Stop and look at the errors reported in ppm.   If they are over 1000 the sound card is suspect.  If you see over 1000 on the capture rate try another test at 48000 and see how it compares to the first test. (normally 48000 and 12000 are more “cardinal” rates which have been shown in many tests to have the lowest errors on most sound cards).
                    There are some ways to compensate for this error in the software (the “calibration” is really just a compensation in the H4TNC DSP code) and I am evaluating some different ways.  If  you have a sound card or sound card interface that is over 1000 ppm please report it to the reflector along with the sound card type so we have some idea of what percentage of those with problems we have.
                    Just for reference: The SCS PTC II and DR7xxx DSP TNC have a very stable temperature compensated Xtal oscillator that is better than 10 ppm.  This is what is required for really high performance DSP.  Some tests I have done on the built in sound cards on the newer radios (Kenwood TS-590, Icom 7100, 7200, 7600) have been very good too...< 100 ppm.
                     
                    73,
                     
                    Rick KN6KB
                     
                  • William Calderwood
                    All -     My Signalink checks out thus: Input: -79 ppm  Output: -28 ppm                                   73, Bill K1CT On Sunday, October
                    Message 9 of 25 , Oct 27, 2013
                    • 0 Attachment
                      All -

                          My Signalink checks out thus:

                      Input: -79 ppm  Output: -28 ppm

                                                        73, Bill K1CT



                      On Sunday, October 27, 2013 8:03 AM, "russtower@..." <russtower@...> wrote:
                       
                      Rick;

                      Great idea on suggesting the group check their sound cards for error!

                      Here is my report for my Creative Soundblaster PCI sound card tested at 12000Hz:
                         
                                                                            INPUT            OUTPUT
                                                                            ----------            ------------
                      Difference in ppm:                              184                   7


                      73/Russ
                      K1DOW

                                         
                         


                      ---In V4Protocol@yahoogroups.com, <v4protocol@yahoogroups.com> wrote:

                       
                       
                      All,
                       
                      Here in LA.  No radio equipment so can’t participate in any over the air testing.
                       
                      I have investigated some of the decoding problems observed showing a cycling of the decode quality  which look something like this:
                       
                      image
                       
                      I am now pretty convinced this is due to some significant errors in the sound card capture or Playback rate.  The current beta software should have enough correction capability to handle up to +/- 1000 ppm (.1% ) error but the above type of distortion seen in some cases is something more like 10,000 ppm or 1% .Pretty bad for even a cheap sound card
                      I have uploaded a utility program found in the MixW software package called CheckSR.exe   to the V4Protocol FTP section.  If you download this simple utility you can check the accuracy of your sound card. 
                      I would suggest if you are seeing results like the above or others have reported seeing your signal like this then run the simple CheckSR.exe program.
                      To run select 12000 Hz sample rate (that is what is used on the H4Chat program.  Start it and let it run at least a minute or more. Stop and look at the errors reported in ppm.   If they are over 1000 the sound card is suspect.  If you see over 1000 on the capture rate try another test at 48000 and see how it compares to the first test. (normally 48000 and 12000 are more “cardinal” rates which have been shown in many tests to have the lowest errors on most sound cards).
                      There are some ways to compensate for this error in the software (the “calibration” is really just a compensation in the H4TNC DSP code) and I am evaluating some different ways.  If  you have a sound card or sound card interface that is over 1000 ppm please report it to the reflector along with the sound card type so we have some idea of what percentage of those with problems we have.
                      Just for reference: The SCS PTC II and DR7xxx DSP TNC have a very stable temperature compensated Xtal oscillator that is better than 10 ppm.  This is what is required for really high performance DSP.  Some tests I have done on the built in sound cards on the newer radios (Kenwood TS-590, Icom 7100, 7200, 7600) have been very good too...< 100 ppm.
                       
                      73,
                       
                      Rick KN6KB
                       


                    • pc_cruiser_pa
                      Hi Rick, I used the CheckSR.exe against the internal USB soundcard in my Icom 7600 and the external SignaLink USB. The numbers just jumped all over, but
                      Message 10 of 25 , Oct 27, 2013
                      • 0 Attachment

                         Hi Rick,


                        I used the CheckSR.exe against the internal USB soundcard in my Icom 7600 and the external SignaLink USB. The numbers just jumped all over, but neither was over 144. Then, without making any changes I checked both with fldigi using its WWV check. Using an adjustment of 0 for both sound inputs produced a straight line. I hope that small contribution will help.


                        I am, however, still puzzled by the very obvious dips in output when running FEC. This is not evident with the FSKID, just FEC text and the beacon.


                        Doug, N3JXB



                        ---In V4Protocol@yahoogroups.com, <v4protocol@yahoogroups.com> wrote:

                         
                         
                        All,
                         
                        Here in LA.  No radio equipment so can’t participate in any over the air testing.
                         
                        I have investigated some of the decoding problems observed showing a cycling of the decode quality  which look something like this:
                         
                        image
                         
                        I am now pretty convinced this is due to some significant errors in the sound card capture or Playback rate.  The current beta software should have enough correction capability to handle up to +/- 1000 ppm (.1% ) error but the above type of distortion seen in some cases is something more like 10,000 ppm or 1% .Pretty bad for even a cheap sound card
                        I have uploaded a utility program found in the MixW software package called CheckSR.exe   to the V4Protocol FTP section.  If you download this simple utility you can check the accuracy of your sound card. 
                        I would suggest if you are seeing results like the above or others have reported seeing your signal like this then run the simple CheckSR.exe program.
                        To run select 12000 Hz sample rate (that is what is used on the H4Chat program.  Start it and let it run at least a minute or more. Stop and look at the errors reported in ppm.   If they are over 1000 the sound card is suspect.  If you see over 1000 on the capture rate try another test at 48000 and see how it compares to the first test. (normally 48000 and 12000 are more “cardinal” rates which have been shown in many tests to have the lowest errors on most sound cards).
                        There are some ways to compensate for this error in the software (the “calibration” is really just a compensation in the H4TNC DSP code) and I am evaluating some different ways.  If  you have a sound card or sound card interface that is over 1000 ppm please report it to the reflector along with the sound card type so we have some idea of what percentage of those with problems we have.
                        Just for reference: The SCS PTC II and DR7xxx DSP TNC have a very stable temperature compensated Xtal oscillator that is better than 10 ppm.  This is what is required for really high performance DSP.  Some tests I have done on the built in sound cards on the newer radios (Kenwood TS-590, Icom 7100, 7200, 7600) have been very good too...< 100 ppm.
                         
                        73,
                         
                        Rick KN6KB
                         
                      • Ricky Faulkenberry
                        Sound blaster is an excellent card for digital modes; got two of them. KJ1j and I had qso about 12 cst. 28.230 worked fine, heard no beacons. ... Sound blaster
                        Message 11 of 25 , Oct 27, 2013
                        • 0 Attachment

                          Sound blaster is an excellent card for digital modes; got two of them. KJ1j and I had qso about 12 cst. 28.230 worked fine, heard no beacons.

                          On Oct 27, 2013 10:03 AM, <russtower@...> wrote:
                           

                          Rick;

                          Great idea on suggesting the group check their sound cards for error!

                          Here is my report for my Creative Soundblaster PCI sound card tested at 12000Hz:
                             
                                                                                INPUT            OUTPUT
                                                                                ----------            ------------
                          Difference in ppm:                              184                   7


                          73/Russ
                          K1DOW

                                             

                             


                          ---In V4Protocol@yahoogroups.com, <v4protocol@yahoogroups.com> wrote:

                           
                           
                          All,
                           
                          Here in LA.  No radio equipment so can’t participate in any over the air testing.
                           
                          I have investigated some of the decoding problems observed showing a cycling of the decode quality  which look something like this:
                           
                          image
                           
                          I am now pretty convinced this is due to some significant errors in the sound card capture or Playback rate.  The current beta software should have enough correction capability to handle up to +/- 1000 ppm (.1% ) error but the above type of distortion seen in some cases is something more like 10,000 ppm or 1% .Pretty bad for even a cheap sound card
                          I have uploaded a utility program found in the MixW software package called CheckSR.exe   to the V4Protocol FTP section.  If you download this simple utility you can check the accuracy of your sound card. 
                          I would suggest if you are seeing results like the above or others have reported seeing your signal like this then run the simple CheckSR.exe program.
                          To run select 12000 Hz sample rate (that is what is used on the H4Chat program.  Start it and let it run at least a minute or more. Stop and look at the errors reported in ppm.   If they are over 1000 the sound card is suspect.  If you see over 1000 on the capture rate try another test at 48000 and see how it compares to the first test. (normally 48000 and 12000 are more “cardinal” rates which have been shown in many tests to have the lowest errors on most sound cards).
                          There are some ways to compensate for this error in the software (the “calibration” is really just a compensation in the H4TNC DSP code) and I am evaluating some different ways.  If  you have a sound card or sound card interface that is over 1000 ppm please report it to the reflector along with the sound card type so we have some idea of what percentage of those with problems we have.
                          Just for reference: The SCS PTC II and DR7xxx DSP TNC have a very stable temperature compensated Xtal oscillator that is better than 10 ppm.  This is what is required for really high performance DSP.  Some tests I have done on the built in sound cards on the newer radios (Kenwood TS-590, Icom 7100, 7200, 7600) have been very good too...< 100 ppm.
                           
                          73,
                           
                          Rick KN6KB
                           
                        • Hoby KK8G
                          After using CHECKSR for several years, I have found that the samplerate of the soundcard is determined in reference to the time base of the computer. So the
                          Message 12 of 25 , Oct 27, 2013
                          • 0 Attachment
                            
                            After using CHECKSR for several years, I have found that the samplerate of the soundcard is determined in reference to the time base of the computer. So the card is usually very good unless the time base of the computer is poor. You can take the same soundcard and use it with two different computers and get two different results.
                            My two radio computers (one WIN7, AMD ATHLON II X4 650  GHz w/8 Gig RAM) routinely gives much better results from CHECKSR. The other (XP CELERON 440 - 2GHz w/1 Gig RAM) is not so stable. My SL on the WIN7 shows a +31 PPM Input, and + 37 PPM Output error. My XP shows the same SL with -97 PPM Input and -101 PPM Output error.
                            The clock on the Win7 machine will normally run several days with the clock varying only a very few seconds. The XP will run the same time, but the clock error will be many seconds, maybe even close to a minute.
                            This stability is shown when running M110A (MARS, not legal on Ham bands) the XP will lose sync much sooner than the more stable WIN7 one.
                            So, don't be in a hurry to blame the soundcard itself.
                            As a contradiction, my XP machine seems to decode H4Chat much better than the WIN7.. Go figure..
                            Hoby KK8G
                             
                            ----- Original Message -----
                            Sent: Saturday, October 26, 2013 4:13 PM
                            Subject: [V4Protocol] Sound Card Sample rate checker

                             

                             
                             
                            All,
                             
                            Here in LA.  No radio equipment so can’t participate in any over the air testing.
                             
                            I have investigated some of the decoding problems observed showing a cycling of the decode quality  which look something like this:
                             
                            image
                             
                            I am now pretty convinced this is due to some significant errors in the sound card capture or Playback rate.  The current beta software should have enough correction capability to handle up to +/- 1000 ppm (.1% ) error but the above type of distortion seen in some cases is something more like 10,000 ppm or 1% .Pretty bad for even a cheap sound card
                            I have uploaded a utility program found in the MixW software package called CheckSR.exe   to the V4Protocol FTP section.  If you download this simple utility you can check the accuracy of your sound card. 
                            I would suggest if you are seeing results like the above or others have reported seeing your signal like this then run the simple CheckSR.exe program.
                            To run select 12000 Hz sample rate (that is what is used on the H4Chat program.  Start it and let it run at least a minute or more. Stop and look at the errors reported in ppm.   If they are over 1000 the sound card is suspect.  If you see over 1000 on the capture rate try another test at 48000 and see how it compares to the first test. (normally 48000 and 12000 are more “cardinal” rates which have been shown in many tests to have the lowest errors on most sound cards).
                            There are some ways to compensate for this error in the software (the “calibration” is really just a compensation in the H4TNC DSP code) and I am evaluating some different ways.  If  you have a sound card or sound card interface that is over 1000 ppm please report it to the reflector along with the sound card type so we have some idea of what percentage of those with problems we have.
                            Just for reference: The SCS PTC II and DR7xxx DSP TNC have a very stable temperature compensated Xtal oscillator that is better than 10 ppm.  This is what is required for really high performance DSP.  Some tests I have done on the built in sound cards on the newer radios (Kenwood TS-590, Icom 7100, 7200, 7600) have been very good too...< 100 ppm.
                             
                            73,
                             
                            Rick KN6KB
                             
                          • Steve Michaels
                            Over the weekend I experimented with my external soundcard (a Soundblaster X-Fi go pro) and found that if I deleted all the Creative Sound Labs
                            Message 13 of 25 , Oct 28, 2013
                            • 0 Attachment
                              Over the weekend I experimented with my external soundcard (a Soundblaster X-Fi go pro) and found that if I deleted all the Creative Sound Labs software/drivers and started using the generic microsoft driver from 2001 ( old XP Pro computer)   all of my problems disappeared!  Before I was seeing the Decode Quality display on beacon transmit and QBF cycling up and down--now I see a solid green line with perfect decode.  CheckSR is showing (at 44100 sample rate)  -51 Input and -76 Output.

                              I discovered the possible source of the problem while researching HF ALE software PCALE,  the developer said there were numerous problems relating to driver software in his application.

                              Beaconing 28.230 for a few hours today.

                              73 de N4AY
                            • Rick Muething
                              Hoby, If you read about he utility program CheckSR.exe you will find that it uses the computer time of day clock for the reference interval. (if you are going
                              Message 14 of 25 , Oct 28, 2013
                              • 0 Attachment
                                Hoby,
                                 
                                If you read about he utility program CheckSR.exe you will find that it uses the computer time of day clock for the reference interval. (if you are going to count the number of samples per second ...actually averaged over MANY seconds...  you MUST have some absolute reference for the “second” and that is when the computer clock is used. In virtually any computer the time of day computer clock is not the same reference used for the sound card.  So the bottom line is you can expect normal variations from the same sound card across different computers due to their non perfect time-of-day reference clocks.  The instructions suggests these variations (due to computer clock) are in the range of +/- 100 ppm or less (that would be on the order of a 10 sec/day real time clock error.
                                 
                                What the CheckSR does is measure the actual sample rate of the sound card (by counting the number of samples in a given period of seconds) .  There are two major problems seen on sound cards...especially low cost ones:
                                1) All the rates  are off by approximately the same ppm. This is due to error in the sound card’s reference oscillator...Some you cheap $.25 “xtals” with poor frequency tolerance and high drift.   Errors of +/- 2000 ppm are not that uncommon.     (as a contrary example the sample rate reference in the SCS PTC II and similar DSP TNCs is on the order of +/- 10 ppm or better)
                                 
                                2) Some sound card chips use logic to approximate the sound card sample rate on some settings. On these you may find the cardinal settings like 96000, 48000 etc.  Very good (say < 200 ppm) but some of the other rates (especially 44.1KHz   11.025, 8 KHz ) to have much larger errors say +/- 2000 ppm or more. This is not due to the xtal frequency but to approximations in the reference generator mechanism. For playing a song you would probably never notice 2000 ppm error but when it comes to demodulating certain modulation types using DSP it is a BIG error and depending on the modulation scheme and number of carriers used and their spacing NOT easily corrected (e.g. with a reasonable amount of CPU resources) by software.
                                 
                                Rick KN6KB
                                 
                                 
                                From: Hoby KK8G
                                Sent: Sunday, October 27, 2013 3:46 PM
                                Subject: Re: [V4Protocol] Sound Card Sample rate checker
                                 
                                 

                                

                                After using CHECKSR for several years, I have found that the samplerate of the soundcard is determined in reference to the time base of the computer. So the card is usually very good unless the time base of the computer is poor. You can take the same soundcard and use it with two different computers and get two different results.
                                My two radio computers (one WIN7, AMD ATHLON II X4 650  GHz w/8 Gig RAM) routinely gives much better results from CHECKSR. The other (XP CELERON 440 - 2GHz w/1 Gig RAM) is not so stable. My SL on the WIN7 shows a +31 PPM Input, and + 37 PPM Output error. My XP shows the same SL with -97 PPM Input and -101 PPM Output error.
                                The clock on the Win7 machine will normally run several days with the clock varying only a very few seconds. The XP will run the same time, but the clock error will be many seconds, maybe even close to a minute.
                                This stability is shown when running M110A (MARS, not legal on Ham bands) the XP will lose sync much sooner than the more stable WIN7 one.
                                So, don't be in a hurry to blame the soundcard itself.
                                As a contradiction, my XP machine seems to decode H4Chat much better than the WIN7.. Go figure..
                                Hoby KK8G
                                 
                                ----- Original Message -----
                                Sent: Saturday, October 26, 2013 4:13 PM
                                Subject: [V4Protocol] Sound Card Sample rate checker
                                 
                                 

                                 
                                 
                                All,
                                 
                                Here in LA.  No radio equipment so can’t participate in any over the air testing.
                                 
                                I have investigated some of the decoding problems observed showing a cycling of the decode quality  which look something like this:
                                 
                                image
                                 
                                I am now pretty convinced this is due to some significant errors in the sound card capture or Playback rate.  The current beta software should have enough correction capability to handle up to +/- 1000 ppm (.1% ) error but the above type of distortion seen in some cases is something more like 10,000 ppm or 1% .Pretty bad for even a cheap sound card
                                I have uploaded a utility program found in the MixW software package called CheckSR.exe   to the V4Protocol FTP section.  If you download this simple utility you can check the accuracy of your sound card. 
                                I would suggest if you are seeing results like the above or others have reported seeing your signal like this then run the simple CheckSR.exe program.
                                To run select 12000 Hz sample rate (that is what is used on the H4Chat program.  Start it and let it run at least a minute or more. Stop and look at the errors reported in ppm.   If they are over 1000 the sound card is suspect.  If you see over 1000 on the capture rate try another test at 48000 and see how it compares to the first test. (normally 48000 and 12000 are more “cardinal” rates which have been shown in many tests to have the lowest errors on most sound cards).
                                There are some ways to compensate for this error in the software (the “calibration” is really just a compensation in the H4TNC DSP code) and I am evaluating some different ways.  If  you have a sound card or sound card interface that is over 1000 ppm please report it to the reflector along with the sound card type so we have some idea of what percentage of those with problems we have.
                                Just for reference: The SCS PTC II and DR7xxx DSP TNC have a very stable temperature compensated Xtal oscillator that is better than 10 ppm.  This is what is required for really high performance DSP.  Some tests I have done on the built in sound cards on the newer radios (Kenwood TS-590, Icom 7100, 7200, 7600) have been very good too...< 100 ppm.
                                 
                                73,
                                 
                                Rick KN6KB
                                 

                                No virus found in this message.
                                Checked by AVG - www.avg.com
                                Version: 2014.0.4158 / Virus Database: 3615/1 - Release Date: 10/07/13
                                Internal Virus Database is out of date.

                              • Rick Muething
                                Thanks for the feedback Steve. For the H4 TNC we have been testing the critical rate is 12000 samples per second. It is important to test on this rate as it
                                Message 15 of 25 , Oct 28, 2013
                                • 0 Attachment
                                  Thanks for the feedback Steve.  For the H4 TNC we have been testing the critical rate is 12000 samples per second.  It is important to test on this rate as it is not uncommon for certain rates to have large differences so it is important to run the test at the desired sample rate. 
                                   
                                  Rick KN6KB
                                   
                                   
                                  Sent: Monday, October 28, 2013 9:12 AM
                                  Subject: Re: [V4Protocol] Sound Card Sample rate checker
                                   
                                   

                                  Over the weekend I experimented with my external soundcard (a Soundblaster X-Fi go pro) and found that if I deleted all the Creative Sound Labs software/drivers and started using the generic microsoft driver from 2001 ( old XP Pro computer)   all of my problems disappeared!  Before I was seeing the Decode Quality display on beacon transmit and QBF cycling up and down--now I see a solid green line with perfect decode.  CheckSR is showing (at 44100 sample rate)  -51 Input and -76 Output.
                                   
                                  I discovered the possible source of the problem while researching HF ALE software PCALE,  the developer said there were numerous problems relating to driver software in his application.
                                   
                                  Beaconing 28.230 for a few hours today.
                                   
                                  73 de N4AY

                                  No virus found in this message.
                                  Checked by AVG - www.avg.com
                                  Version: 2014.0.4158 / Virus Database: 3615/1 - Release Date: 10/07/13
                                  Internal Virus Database is out of date.

                                • Rick Muething
                                  Ron, Not necessarily propagation...If the station sending to you for example has a playback rate of say 12024 Hz instead of 12000 (a + 2000 ppm error) you
                                  Message 16 of 25 , Oct 28, 2013
                                  • 0 Attachment

                                    Ron,
                                     
                                    Not necessarily propagation...If the station sending to you for example has a playback rate of say 12024 Hz instead of 12000 (a + 2000 ppm error) you could see poor decoding since your rate does not closely match his.
                                     
                                     
                                    I have some ideas on some more aggressive compensation that I will look at...but having correct sample rates (say +/- 500 ppm) makes a big difference.
                                     
                                    73,
                                     
                                    Rick KN6KB
                                     
                                     
                                    From: Ron Wenig
                                    Sent: Sunday, October 27, 2013 7:18 AM
                                    Subject: Re: [V4Protocol] Sound Card Sample rate checker
                                     
                                     

                                    Hi Rick,

                                    I ran the CheckSR program on my Kenwood TS-590 with its built in sound card.  The difference PPM was 0, both input and output.

                                    When I see this kind of decoding problems I just take it for granted that it's just poor propagation.

                                    73, Ron ny3j

                                    On 10/26/2013 4:13 PM, Rick Muething wrote:
                                     
                                     
                                     
                                    All,
                                     
                                    Here in LA.  No radio equipment so can’t participate in any over the air testing.
                                     
                                    I have investigated some of the decoding problems observed showing a cycling of the decode quality  which look something like this:
                                     

                                    No virus found in this message.
                                    Checked by AVG - www.avg.com
                                    Version: 2014.0.4158 / Virus Database: 3615/1 - Release Date: 10/07/13
                                    Internal Virus Database is out of date.

                                  • ric5
                                    Rick would it be in my best interest to go out and buy a high in sound card at this point? See attachment... Rick Tharrington Kd4jrx From: Rick Muething Sent:
                                    Message 17 of 25 , Oct 28, 2013
                                    Rick would it be in my best interest to go out and buy a high in sound card at this point? See attachment...
                                    Rick Tharrington Kd4jrx

                                    Sent: Monday, October 28, 2013 12:34 PM
                                    Subject: Re: [V4Protocol] Sound Card Sample rate checker

                                     

                                    Hoby,
                                     
                                    If you read about he utility program CheckSR.exe you will find that it uses the computer time of day clock for the reference interval. (if you are going to count the number of samples per second ...actually averaged over MANY seconds...  you MUST have some absolute reference for the “second” and that is when the computer clock is used. In virtually any computer the time of day computer clock is not the same reference used for the sound card.  So the bottom line is you can expect normal variations from the same sound card across different computers due to their non perfect time-of-day reference clocks.  The instructions suggests these variations (due to computer clock) are in the range of +/- 100 ppm or less (that would be on the order of a 10 sec/day real time clock error.
                                     
                                    What the CheckSR does is measure the actual sample rate of the sound card (by counting the number of samples in a given period of seconds) .  There are two major problems seen on sound cards...especially low cost ones:
                                    1) All the rates  are off by approximately the same ppm. This is due to error in the sound card’s reference oscillator...Some you cheap $.25 “xtals” with poor frequency tolerance and high drift.   Errors of +/- 2000 ppm are not that uncommon.     (as a contrary example the sample rate reference in the SCS PTC II and similar DSP TNCs is on the order of +/- 10 ppm or better)
                                     
                                    2) Some sound card chips use logic to approximate the sound card sample rate on some settings. On these you may find the cardinal settings like 96000, 48000 etc.  Very good (say < 200 ppm) but some of the other rates (especially 44.1KHz   11.025, 8 KHz ) to have much larger errors say +/- 2000 ppm or more. This is not due to the xtal frequency but to approximations in the reference generator mechanism. For playing a song you would probably never notice 2000 ppm error but when it comes to demodulating certain modulation types using DSP it is a BIG error and depending on the modulation scheme and number of carriers used and their spacing NOT easily corrected (e.g. with a reasonable amount of CPU resources) by software.
                                     
                                    Rick KN6KB
                                     
                                     
                                    From: Hoby KK8G
                                    Sent: Sunday, October 27, 2013 3:46 PM
                                    Subject: Re: [V4Protocol] Sound Card Sample rate checker
                                     
                                     

                                    

                                    After using CHECKSR for several years, I have found that the samplerate of the soundcard is determined in reference to the time base of the computer. So the card is usually very good unless the time base of the computer is poor. You can take the same soundcard and use it with two different computers and get two different results.
                                    My two radio computers (one WIN7, AMD ATHLON II X4 650  GHz w/8 Gig RAM) routinely gives much better results from CHECKSR. The other (XP CELERON 440 - 2GHz w/1 Gig RAM) is not so stable. My SL on the WIN7 shows a +31 PPM Input, and + 37 PPM Output error. My XP shows the same SL with -97 PPM Input and -101 PPM Output error.
                                    The clock on the Win7 machine will normally run several days with the clock varying only a very few seconds. The XP will run the same time, but the clock error will be many seconds, maybe even close to a minute.
                                    This stability is shown when running M110A (MARS, not legal on Ham bands) the XP will lose sync much sooner than the more stable WIN7 one.
                                    So, don't be in a hurry to blame the soundcard itself.
                                    As a contradiction, my XP machine seems to decode H4Chat much better than the WIN7.. Go figure..
                                    Hoby KK8G
                                     
                                    ----- Original Message -----
                                    Sent: Saturday, October 26, 2013 4:13 PM
                                    Subject: [V4Protocol] Sound Card Sample rate checker
                                     
                                     

                                     
                                     
                                    All,
                                     
                                    Here in LA.  No radio equipment so can’t participate in any over the air testing.
                                     
                                    I have investigated some of the decoding problems observed showing a cycling of the decode quality  which look something like this:
                                     
                                    image
                                     
                                    I am now pretty convinced this is due to some significant errors in the sound card capture or Playback rate.  The current beta software should have enough correction capability to handle up to +/- 1000 ppm (.1% ) error but the above type of distortion seen in some cases is something more like 10,000 ppm or 1% .Pretty bad for even a cheap sound card
                                    I have uploaded a utility program found in the MixW software package called CheckSR.exe   to the V4Protocol FTP section.  If you download this simple utility you can check the accuracy of your sound card. 
                                    I would suggest if you are seeing results like the above or others have reported seeing your signal like this then run the simple CheckSR.exe program.
                                    To run select 12000 Hz sample rate (that is what is used on the H4Chat program.  Start it and let it run at least a minute or more. Stop and look at the errors reported in ppm.   If they are over 1000 the sound card is suspect.  If you see over 1000 on the capture rate try another test at 48000 and see how it compares to the first test. (normally 48000 and 12000 are more “cardinal” rates which have been shown in many tests to have the lowest errors on most sound cards).
                                    There are some ways to compensate for this error in the software (the “calibration” is really just a compensation in the H4TNC DSP code) and I am evaluating some different ways.  If  you have a sound card or sound card interface that is over 1000 ppm please report it to the reflector along with the sound card type so we have some idea of what percentage of those with problems we have.
                                    Just for reference: The SCS PTC II and DR7xxx DSP TNC have a very stable temperature compensated Xtal oscillator that is better than 10 ppm.  This is what is required for really high performance DSP.  Some tests I have done on the built in sound cards on the newer radios (Kenwood TS-590, Icom 7100, 7200, 7600) have been very good too...< 100 ppm.
                                     
                                    73,
                                     
                                    Rick KN6KB
                                     

                                    No virus found in this message.
                                    Checked by AVG - www.avg.com
                                    Version: 2014.0.4158 / Virus Database: 3615/1 - Release Date: 10/07/13
                                    Internal Virus Database is out of date.

                                  • Rick Muething
                                    Mike, I knew the 7100 was good but 1 or 2 ppm is in a league by itself! It really isn’t hard when you use a good sound card chip and a decent Xtal and
                                    Message 18 of 25 , Oct 28, 2013
                                    • 0 Attachment
                                      Mike,
                                       
                                      I knew the 7100 was good but 1 or 2 ppm is in a league by itself!  It really isn’t hard when you use a good sound card chip and a decent Xtal and oscillator!
                                       
                                      Rick
                                       
                                       
                                      Sent: Saturday, October 26, 2013 5:19 PM
                                      Subject: [V4Protocol] RE: Sound Card Sample rate checker
                                       
                                       

                                      Rick in answer to your query, ICOM 7100 sound card rate after 25 minutes with ratechecker is:

                                      Input diff in PPM   -2  Output diff in PPM -1

                                       

                                      Mike

                                       



                                      ---In V4Protocol@yahoogroups.com, <v4protocol@yahoogroups.com> wrote:

                                       
                                       
                                      All,
                                       
                                      Here in LA.  No radio equipment so can’t participate in any over the air testing.
                                       
                                      I have investigated some of the decoding problems observed showing a cycling of the decode quality  which look something like this:
                                       
                                      image
                                       
                                      I am now pretty convinced this is due to some significant errors in the sound card capture or Playback rate.  The current beta software should have enough correction capability to handle up to +/- 1000 ppm (.1% ) error but the above type of distortion seen in some cases is something more like 10,000 ppm or 1% .Pretty bad for even a cheap sound card
                                      I have uploaded a utility program found in the MixW software package called CheckSR.exe   to the V4Protocol FTP section.  If you download this simple utility you can check the accuracy of your sound card. 
                                      I would suggest if you are seeing results like the above or others have reported seeing your signal like this then run the simple CheckSR.exe program.
                                      To run select 12000 Hz sample rate (that is what is used on the H4Chat program.  Start it and let it run at least a minute or more. Stop and look at the errors reported in ppm.   If they are over 1000 the sound card is suspect.  If you see over 1000 on the capture rate try another test at 48000 and see how it compares to the first test. (normally 48000 and 12000 are more “cardinal” rates which have been shown in many tests to have the lowest errors on most sound cards).
                                      There are some ways to compensate for this error in the software (the “calibration” is really just a compensation in the H4TNC DSP code) and I am evaluating some different ways.  If  you have a sound card or sound card interface that is over 1000 ppm please report it to the reflector along with the sound card type so we have some idea of what percentage of those with problems we have.
                                      Just for reference: The SCS PTC II and DR7xxx DSP TNC have a very stable temperature compensated Xtal oscillator that is better than 10 ppm.  This is what is required for really high performance DSP.  Some tests I have done on the built in sound cards on the newer radios (Kenwood TS-590, Icom 7100, 7200, 7600) have been very good too...< 100 ppm.
                                       
                                      73,
                                       
                                      Rick KN6KB
                                       

                                      No virus found in this message.
                                      Checked by AVG - www.avg.com
                                      Version: 2014.0.4158 / Virus Database: 3615/1 - Release Date: 10/07/13
                                      Internal Virus Database is out of date.

                                    • Rick Muething
                                      Doug, To get the numbers to be stable you have to let the CheckSR run a long time (a minute or more). Remember it is averaging results taken over many trials.
                                      Message 19 of 25 , Oct 28, 2013
                                      • 0 Attachment
                                        Doug,
                                         
                                        To get the numbers to be stable you have to let the CheckSR run a long time (a minute or more).  Remember it is averaging results taken over many trials.
                                         
                                        Also remember it is not just YOUR sample rates but he combination of  the playback and sample rates of the two stations. If your station is < 200 ppm capture rate error but the remote station has 5000 ppm playback error you will get the cyclic decoding issues observed. I am working on some code that will minimize this.  But IMHO trying to play around with any sound card mode with a sound card with > 1000 ppm error is like threading a needle wearing boxing gloves!  There are good $10 solutions out there if one looks and tests with the CheckSR.exe program.
                                         
                                        73,

                                        Rick KN6KB
                                         
                                         
                                        Sent: Sunday, October 27, 2013 2:50 PM
                                        Subject: [V4Protocol] RE: Sound Card Sample rate checker
                                         
                                         

                                        Hi Rick,

                                         

                                        I used the CheckSR.exe against the internal USB soundcard in my Icom 7600 and the external SignaLink USB. The numbers just jumped all over, but neither was over 144. Then, without making any changes I checked both with fldigi using its WWV check. Using an adjustment of 0 for both sound inputs produced a straight line. I hope that small contribution will help.

                                         

                                        I am, however, still puzzled by the very obvious dips in output when running FEC. This is not evident with the FSKID, just FEC text and the beacon.

                                         

                                        Doug, N3JXB



                                        ---In V4Protocol@yahoogroups.com, <v4protocol@yahoogroups.com> wrote:

                                         
                                         
                                        All,
                                         
                                        Here in LA.  No radio equipment so can’t participate in any over the air testing.
                                         
                                        I have investigated some of the decoding problems observed showing a cycling of the decode quality  which look something like this:
                                         
                                        image
                                         
                                        I am now pretty convinced this is due to some significant errors in the sound card capture or Playback rate.  The current beta software should have enough correction capability to handle up to +/- 1000 ppm (.1% ) error but the above type of distortion seen in some cases is something more like 10,000 ppm or 1% .Pretty bad for even a cheap sound card
                                        I have uploaded a utility program found in the MixW software package called CheckSR.exe   to the V4Protocol FTP section.  If you download this simple utility you can check the accuracy of your sound card. 
                                        I would suggest if you are seeing results like the above or others have reported seeing your signal like this then run the simple CheckSR.exe program.
                                        To run select 12000 Hz sample rate (that is what is used on the H4Chat program.  Start it and let it run at least a minute or more. Stop and look at the errors reported in ppm.   If they are over 1000 the sound card is suspect.  If you see over 1000 on the capture rate try another test at 48000 and see how it compares to the first test. (normally 48000 and 12000 are more “cardinal” rates which have been shown in many tests to have the lowest errors on most sound cards).
                                        There are some ways to compensate for this error in the software (the “calibration” is really just a compensation in the H4TNC DSP code) and I am evaluating some different ways.  If  you have a sound card or sound card interface that is over 1000 ppm please report it to the reflector along with the sound card type so we have some idea of what percentage of those with problems we have.
                                        Just for reference: The SCS PTC II and DR7xxx DSP TNC have a very stable temperature compensated Xtal oscillator that is better than 10 ppm.  This is what is required for really high performance DSP.  Some tests I have done on the built in sound cards on the newer radios (Kenwood TS-590, Icom 7100, 7200, 7600) have been very good too...< 100 ppm.
                                         
                                        73,
                                         
                                        Rick KN6KB
                                         

                                        No virus found in this message.
                                        Checked by AVG - www.avg.com
                                        Version: 2014.0.4158 / Virus Database: 3615/1 - Release Date: 10/07/13
                                        Internal Virus Database is out of date.

                                      • pc_cruiser_pa
                                        Hi Rick, I re-ran the CheckSR app against both the IC-7600 and SignaLink at 12000 HZ setting, allowing each to run for 2 minutes. The 7600 produces +30 In, +30
                                        Message 20 of 25 , Oct 28, 2013
                                        • 0 Attachment

                                          Hi Rick,


                                          I re-ran the CheckSR app against both the IC-7600 and SignaLink at 12000 HZ setting, allowing each to run for 2 minutes. The 7600 produces +30 In, +30 Out. The SignaLink produces +241 In and +120 Out. I hope that adds some useful info to solution.


                                          Doug, N3JXB




                                          ---In V4Protocol@yahoogroups.com, <v4protocol@yahoogroups.com> wrote:

                                          Doug,
                                           
                                          To get the numbers to be stable you have to let the CheckSR run a long time (a minute or more).  Remember it is averaging results taken over many trials.
                                           
                                          Also remember it is not just YOUR sample rates but he combination of  the playback and sample rates of the two stations. If your station is < 200 ppm capture rate error but the remote station has 5000 ppm playback error you will get the cyclic decoding issues observed. I am working on some code that will minimize this.  But IMHO trying to play around with any sound card mode with a sound card with > 1000 ppm error is like threading a needle wearing boxing gloves!  There are good $10 solutions out there if one looks and tests with the CheckSR.exe program.
                                           
                                          73,

                                          Rick KN6KB
                                           
                                           
                                          Sent: Sunday, October 27, 2013 2:50 PM
                                          Subject: [V4Protocol] RE: Sound Card Sample rate checker
                                           
                                           

                                          Hi Rick,

                                           

                                          I used the CheckSR.exe against the internal USB soundcard in my Icom 7600 and the external SignaLink USB. The numbers just jumped all over, but neither was over 144. Then, without making any changes I checked both with fldigi using its WWV check. Using an adjustment of 0 for both sound inputs produced a straight line. I hope that small contribution will help.

                                           

                                          I am, however, still puzzled by the very obvious dips in output when running FEC. This is not evident with the FSKID, just FEC text and the beacon.

                                           

                                          Doug, N3JXB



                                          ---In V4Protocol@yahoogroups.com, <v4protocol@yahoogroups.com> wrote:

                                           
                                           
                                          All,
                                           
                                          Here in LA.  No radio equipment so can’t participate in any over the air testing.
                                           
                                          I have investigated some of the decoding problems observed showing a cycling of the decode quality  which look something like this:
                                           
                                          image
                                           
                                          I am now pretty convinced this is due to some significant errors in the sound card capture or Playback rate.  The current beta software should have enough correction capability to handle up to +/- 1000 ppm (.1% ) error but the above type of distortion seen in some cases is something more like 10,000 ppm or 1% .Pretty bad for even a cheap sound card
                                          I have uploaded a utility program found in the MixW software package called CheckSR.exe   to the V4Protocol FTP section.  If you download this simple utility you can check the accuracy of your sound card. 
                                          I would suggest if you are seeing results like the above or others have reported seeing your signal like this then run the simple CheckSR.exe program.
                                          To run select 12000 Hz sample rate (that is what is used on the H4Chat program.  Start it and let it run at least a minute or more. Stop and look at the errors reported in ppm.   If they are over 1000 the sound card is suspect.  If you see over 1000 on the capture rate try another test at 48000 and see how it compares to the first test. (normally 48000 and 12000 are more “cardinal” rates which have been shown in many tests to have the lowest errors on most sound cards).
                                          There are some ways to compensate for this error in the software (the “calibration” is really just a compensation in the H4TNC DSP code) and I am evaluating some different ways.  If  you have a sound card or sound card interface that is over 1000 ppm please report it to the reflector along with the sound card type so we have some idea of what percentage of those with problems we have.
                                          Just for reference: The SCS PTC II and DR7xxx DSP TNC have a very stable temperature compensated Xtal oscillator that is better than 10 ppm.  This is what is required for really high performance DSP.  Some tests I have done on the built in sound cards on the newer radios (Kenwood TS-590, Icom 7100, 7200, 7600) have been very good too...< 100 ppm.
                                           
                                          73,
                                           
                                          Rick KN6KB
                                           

                                          No virus found in this message.
                                          Checked by AVG - www.avg.com
                                          Version: 2014.0.4158 / Virus Database: 3615/1 - Release Date: 10/07/13
                                          Internal Virus Database is out of date.

                                        • W6IDS
                                          Hey Rick. . . I downloaded the utility tonight and let it run for about 10 minutes on the microHAM micro KEYER II. I was busy with a chore and let it run.
                                          Message 21 of 25 , Oct 28, 2013
                                          • 0 Attachment
                                                Hey Rick. . .
                                             
                                            I downloaded the utility tonight and let it run for about 10
                                            minutes on the microHAM micro KEYER II.  I was busy with
                                            a chore and let it run. 
                                             
                                            When I returned, I found the following:
                                             
                                            Sample Rate 12,000
                                             
                                                                         Input         Output
                                             
                                            Sample rate, Hz  11999.90   11999.98
                                             
                                            Difference ppm   -8                -1
                                             
                                             
                                            Just for grins, I went ahead and ran the test at 48000
                                            and came up with these results:
                                             
                                                                           Input        Output
                                             
                                            Sample rate, Hz  48000.63    47997.71
                                             
                                            Difference ppm     13               -47
                                             
                                             
                                            Hope that helps. . .
                                             
                                            Howard W6IDS
                                            Richmond, IN
                                             
                                             
                                            Sent: Monday, October 28, 2013 9:44 PM
                                            Subject: RE: Re: [V4Protocol] RE: Sound Card Sample rate checker
                                             
                                            Hi Rick,

                                            I re-ran the CheckSR app against both the IC-7600 and SignaLink at 12000 HZ setting, allowing each to run for 2 minutes. The 7600 produces +30 In, +30 Out. The SignaLink produces +241 In and +120 Out. I hope that adds some useful info to solution.

                                            Doug, N3JXB

                                            ---In V4Protocol@yahoogroups.com, <v4protocol@yahoogroups.com> wrote:

                                            Doug,
                                             
                                            To get the numbers to be stable you have to let the CheckSR run a long time (a minute or more).  Remember it is averaging results taken over many trials.
                                             
                                            Also remember it is not just YOUR sample rates but he combination of  the playback and sample rates of the two stations. If your station is < 200 ppm capture rate error but the remote station has 5000 ppm playback error you will get the cyclic decoding issues observed. I am working on some code that will minimize this.  But IMHO trying to play around with any sound card mode with a sound card with > 1000 ppm error is like threading a needle wearing boxing gloves!  There are good $10 solutions out there if one looks and tests with the CheckSR.exe program.
                                             
                                            73,

                                            Rick KN6KB
                                             
                                            <SNIP> <SNIP>
                                          • Rick Muething
                                            Howard, Those are very good numbers and would not require any compensation in virtually any application. The problems observed were mainly from those using
                                            Message 22 of 25 , Oct 29, 2013
                                            • 0 Attachment
                                              Howard,
                                               
                                              Those are very good numbers and would not require any compensation in virtually any application. The problems observed were mainly from those using sound cards with sample rate errors of 10000 ppm ...about 1% .  With that much error correction is difficult or impossible.
                                               
                                              I am working on a scheme (now testing) that would accommodate up to about +/- 4000 ppm which is way over what any decent sound card should have.
                                               
                                              Thanks for the feedback....I suspected the microHam to be good.
                                               
                                              Rick KN6KB
                                               
                                               
                                              From: W6IDS
                                              Sent: Tuesday, October 29, 2013 1:13 AM
                                              Subject: Re: [V4Protocol] RE: Sound Card Sample rate checker
                                               
                                               

                                                  Hey Rick. . .
                                               
                                              I downloaded the utility tonight and let it run for about 10
                                              minutes on the microHAM micro KEYER II.  I was busy with
                                              a chore and let it run. 
                                               
                                              When I returned, I found the following:
                                               
                                              Sample Rate 12,000
                                               
                                                                           Input         Output
                                               
                                              Sample rate, Hz  11999.90   11999.98
                                               
                                              Difference ppm   -8                -1
                                               
                                               
                                              Just for grins, I went ahead and ran the test at 48000
                                              and came up with these results:
                                               
                                                                             Input        Output
                                               
                                              Sample rate, Hz  48000.63    47997.71
                                               
                                              Difference ppm     13               -47
                                               
                                               
                                              Hope that helps. . .
                                               
                                              Howard W6IDS
                                              Richmond, IN
                                               
                                               
                                              Sent: Monday, October 28, 2013 9:44 PM
                                              Subject: RE: Re: [V4Protocol] RE: Sound Card Sample rate checker
                                               
                                              Hi Rick,

                                              I re-ran the CheckSR app against both the IC-7600 and SignaLink at 12000 HZ setting, allowing each to run for 2 minutes. The 7600 produces +30 In, +30 Out. The SignaLink produces +241 In and +120 Out. I hope that adds some useful info to solution.

                                              Doug, N3JXB

                                              ---In V4Protocol@yahoogroups.com, <v4protocol@yahoogroups.com> wrote:

                                              Doug,
                                               
                                              To get the numbers to be stable you have to let the CheckSR run a long time (a minute or more).  Remember it is averaging results taken over many trials.
                                               
                                              Also remember it is not just YOUR sample rates but he combination of  the playback and sample rates of the two stations. If your station is < 200 ppm capture rate error but the remote station has 5000 ppm playback error you will get the cyclic decoding issues observed. I am working on some code that will minimize this.  But IMHO trying to play around with any sound card mode with a sound card with > 1000 ppm error is like threading a needle wearing boxing gloves!  There are good $10 solutions out there if one looks and tests with the CheckSR.exe program.
                                               
                                              73,

                                              Rick KN6KB
                                               
                                              <SNIP> <SNIP>

                                              No virus found in this message.
                                              Checked by AVG - www.avg.com
                                              Version: 2014.0.4158 / Virus Database: 3615/1 - Release Date: 10/07/13
                                              Internal Virus Database is out of date.

                                            • Rick Muething
                                              Doug, Both those are good numbers and easily correctable with DSP compensation. The main problem appears to be from those using sound cards with sample rate
                                              Message 23 of 25 , Oct 29, 2013
                                              • 0 Attachment
                                                Doug,
                                                 
                                                Both those are good numbers and easily correctable with DSP compensation. The main problem appears to be from those using sound cards with sample rate errors in the 10,000 ppm range...1%.  This is beyond what is really practical to do in most DSP correction schemes.   I am working on a new mechanism that will compensate up to about +/- 4000 ppm which should cover most every case.
                                                 
                                                Thanks for the feedback
                                                 
                                                Rick KN6KB
                                                 
                                                Sent: Monday, October 28, 2013 9:44 PM
                                                Subject: RE: Re: [V4Protocol] RE: Sound Card Sample rate checker
                                                 
                                                 

                                                Hi Rick,

                                                 

                                                I re-ran the CheckSR app against both the IC-7600 and SignaLink at 12000 HZ setting, allowing each to run for 2 minutes. The 7600 produces +30 In, +30 Out. The SignaLink produces +241 In and +120 Out. I hope that adds some useful info to solution.

                                                 

                                                Doug, N3JXB

                                                 



                                                ---In V4Protocol@yahoogroups.com, <v4protocol@yahoogroups.com> wrote:

                                                Doug,
                                                 
                                                To get the numbers to be stable you have to let the CheckSR run a long time (a minute or more).  Remember it is averaging results taken over many trials.
                                                 
                                                Also remember it is not just YOUR sample rates but he combination of  the playback and sample rates of the two stations. If your station is < 200 ppm capture rate error but the remote station has 5000 ppm playback error you will get the cyclic decoding issues observed. I am working on some code that will minimize this.  But IMHO trying to play around with any sound card mode with a sound card with > 1000 ppm error is like threading a needle wearing boxing gloves!  There are good $10 solutions out there if one looks and tests with the CheckSR.exe program.
                                                 
                                                73,

                                                Rick KN6KB
                                                 
                                                 
                                                Sent: Sunday, October 27, 2013 2:50 PM
                                                Subject: [V4Protocol] RE: Sound Card Sample rate checker
                                                 
                                                 

                                                Hi Rick,

                                                 

                                                I used the CheckSR.exe against the internal USB soundcard in my Icom 7600 and the external SignaLink USB. The numbers just jumped all over, but neither was over 144. Then, without making any changes I checked both with fldigi using its WWV check. Using an adjustment of 0 for both sound inputs produced a straight line. I hope that small contribution will help.

                                                 

                                                I am, however, still puzzled by the very obvious dips in output when running FEC. This is not evident with the FSKID, just FEC text and the beacon.

                                                 

                                                Doug, N3JXB



                                                ---In V4Protocol@yahoogroups.com, <v4protocol@yahoogroups.com> wrote:

                                                 
                                                 
                                                All,
                                                 
                                                Here in LA.  No radio equipment so can’t participate in any over the air testing.
                                                 
                                                I have investigated some of the decoding problems observed showing a cycling of the decode quality  which look something like this:
                                                 
                                                image
                                                 
                                                I am now pretty convinced this is due to some significant errors in the sound card capture or Playback rate.  The current beta software should have enough correction capability to handle up to +/- 1000 ppm (.1% ) error but the above type of distortion seen in some cases is something more like 10,000 ppm or 1% .Pretty bad for even a cheap sound card
                                                I have uploaded a utility program found in the MixW software package called CheckSR.exe   to the V4Protocol FTP section.  If you download this simple utility you can check the accuracy of your sound card. 
                                                I would suggest if you are seeing results like the above or others have reported seeing your signal like this then run the simple CheckSR.exe program.
                                                To run select 12000 Hz sample rate (that is what is used on the H4Chat program.  Start it and let it run at least a minute or more. Stop and look at the errors reported in ppm.   If they are over 1000 the sound card is suspect.  If you see over 1000 on the capture rate try another test at 48000 and see how it compares to the first test. (normally 48000 and 12000 are more “cardinal” rates which have been shown in many tests to have the lowest errors on most sound cards).
                                                There are some ways to compensate for this error in the software (the “calibration” is really just a compensation in the H4TNC DSP code) and I am evaluating some different ways.  If  you have a sound card or sound card interface that is over 1000 ppm please report it to the reflector along with the sound card type so we have some idea of what percentage of those with problems we have.
                                                Just for reference: The SCS PTC II and DR7xxx DSP TNC have a very stable temperature compensated Xtal oscillator that is better than 10 ppm.  This is what is required for really high performance DSP.  Some tests I have done on the built in sound cards on the newer radios (Kenwood TS-590, Icom 7100, 7200, 7600) have been very good too...< 100 ppm.
                                                 
                                                73,
                                                 
                                                Rick KN6KB
                                                 

                                                No virus found in this message.
                                                Checked by AVG - www.avg.com
                                                Version: 2014.0.4158 / Virus Database: 3615/1 - Release Date: 10/07/13
                                                Internal Virus Database is out of date.

                                                No virus found in this message.
                                                Checked by AVG - www.avg.com
                                                Version: 2014.0.4158 / Virus Database: 3615/1 - Release Date: 10/07/13
                                                Internal Virus Database is out of date.

                                              • w4wnt
                                                I have an old Dell Dimension 4500 desktop that I use for digital radio using a Signalink USB interface. I tested the sample rates at 12000 and initially the
                                                Message 24 of 25 , Oct 29, 2013
                                                • 0 Attachment
                                                  I have an old Dell Dimension 4500 desktop that I use for digital radio using a Signalink USB interface.  I tested the sample rates at 12000 and initially the in and out numbers were all over the place settling around 150-200 In and -250 to -300ppm Out.  These numbers were without any digital program running, just the Signalink on.

                                                  I switched screens to Firefox to look something up and when I went back to the checker, the numbers were in the thousands.  After a while they settled down to lower numbers.

                                                  I was also running a VOIP phone program, when I shut it off and let the checker run for a few minutes, the numbers went to In -20 to +60, Out -40 to +80.

                                                  I let the checker run for an hour and they settled to In -10 and Out -24.

                                                  Looks like for older computers, multiple running programs will affect the results.


                                                  Bill, W4WNT

                                                   



                                                  ---In v4protocol@yahoogroups.com, <rmuething@...> wrote:

                                                  Doug,
                                                   
                                                  Both those are good numbers and easily correctable with DSP compensation. The main problem appears to be from those using sound cards with sample rate errors in the 10,000 ppm range...1%.  This is beyond what is really practical to do in most DSP correction schemes.   I am working on a new mechanism that will compensate up to about +/- 4000 ppm which should cover most every case.
                                                   
                                                  Thanks for the feedback
                                                   
                                                  Rick KN6KB
                                                   
                                                  Sent: Monday, October 28, 2013 9:44 PM
                                                  Subject: RE: Re: [V4Protocol] RE: Sound Card Sample rate checker
                                                   
                                                   

                                                  Hi Rick,

                                                   

                                                  I re-ran the CheckSR app against both the IC-7600 and SignaLink at 12000 HZ setting, allowing each to run for 2 minutes. The 7600 produces +30 In, +30 Out. The SignaLink produces +241 In and +120 Out. I hope that adds some useful info to solution.

                                                   

                                                  Doug, N3JXB

                                                   



                                                  ---In V4Protocol@yahoogroups.com, <v4protocol@yahoogroups.com> wrote:

                                                  Doug,
                                                   
                                                  To get the numbers to be stable you have to let the CheckSR run a long time (a minute or more).  Remember it is averaging results taken over many trials.
                                                   
                                                  Also remember it is not just YOUR sample rates but he combination of  the playback and sample rates of the two stations. If your station is < 200 ppm capture rate error but the remote station has 5000 ppm playback error you will get the cyclic decoding issues observed. I am working on some code that will minimize this.  But IMHO trying to play around with any sound card mode with a sound card with > 1000 ppm error is like threading a needle wearing boxing gloves!  There are good $10 solutions out there if one looks and tests with the CheckSR.exe program.
                                                   
                                                  73,

                                                  Rick KN6KB
                                                   
                                                   
                                                  Sent: Sunday, October 27, 2013 2:50 PM
                                                  Subject: [V4Protocol] RE: Sound Card Sample rate checker
                                                   
                                                   

                                                  Hi Rick,

                                                   

                                                  I used the CheckSR.exe against the internal USB soundcard in my Icom 7600 and the external SignaLink USB. The numbers just jumped all over, but neither was over 144. Then, without making any changes I checked both with fldigi using its WWV check. Using an adjustment of 0 for both sound inputs produced a straight line. I hope that small contribution will help.

                                                   

                                                  I am, however, still puzzled by the very obvious dips in output when running FEC. This is not evident with the FSKID, just FEC text and the beacon.

                                                   

                                                  Doug, N3JXB



                                                  ---In V4Protocol@yahoogroups.com, <v4protocol@yahoogroups.com> wrote:

                                                   
                                                   
                                                  All,
                                                   
                                                  Here in LA.  No radio equipment so can’t participate in any over the air testing.
                                                   
                                                  I have investigated some of the decoding problems observed showing a cycling of the decode quality  which look something like this:
                                                   
                                                  image
                                                   
                                                  I am now pretty convinced this is due to some significant errors in the sound card capture or Playback rate.  The current beta software should have enough correction capability to handle up to +/- 1000 ppm (.1% ) error but the above type of distortion seen in some cases is something more like 10,000 ppm or 1% .Pretty bad for even a cheap sound card
                                                  I have uploaded a utility program found in the MixW software package called CheckSR.exe   to the V4Protocol FTP section.  If you download this simple utility you can check the accuracy of your sound card. 
                                                  I would suggest if you are seeing results like the above or others have reported seeing your signal like this then run the simple CheckSR.exe program.
                                                  To run select 12000 Hz sample rate (that is what is used on the H4Chat program.  Start it and let it run at least a minute or more. Stop and look at the errors reported in ppm.   If they are over 1000 the sound card is suspect.  If you see over 1000 on the capture rate try another test at 48000 and see how it compares to the first test. (normally 48000 and 12000 are more “cardinal” rates which have been shown in many tests to have the lowest errors on most sound cards).
                                                  There are some ways to compensate for this error in the software (the “calibration” is really just a compensation in the H4TNC DSP code) and I am evaluating some different ways.  If  you have a sound card or sound card interface that is over 1000 ppm please report it to the reflector along with the sound card type so we have some idea of what percentage of those with problems we have.
                                                  Just for reference: The SCS PTC II and DR7xxx DSP TNC have a very stable temperature compensated Xtal oscillator that is better than 10 ppm.  This is what is required for really high performance DSP.  Some tests I have done on the built in sound cards on the newer radios (Kenwood TS-590, Icom 7100, 7200, 7600) have been very good too...< 100 ppm.
                                                   
                                                  73,
                                                   
                                                  Rick KN6KB
                                                   

                                                  No virus found in this message.
                                                  Checked by AVG - www.avg.com
                                                  Version: 2014.0.4158 / Virus Database: 3615/1 - Release Date: 10/07/13
                                                  Internal Virus Database is out of date.

                                                  No virus found in this message.
                                                  Checked by AVG - www.avg.com
                                                  Version: 2014.0.4158 / Virus Database: 3615/1 - Release Date: 10/07/13
                                                  Internal Virus Database is out of date.

                                                • Rick Muething
                                                  Bill, It is important to remember that the CheckSR is a measurement program and by nature is not super precise. ...Consider this: 100 ppm is an error of .01%
                                                  Message 25 of 25 , Oct 29, 2013
                                                  • 0 Attachment
                                                    Bill,
                                                     
                                                    It is important to remember that the CheckSR is a measurement program and by nature is not super precise. ...Consider this: 100 ppm is an error of .01%  pretty small. To make that precise of measurement using the computers Time of Day clock requires two things:
                                                    1) That the time of day clock itself  is precise ...most computer clocks aren’t better than about 100 ppm or so (about 10 seconds per day). So your actual measured value using the TOD clock can never be better than the reference.
                                                    2) That the interval needs be fairly long (or the number of repeated measurements to average large) to get any kind of accuracy.  That is why the CheckSR values jump around a lot until it gets many numbers of samples to average.   It takes at least  about 60 seconds of operation to get fairly stable values.  If you let it run long the number will converge on the actual measurement value BUT it can never be more accurate than the reference time base which is the computers time of day clock.
                                                     
                                                    Running multiple programs at the same time has some affect on the ability to accurately capture timing details due to OS latency issues.  The multiple programs aren’t really changing the sample rate...they are just slightly changing the measurement (by the CheckSR and computer time of day clock).
                                                     
                                                    In any event if your sound card measures less than 1000 ppm error  it will certainly work with the DSP and Compensation code in the H4 TNC.  I  am working (still needs some testing) on a sound card compensation technique (does not require any “calibration” to accommodate +/- 4000 ppm error...If it is worse than that you need a new sound card!
                                                     
                                                    73,
                                                     
                                                    Rick KN6KB
                                                     
                                                    From: w4wnt@...
                                                    Sent: Tuesday, October 29, 2013 2:41 PM
                                                    Subject: [V4Protocol] RE: 28230 Print
                                                     
                                                     

                                                    I have an old Dell Dimension 4500 desktop that I use for digital radio using a Signalink USB interface.  I tested the sample rates at 12000 and initially the in and out numbers were all over the place settling around 150-200 In and -250 to -300ppm Out.  These numbers were without any digital program running, just the Signalink on.

                                                    I switched screens to Firefox to look something up and when I went back to the checker, the numbers were in the thousands.  After a while they settled down to lower numbers.

                                                    I was also running a VOIP phone program, when I shut it off and let the checker run for a few minutes, the numbers went to In -20 to +60, Out -40 to +80.

                                                    I let the checker run for an hour and they settled to In -10 and Out -24.

                                                    Looks like for older computers, multiple running programs will affect the results.


                                                    Bill, W4WNT

                                                     



                                                    ---In v4protocol@yahoogroups.com, <rmuething@...> wrote:

                                                    Doug,
                                                     
                                                    Both those are good numbers and easily correctable with DSP compensation. The main problem appears to be from those using sound cards with sample rate errors in the 10,000 ppm range...1%.  This is beyond what is really practical to do in most DSP correction schemes.   I am working on a new mechanism that will compensate up to about +/- 4000 ppm which should cover most every case.
                                                     
                                                    Thanks for the feedback
                                                     
                                                    Rick KN6KB
                                                     
                                                    Sent: Monday, October 28, 2013 9:44 PM
                                                    Subject: RE: Re: [V4Protocol] RE: Sound Card Sample rate checker
                                                     
                                                     

                                                    Hi Rick,

                                                     

                                                    I re-ran the CheckSR app against both the IC-7600 and SignaLink at 12000 HZ setting, allowing each to run for 2 minutes. The 7600 produces +30 In, +30 Out. The SignaLink produces +241 In and +120 Out. I hope that adds some useful info to solution.

                                                     

                                                    Doug, N3JXB

                                                     



                                                    ---In V4Protocol@yahoogroups.com, <v4protocol@yahoogroups.com> wrote:

                                                    Doug,
                                                     
                                                    To get the numbers to be stable you have to let the CheckSR run a long time (a minute or more).  Remember it is averaging results taken over many trials.
                                                     
                                                    Also remember it is not just YOUR sample rates but he combination of  the playback and sample rates of the two stations. If your station is < 200 ppm capture rate error but the remote station has 5000 ppm playback error you will get the cyclic decoding issues observed. I am working on some code that will minimize this.  But IMHO trying to play around with any sound card mode with a sound card with > 1000 ppm error is like threading a needle wearing boxing gloves!  There are good $10 solutions out there if one looks and tests with the CheckSR.exe program.
                                                     
                                                    73,

                                                    Rick KN6KB
                                                     
                                                     
                                                    Sent: Sunday, October 27, 2013 2:50 PM
                                                    Subject: [V4Protocol] RE: Sound Card Sample rate checker
                                                     
                                                     

                                                    Hi Rick,

                                                     

                                                    I used the CheckSR.exe against the internal USB soundcard in my Icom 7600 and the external SignaLink USB. The numbers just jumped all over, but neither was over 144. Then, without making any changes I checked both with fldigi using its WWV check. Using an adjustment of 0 for both sound inputs produced a straight line. I hope that small contribution will help.

                                                     

                                                    I am, however, still puzzled by the very obvious dips in output when running FEC. This is not evident with the FSKID, just FEC text and the beacon.

                                                     

                                                    Doug, N3JXB



                                                    ---In V4Protocol@yahoogroups.com, <v4protocol@yahoogroups.com> wrote:

                                                     
                                                     
                                                    All,
                                                     
                                                    Here in LA.  No radio equipment so can’t participate in any over the air testing.
                                                     
                                                    I have investigated some of the decoding problems observed showing a cycling of the decode quality  which look something like this:
                                                     
                                                    image
                                                     
                                                    I am now pretty convinced this is due to some significant errors in the sound card capture or Playback rate.  The current beta software should have enough correction capability to handle up to +/- 1000 ppm (.1% ) error but the above type of distortion seen in some cases is something more like 10,000 ppm or 1% .Pretty bad for even a cheap sound card
                                                    I have uploaded a utility program found in the MixW software package called CheckSR.exe   to the V4Protocol FTP section.  If you download this simple utility you can check the accuracy of your sound card. 
                                                    I would suggest if you are seeing results like the above or others have reported seeing your signal like this then run the simple CheckSR.exe program.
                                                    To run select 12000 Hz sample rate (that is what is used on the H4Chat program.  Start it and let it run at least a minute or more. Stop and look at the errors reported in ppm.   If they are over 1000 the sound card is suspect.  If you see over 1000 on the capture rate try another test at 48000 and see how it compares to the first test. (normally 48000 and 12000 are more “cardinal” rates which have been shown in many tests to have the lowest errors on most sound cards).
                                                    There are some ways to compensate for this error in the software (the “calibration” is really just a compensation in the H4TNC DSP code) and I am evaluating some different ways.  If  you have a sound card or sound card interface that is over 1000 ppm please report it to the reflector along with the sound card type so we have some idea of what percentage of those with problems we have.
                                                    Just for reference: The SCS PTC II and DR7xxx DSP TNC have a very stable temperature compensated Xtal oscillator that is better than 10 ppm.  This is what is required for really high performance DSP.  Some tests I have done on the built in sound cards on the newer radios (Kenwood TS-590, Icom 7100, 7200, 7600) have been very good too...< 100 ppm.
                                                     
                                                    73,
                                                     
                                                    Rick KN6KB
                                                     

                                                    No virus found in this message.
                                                    Checked by AVG - www.avg.com
                                                    Version: 2014.0.4158 / Virus Database: 3615/1 - Release Date: 10/07/13
                                                    Internal Virus Database is out of date.

                                                    No virus found in this message.
                                                    Checked by AVG - www.avg.com
                                                    Version: 2014.0.4158 / Virus Database: 3615/1 - Release Date: 10/07/13
                                                    Internal Virus Database is out of date.

                                                    No virus found in this message.
                                                    Checked by AVG - www.avg.com
                                                    Version: 2014.0.4158 / Virus Database: 3615/1 - Release Date: 10/07/13
                                                    Internal Virus Database is out of date.

                                                  Your message has been successfully submitted and would be delivered to recipients shortly.