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

New Perseus Databases 09062008

Expand Messages
  • Willi Passmann
    Hello all, a new version of the Perseus databases is available at http://www.4shared.com/dir/5567845/166a39bd/sharing.html Included are the following updates:
    Message 1 of 27 , Jun 10, 2008
    • 0 Attachment
      Hello all,

      a new version of the Perseus databases is available at
      http://www.4shared.com/dir/5567845/166a39bd/sharing.html

      Included are the following updates:

      Niven 09062008
      DXCP 02062008


      vy 73
      Willi, DJ6JZ

      http://www.radio-portal.org/sdr.html
      SDR-Special
    • Roelof Bakker
      Hello all, PERSEUS minimum bandwidth of 50 Hz leaves still something to be desired for NDB chasers. Fortunately there is a work around: by careful adjusting
      Message 2 of 27 , Jun 11, 2008
      • 0 Attachment
        Hello all,

        PERSEUS minimum bandwidth of 50 Hz leaves still something to be desired
        for NDB chasers.
        Fortunately there is a work around: by careful adjusting the NOTCH, the
        bandwidth can be reduced to around 20 Hz.

        I have made a filter plot with SpectraPlus and a simple home made noise
        generator.
        PERSEUS was tuned to 3000 kHz and the noise input was set well below the
        maximum input.

        These measurements are not trivial.
        The program is running on 16 bit and is easily overloaded.
        Hence the useable range is not much more than 40 - 50 dB.
        But as a Dutch saying goes, better half an egg than an empty shell.
        It still provides interesting data.

        I have uploaded the filter plot to the photo section, but the size is so
        much reduced that details are suffering quite a bit. The plot can also be
        found on my webspace at:

        www.ndb.demon.nl/PERSEUS

        It should be emphasised that only the upper part of the plot represents a
        good view of the real filter.
        In other words, ultimate filter rejection is much better than displayed
        and this plot should certainly not be taken seriously below 40 dB's down
        the filter slope.

        As you can see, the notch reduces the bandwidth for the first 30 dB, which
        is good enough in many cases.
        The extra loss is negligible.

        It works for me to get rid of adjacent carriers.

        Best regards,
        Roelof Bakker, pa0rdt
      • Guy Atkins
        Hi Roelof, Thank you for sharing the filter plots. What sampling rate and span combination were you using? If I remember correctly, Nico once said that the
        Message 3 of 27 , Jun 11, 2008
        • 0 Attachment
          Hi Roelof,

          Thank you for sharing the filter plots.

          What sampling rate and span combination were you using? If I remember
          correctly, Nico once said that the shape factor (and possibly ultimate
          rejection) of Perseus filters are related to the span and sampling rate in
          use, due to the number of "FFT bins" available for the filter.

          Regards,

          Guy Atkins
          Puyallup, WA USA
          www.perseus-sdr.blogspot.com




          -----Original Message-----
          From: perseus_SDR@yahoogroups.com [mailto:perseus_SDR@yahoogroups.com]On
          Behalf Of Roelof Bakker
          Sent: Wednesday, June 11, 2008 5:57 AM
          To: perseus_SDR@yahoogroups.com
          Subject: [perseus_SDR] Narrow Filter plot


          Hello all,

          PERSEUS minimum bandwidth of 50 Hz leaves still something to be desired
          for NDB chasers.
          Fortunately there is a work around: by careful adjusting the NOTCH, the
          bandwidth can be reduced to around 20 Hz.
        • Roelof Bakker
          Hello all, A screenshot of a second set of filter plots has been uploaded to: www.ndb.demon.nl/PERSEUS These plots clearly show the difference between the
          Message 4 of 27 , Jun 11, 2008
          • 0 Attachment
            Hello all,

            A screenshot of a second set of filter plots has been uploaded to:

            www.ndb.demon.nl/PERSEUS

            These plots clearly show the difference between the settings of the
            PERSEUS "BW" pushbuttons.
            The filters were adjusted for an audio bandwith of about 2500 Hz.
            You will note that the filter slope is much steeper with a BW setting of 3
            kHz than with a BW setting of 25 kHz, though the actual bandwith is the
            same.
            But your ears will already have told you that!

            The last file contains the plot with PERSEUS BW set at 3 kHz and the
            filter width at 2000 Hz and 2400 Hz respectively. The NOTCH has also been
            engaged with a depth of 30 dB, centered around 1000 Hz.
            It can be seen that the filter slope is not affected by adjusting the
            filter width within a give PERSEUS BW setting.

            The plots were taken in mode USB.

            As mentioned before, the PERSEUS filters are really excellent, with a lot
            of flexibility build in.

            Best regards,
            Roelof Bakker, pa0rdt
          • Roelof Bakker
            Hello Guy, I did not pay attention to either span or sampling rate. I will run the tests again and see if changing these makes a difference. 73, Roelof Bakker
            Message 5 of 27 , Jun 11, 2008
            • 0 Attachment
              Hello Guy,

              I did not pay attention to either span or sampling rate.
              I will run the tests again and see if changing these makes a difference.

              73,
              Roelof Bakker
            • Guy Atkins
              Hi Roelof, It would be good to try the two extremes-- 1000 kHz sampling/800 kHz span, and 125 kHz sampling/0.4 kHz span. I *think* Nico said there are sharper
              Message 6 of 27 , Jun 11, 2008
              • 0 Attachment
                Hi Roelof,

                It would be good to try the two extremes-- 1000 kHz sampling/800 kHz span,
                and 125 kHz sampling/0.4 kHz span.

                I *think* Nico said there are sharper filters with the narrower sampling
                rates and spans, but I do not remember for sure.

                73,

                Guy


                -----Original Message-----
                From: perseus_SDR@yahoogroups.com [mailto:perseus_SDR@yahoogroups.com]On
                Behalf Of Roelof Bakker
                Sent: Wednesday, June 11, 2008 1:50 PM
                To: perseus_SDR@yahoogroups.com
                Subject: Re: [perseus_SDR] Narrow Filter plot


                Hello Guy,

                I did not pay attention to either span or sampling rate.
                I will run the tests again and see if changing these makes a difference.

                73,
                Roelof Bakker

                ------------------------------------
              • Nico iv3nwv
                Guy, the filters roll-off factor (sharpness) does not depend upon the DDC output sample rate. In front of the selectivity filters there s a configurable
                Message 7 of 27 , Jun 11, 2008
                • 0 Attachment
                  Guy,
                  the filters roll-off factor (sharpness) does not depend upon the DDC
                  output sample rate.
                  In front of the selectivity filters there's a configurable decimation
                  dsp chain which present them the data to be filtered always at the
                  same sample rate, i.e. 15,625 samples per second when you select the
                  12 kHz filter, 7,812.5 when you select the 6 kHz filter and so on.

                  The scope of this decimation chain is to reduce the filtering
                  processing power at its minimum since it would be computationally very
                  expensive to make a 3 kHz FIR filter very sharp when the sample rate
                  is much higher than required, say 125 kS/s. So far, the Perseus
                  software reduces the DDC sample rate to a value which does not violate
                  the Nyquist criterion (sample rate greater than the bandwidth of the
                  signal to be filtered) and then filters the result.
                  Doing this way and once the selectivity filters form factor is kept
                  the same, the processing power required to filter a narrow CW signal
                  is less than that required to filter an AM or WBFM signal, even if the
                  transition band of the narrow filter is much sharper than the wider one.

                  Nico
                • Guy Atkins
                  Nico, after you described why the filter roll-off does not depend on the sample rate, I remembered that you have discussed this before on the Perseus
                  Message 8 of 27 , Jun 12, 2008
                  • 0 Attachment
                    Nico, after you described why the filter roll-off does not depend on the sample rate, I remembered that you have discussed this before on the Perseus Yahoogroup. Sorry that you had to explain it again :^)
                     
                    I guess I've been locked into the mindset of how the PowerSDR software works for the Flex Radio products; DSP buffer options in PowerSDR effectively let you choose the shape factor of the filtering, with a trade-off of CPU load.
                     
                    Regards,
                     
                    Guy
                     
                     
                    -----Original Message-----
                    From: perseus_SDR@yahoogroups.com [mailto:perseus_SDR@yahoogroups.com]On Behalf Of Nico iv3nwv
                    Sent: Wednesday, June 11, 2008 4:56 PM
                    To: perseus_SDR@yahoogroups.com
                    Subject: [perseus_SDR] Re: Narrow Filter plot

                    Guy,
                    the filters roll-off factor (sharpness) does not depend upon the DDC
                    output sample rate.
                    In front of the selectivity filters there's a configurable decimation
                    dsp chain which present them the data to be filtered always at the
                    same sample rate, i.e. 15,625 samples per second when you select the
                    12 kHz filter, 7,812.5 when you select the 6 kHz filter and so on.

                    The scope of this decimation chain is to reduce the filtering
                    processing power at its minimum since it would be computationally very
                    expensive to make a 3 kHz FIR filter very sharp when the sample rate
                    is much higher than required, say 125 kS/s. So far, the Perseus
                    software reduces the DDC sample rate to a value which does not violate
                    the Nyquist criterion (sample rate greater than the bandwidth of the
                    signal to be filtered) and then filters the result.
                    Doing this way and once the selectivity filters form factor is kept
                    the same, the processing power required to filter a narrow CW signal
                    is less than that required to filter an AM or WBFM signal, even if the
                    transition band of the narrow filter is much sharper than the wider one.

                    Nico

                  • Nico iv3nwv
                    ... Never mind, Guy. I do not remember if in the past I ve tried to describe some detail of the dsp performed by the Perseus software, so this was a good
                    Message 9 of 27 , Jun 12, 2008
                    • 0 Attachment
                      > Sorry that you had to explain it again :^)

                      Never mind, Guy. I do not remember if in the past I've tried to
                      describe some detail of the dsp performed by the Perseus software, so
                      this was a good occasion to do it.

                      > DSP buffer options in PowerSDR effectively let
                      > you choose the shape factor of the filtering,
                      > with a trade-off of CPU load.

                      This is a good idea and not only for the trade-off you mentioned but
                      also for another important issue which is the filter delay.
                      I've never given a look to PowerSDR but I guess that filtering is
                      performed in it by a FIR filter, like i.e. in the Perseus software.
                      Increasing the number of the taps of such filters improves the shape
                      factor, the CPU load and ALSO the filter delay.
                      Usually this is not a big concern (not for BC listeners at least) for
                      wide filters as the main source of the delay is the audio buffering
                      required by the OS, but if the filter bandwidth is very narrow the
                      filter itself can contribute significantly to the total audio delay
                      (and this is a problem for amateur radios which operate in CW).

                      In my humble opinion filter "sharpness" is often overspecified.
                      Interferences get into the demodulator in two ways. The first one is
                      commonly thought as the only source of troubles, an off frequency
                      interfering signal which is not rejected sufficiently by the
                      selectivity filter. The second one instead is due to the spectral
                      tails of an adiacent interfering signal which fall INTO the filter
                      mask. This happens more frequently than believed, both for SWL and
                      amateur radios. In this case increasing the slopes of the filter
                      serves nothing. Consider a very strong MW transmitter in the channel
                      adiacent to that you're interested in. Suppose that it is really
                      strong, say 100 dB stronger than that you're listening to. In this
                      case you can use even an infinitely long filter with a brick wall
                      response to reject everything outside the +/- 4.5 kHz frequency
                      interval you're tuned to, but you won't suppress the tails of the
                      transmitter spectrum which fall inside it (and they are likely much
                      stronger than the residual interferences which could pass from the
                      filter rejection band if the filter itself were a "real" one, with a
                      finite shape factor and complexity).

                      Anyway, I will try to add a new parameter in the perseus registry so
                      that the filter length will be configurable by the user and he can
                      play with different settings. It should be quite easy to do.

                      Nico
                    • Rob Moore
                      Hi Nice, The WinRadio G313 also lets the user specify the number of taps and hence the sharpness of the filter reponses. I ve taken advantage of this to turn
                      Message 10 of 27 , Jun 12, 2008
                      • 0 Attachment
                        Hi Nice,

                        The WinRadio G313 also lets the user specify the number of taps and hence the sharpness of the filter reponses.  I've  taken advantage of this to turn up the number of taps because as shipped, it's set to a fairly small number (55 I think) so as not to overburdon the slower CPU's out there.

                        Rob


                        Nico iv3nwv <nicopal@...> wrote:
                        > Sorry that you had to explain it again :^)

                        Never mind, Guy. I do not remember if in the past I've tried to
                        describe some detail of the dsp performed by the Perseus software, so
                        this was a good occasion to do it.

                        > DSP buffer options in PowerSDR effectively let
                        > you choose the shape factor of the filtering,
                        > with a trade-off of CPU load.

                        This is a good idea and not only for the trade-off you mentioned but
                        also for another important issue which is the filter delay.
                        I've never given a look to PowerSDR but I guess that filtering is
                        performed in it by a FIR filter, like i.e. in the Perseus software.
                        Increasing the number of the taps of such filters improves the shape
                        factor, the CPU load and ALSO the filter delay.
                        Usually this is not a big concern (not for BC listeners at least) for
                        wide filters as the main source of the delay is the audio buffering
                        required by the OS, but if the filter bandwidth is very narrow the
                        filter itself can contribute significantly to the total audio delay
                        (and this is a problem for amateur radios which operate in CW).

                        In my humble opinion filter "sharpness" is often overspecified.
                        Interferences get into the demodulator in two ways. The first one is
                        commonly thought as the only source of troubles, an off frequency
                        interfering signal which is not rejected sufficiently by the
                        selectivity filter. The second one instead is due to the spectral
                        tails of an adiacent interfering signal which fall INTO the filter
                        mask. This happens more frequently than believed, both for SWL and
                        amateur radios. In this case increasing the slopes of the filter
                        serves nothing. Consider a very strong MW transmitter in the channel
                        adiacent to that you're interested in. Suppose that it is really
                        strong, say 100 dB stronger than that you're listening to. In this
                        case you can use even an infinitely long filter with a brick wall
                        response to reject everything outside the +/- 4.5 kHz frequency
                        interval you're tuned to, but you won't suppress the tails of the
                        transmitter spectrum which fall inside it (and they are likely much
                        stronger than the residual interferences which could pass from the
                        filter rejection band if the filter itself were a "real" one, with a
                        finite shape factor and complexity).

                        Anyway, I will try to add a new parameter in the perseus registry so
                        that the filter length will be configurable by the user and he can
                        play with different settings. It should be quite easy to do.

                        Nico


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

                        Yahoo! Groups Links

                        <*> To visit your group on the web, go to:
                        http://groups.yahoo.com/group/perseus_SDR/

                        <*> Your email settings:
                        Individual Email | Traditional

                        <*> To change settings online go to:
                        http://groups.yahoo.com/group/perseus_SDR/join
                        (Yahoo! ID required)

                        <*> To change settings via email:
                        mailto:perseus_SDR-digest@yahoogroups.com
                        mailto:perseus_SDR-fullfeatured@yahoogroups.com

                        <*> To unsubscribe from this group, send an email to:
                        perseus_SDR-unsubscribe@yahoogroups.com

                        <*> Your use of Yahoo! Groups is subject to:
                        http://docs.yahoo.com/info/terms/


                      • Jurgen Bartels
                        ... Exactly ! And those spectral lines within the bw SplatAway reduces quite a lot. See http://dx.3sdesign.de/splatter_killer.htm and listen to audio clip #2
                        Message 11 of 27 , Jun 12, 2008
                        • 0 Attachment
                          > interval you're tuned to, but you won't suppress the tails of the
                          > transmitter spectrum which fall inside it (and they are likely much
                          > stronger than the residual interferences which could pass from the
                          > filter rejection band if the filter itself were a "real" one, with a

                          Exactly !
                          And those spectral lines within the bw SplatAway reduces quite a lot.
                          See http://dx.3sdesign.de/splatter_killer.htm and listen to audio clip #2





                          Jurgen Bartels Suellwarden, N. Germany
                          http://dx.3sdesign.de/tv_logs_2008.htm

                          http://nordsee-ferienwohnung-burhave.de - Vacation home at Northsea
                        • Leif Asbrink
                          Hello Nico, ... This is most probably correct always. The power amplifiers of big transmitters are not extremely linear. But there is an interesting
                          Message 12 of 27 , Jun 12, 2008
                          • 0 Attachment
                            Hello Nico,

                            > Consider a very strong MW transmitter in the channel
                            > adiacent to that you're interested in. Suppose that it is really
                            > strong, say 100 dB stronger than that you're listening to. In this
                            > case you can use even an infinitely long filter with a brick wall
                            > response to reject everything outside the +/- 4.5 kHz frequency
                            > interval you're tuned to, but you won't suppress the tails of the
                            > transmitter spectrum which fall inside it (and they are likely much
                            > stronger than the residual interferences which could pass from the
                            > filter rejection band if the filter itself were a "real" one, with a
                            > finite shape factor and complexity).
                            This is most probably correct always. The power amplifiers of
                            big transmitters are not extremely linear. But there is an interesting
                            possibillity and that is to try to model the non-linearity
                            of the strong station. Presumably one could compute the signal
                            outside the nominal bandwith of the strong station using the
                            much stronger signal within the nominal passband.

                            I think it would work for AM broadcast stations because I would
                            guess their splatter is dominated by power amplifier non-linearities.
                            I do not expect it to work for amateur transmitters since their
                            interference is typically generated by ALC action which is probably
                            much harder to compute.

                            In case someone has a recording (125 kHz would be fine) of a strong
                            station causing interference to a weak one to the extent that the
                            weak one is really difficult to copy I would be interested to
                            have a look at it in case the splatter seems to be generated by
                            the RF power amplifier.

                            73

                            Leif / SM5BSZ
                          • David Towers
                            Hi Rob As far as I am aware,the filter length is only adjustable in the WR-G303 pro demodulator version. Please advise if I am incorrect cheers dave G8SZX
                            Message 13 of 27 , Jun 12, 2008
                            • 0 Attachment
                              Hi Rob
                              As far as I am aware,the filter length is only adjustable in the WR-G303
                              pro demodulator version.
                              Please advise if I am incorrect
                              cheers
                              dave
                              G8SZX Glenfield Leicester IO92jp
                              webpage
                              www.g8szx.mediumwaveradio.org


                              Rob Moore wrote:
                              > Hi Nice,
                              >
                              > The WinRadio G313 also lets the user specify the number of taps and
                              > hence the sharpness of the filter reponses. I've taken advantage of
                              > this to turn up the number of taps because as shipped, it's set to a
                              > fairly small number (55 I think) so as not to overburdon the slower
                              > CPU's out there.
                              >
                              > Rob
                              >
                              >
                              > */Nico iv3nwv <nicopal@...>/* wrote:
                              >
                              > > Sorry that you had to explain it again :^)
                              >
                              > Never mind, Guy. I do not remember if in the past I've tried to
                              > describe some detail of the dsp performed by the Perseus software, so
                              > this was a good occasion to do it.
                              >
                              > > DSP buffer options in PowerSDR effectively let
                              > > you choose the shape factor of the filtering,
                              > > with a trade-off of CPU load.
                              >
                              > This is a good idea and not only for the trade-off you mentioned but
                              > also for another important issue which is the filter delay.
                              > I've never given a look to PowerSDR but I guess that filtering is
                              > performed in it by a FIR filter, like i.e. in the Perseus software.
                              > Increasing the number of the taps of such filters improves the shape
                              > factor, the CPU load and ALSO the filter delay.
                              > Usually this is not a big concern (not for BC listeners at least) for
                              > wide filters as the main source of the delay is the audio buffering
                              > required by the OS, but if the filter bandwidth is very narrow the
                              > filter itself can contribute significantly to the total audio delay
                              > (and this is a problem for amateur radios which operate in CW).
                              >
                              > In my humble opinion filter "sharpness" is often overspecified.
                              > Interferences get into the demodulator in two ways. The first one is
                              > commonly thought as the only source of troubles, an off frequency
                              > interfering signal which is not rejected sufficiently by the
                              > selectivity filter. The second one instead is due to the spectral
                              > tails of an adiacent interfering signal which fall INTO the filter
                              > mask. This happens more frequently than believed, both for SWL and
                              > amateur radios. In this case increasing the slopes of the filter
                              > serves nothing. Consider a very strong MW transmitter in the channel
                              > adiacent to that you're interested in. Suppose that it is really
                              > strong, say 100 dB stronger than that you're listening to. In this
                              > case you can use even an infinitely long filter with a brick wall
                              > response to reject everything outside the +/- 4.5 kHz frequency
                              > interval you're tuned to, but you won't suppress the tails of the
                              > transmitter spectrum which fall inside it (and they are likely much
                              > stronger than the residual interferences which could pass from the
                              > filter rejection band if the filter itself were a "real" one, with a
                              > finite shape factor and complexity).
                              >
                              > Anyway, I will try to add a new parameter in the perseus registry so
                              > that the filter length will be configurable by the user and he can
                              > play with different settings. It should be quite easy to do.
                              >
                              > Nico
                              >
                              >
                              > ------------------------------------
                              >
                              > Yahoo! Groups Links
                              >
                              >
                              >
                              >
                              >
                              > This message has been processed by Firetrust Benign
                              > <http://www.firetrust.com/products/benign/>.
                            • Willi Passmann
                              Hello all, a new version of the Perseus databases is available at http://www.4shared.com/dir/5567845/166a39bd/sharing.html Included are the following updates:
                              Message 14 of 27 , Jun 13, 2008
                              • 0 Attachment
                                Hello all,

                                a new version of the Perseus databases is available at
                                http://www.4shared.com/dir/5567845/166a39bd/sharing.html

                                Included are the following updates:

                                Nagoya 12062008
                                Eibi 11062008
                                New: ELWG 10062008

                                vy 73
                                Willi, DJ6JZ

                                http://www.radio-portal.org/sdr.html
                                SDR-Special
                              • Rob Moore
                                I believe you re correct. I have both but nether one was in front of me at work. Rob
                                Message 15 of 27 , Jun 13, 2008
                                • 0 Attachment
                                  I believe you're correct. I have both but nether one was in front of me
                                  at work.

                                  Rob


                                  David Towers wrote:
                                  > Hi Rob
                                  > As far as I am aware,the filter length is only adjustable in the WR-G303
                                  > pro demodulator version.
                                  > Please advise if I am incorrect
                                  > cheers
                                  > dave
                                  > G8SZX Glenfield Leicester IO92jp
                                  > webpage
                                  > www.g8szx.mediumwaveradio.org
                                  >
                                  >
                                  > Rob Moore wrote:
                                  >
                                  >> Hi Nice,
                                  >>
                                  >> The WinRadio G313 also lets the user specify the number of taps and
                                  >> hence the sharpness of the filter reponses. I've taken advantage of
                                  >> this to turn up the number of taps because as shipped, it's set to a
                                  >> fairly small number (55 I think) so as not to overburdon the slower
                                  >> CPU's out there.
                                  >>
                                  >> Rob
                                  >>
                                  >>
                                  >> */Nico iv3nwv <nicopal@...>/* wrote:
                                  >>
                                  >> > Sorry that you had to explain it again :^)
                                  >>
                                  >> Never mind, Guy. I do not remember if in the past I've tried to
                                  >> describe some detail of the dsp performed by the Perseus software, so
                                  >> this was a good occasion to do it.
                                  >>
                                  >> > DSP buffer options in PowerSDR effectively let
                                  >> > you choose the shape factor of the filtering,
                                  >> > with a trade-off of CPU load.
                                  >>
                                  >> This is a good idea and not only for the trade-off you mentioned but
                                  >> also for another important issue which is the filter delay.
                                  >> I've never given a look to PowerSDR but I guess that filtering is
                                  >> performed in it by a FIR filter, like i.e. in the Perseus software.
                                  >> Increasing the number of the taps of such filters improves the shape
                                  >> factor, the CPU load and ALSO the filter delay.
                                  >> Usually this is not a big concern (not for BC listeners at least) for
                                  >> wide filters as the main source of the delay is the audio buffering
                                  >> required by the OS, but if the filter bandwidth is very narrow the
                                  >> filter itself can contribute significantly to the total audio delay
                                  >> (and this is a problem for amateur radios which operate in CW).
                                  >>
                                  >> In my humble opinion filter "sharpness" is often overspecified.
                                  >> Interferences get into the demodulator in two ways. The first one is
                                  >> commonly thought as the only source of troubles, an off frequency
                                  >> interfering signal which is not rejected sufficiently by the
                                  >> selectivity filter. The second one instead is due to the spectral
                                  >> tails of an adiacent interfering signal which fall INTO the filter
                                  >> mask. This happens more frequently than believed, both for SWL and
                                  >> amateur radios. In this case increasing the slopes of the filter
                                  >> serves nothing. Consider a very strong MW transmitter in the channel
                                  >> adiacent to that you're interested in. Suppose that it is really
                                  >> strong, say 100 dB stronger than that you're listening to. In this
                                  >> case you can use even an infinitely long filter with a brick wall
                                  >> response to reject everything outside the +/- 4.5 kHz frequency
                                  >> interval you're tuned to, but you won't suppress the tails of the
                                  >> transmitter spectrum which fall inside it (and they are likely much
                                  >> stronger than the residual interferences which could pass from the
                                  >> filter rejection band if the filter itself were a "real" one, with a
                                  >> finite shape factor and complexity).
                                  >>
                                  >> Anyway, I will try to add a new parameter in the perseus registry so
                                  >> that the filter length will be configurable by the user and he can
                                  >> play with different settings. It should be quite easy to do.
                                  >>
                                  >> Nico
                                  >>
                                  >>
                                  >> ------------------------------------
                                  >>
                                  >> Yahoo! Groups Links
                                  >>
                                  >>
                                  >>
                                  >>
                                  >>
                                  >> This message has been processed by Firetrust Benign
                                  >> <http://www.firetrust.com/products/benign/>.
                                  >>
                                  >
                                  > ------------------------------------
                                  >
                                  > Yahoo! Groups Links
                                  >
                                  >
                                  >
                                  >
                                  >
                                • Nico iv3nwv
                                  Hi Juergen, yesterday afternoon, before I went home unusually early (yes, I like to watch the European Soccer Championship at TV from time to time), I ve added
                                  Message 16 of 27 , Jun 14, 2008
                                  • 0 Attachment
                                    Hi Juergen,
                                    yesterday afternoon, before I went home unusually early (yes, I like
                                    to watch the European Soccer Championship at TV from time to time),
                                    I've added a AM splatter rejection block to the Perseus software too.

                                    It works quite in a different manner from yours (no analysis in the
                                    frequency domain is needed) and does no spectral subtraction at all.
                                    After playing all the morning with Hilbert transforms and other
                                    amazing mathematical tools, I recognized that actually it is very
                                    simple to remove the unwanted sideband of an adiacent channel AM
                                    signal from the intended signal and that no complex filtering or
                                    algorithms are really required.
                                    The block I've added works just on one sideband of the desired signal
                                    and is placed before the selectivity filter, the AGC and the
                                    demodulator. It just remove one interfering sideband so, if both
                                    sidebands of the wanted signal are interfered, it requires that the
                                    one of them is filtered out in the usual way (i.e. dragging the
                                    selectivity filter mask to exclude it). In any case, the choice of
                                    which interferer has to be suppressed by the AM splatter rejection
                                    block is left to the user (just clicking with the mouse over its carrier).
                                    BTW, in the sw version I'm actually working, I've also added a "50
                                    kHz" BW button (I did it because I also like to decode NOAA satellite
                                    APT pictures from time to time). This is the bandwidth that should be
                                    used when working with the AM splatter killer. The "25 kHZ" BW setting
                                    is not sufficient as it introduce significant attenuation at +/- 12.5
                                    kHz and it would distort the AM signal to be killed.
                                    Anyway, the tests I've made in my lab with the aid of a signal
                                    generator which injected the undesired AM splatter are really good.
                                    I've also tested the block with a real MW record (made by Guy Atkins
                                    if I remember correctly) but I was not completely satisfied by the
                                    results (I have to further investigate why: because the interferer
                                    signal had no clean sideband? because multipath propagation destroy
                                    the phase simmetry of the interferer?).

                                    I hope to release this feature soon, maybe in a publicly available
                                    beta version release so that everybody can try it.

                                    Man at work,
                                    Nico

                                    --- In perseus_SDR@yahoogroups.com, "Jurgen Bartels" <jurgenb2@...> wrote:
                                    >
                                    > > interval you're tuned to, but you won't suppress the tails of the
                                    > > transmitter spectrum which fall inside it (and they are likely much
                                    > > stronger than the residual interferences which could pass from the
                                    > > filter rejection band if the filter itself were a "real" one, with a
                                    >
                                    > Exactly !
                                    > And those spectral lines within the bw SplatAway reduces quite a lot.
                                    > See http://dx.3sdesign.de/splatter_killer.htm and listen to audio
                                    clip #2
                                    >
                                    >
                                    >
                                    >
                                    >
                                    > Jurgen Bartels Suellwarden, N. Germany
                                    > http://dx.3sdesign.de/tv_logs_2008.htm
                                    >
                                    > http://nordsee-ferienwohnung-burhave.de - Vacation home at Northsea
                                    >
                                  • Nelu
                                    I wish it were effective in removing SSB splatter too.... 73 Corneliu ... too. ... signal ... carrier). ... satellite ... be ... setting ... 12.5 ... the ...
                                    Message 17 of 27 , Jun 14, 2008
                                    • 0 Attachment
                                      I wish it were effective in removing SSB splatter too....

                                      73 Corneliu

                                      --- In perseus_SDR@yahoogroups.com, "Nico iv3nwv" <nicopal@...> wrote:
                                      >
                                      > Hi Juergen,
                                      > yesterday afternoon, before I went home unusually early (yes, I like
                                      > to watch the European Soccer Championship at TV from time to time),
                                      > I've added a AM splatter rejection block to the Perseus software
                                      too.
                                      >
                                      > It works quite in a different manner from yours (no analysis in the
                                      > frequency domain is needed) and does no spectral subtraction at all.
                                      > After playing all the morning with Hilbert transforms and other
                                      > amazing mathematical tools, I recognized that actually it is very
                                      > simple to remove the unwanted sideband of an adiacent channel AM
                                      > signal from the intended signal and that no complex filtering or
                                      > algorithms are really required.
                                      > The block I've added works just on one sideband of the desired
                                      signal
                                      > and is placed before the selectivity filter, the AGC and the
                                      > demodulator. It just remove one interfering sideband so, if both
                                      > sidebands of the wanted signal are interfered, it requires that the
                                      > one of them is filtered out in the usual way (i.e. dragging the
                                      > selectivity filter mask to exclude it). In any case, the choice of
                                      > which interferer has to be suppressed by the AM splatter rejection
                                      > block is left to the user (just clicking with the mouse over its
                                      carrier).
                                      > BTW, in the sw version I'm actually working, I've also added a "50
                                      > kHz" BW button (I did it because I also like to decode NOAA
                                      satellite
                                      > APT pictures from time to time). This is the bandwidth that should
                                      be
                                      > used when working with the AM splatter killer. The "25 kHZ" BW
                                      setting
                                      > is not sufficient as it introduce significant attenuation at +/-
                                      12.5
                                      > kHz and it would distort the AM signal to be killed.
                                      > Anyway, the tests I've made in my lab with the aid of a signal
                                      > generator which injected the undesired AM splatter are really good.
                                      > I've also tested the block with a real MW record (made by Guy Atkins
                                      > if I remember correctly) but I was not completely satisfied by the
                                      > results (I have to further investigate why: because the interferer
                                      > signal had no clean sideband? because multipath propagation destroy
                                      > the phase simmetry of the interferer?).
                                      >
                                      > I hope to release this feature soon, maybe in a publicly available
                                      > beta version release so that everybody can try it.
                                      >
                                      > Man at work,
                                      > Nico
                                      >
                                      > --- In perseus_SDR@yahoogroups.com, "Jurgen Bartels" <jurgenb2@>
                                      wrote:
                                      > >
                                      > > > interval you're tuned to, but you won't suppress the tails of
                                      the
                                      > > > transmitter spectrum which fall inside it (and they are likely
                                      much
                                      > > > stronger than the residual interferences which could pass from
                                      the
                                      > > > filter rejection band if the filter itself were a "real" one,
                                      with a
                                      > >
                                      > > Exactly !
                                      > > And those spectral lines within the bw SplatAway reduces quite a
                                      lot.
                                      > > See http://dx.3sdesign.de/splatter_killer.htm and listen to audio
                                      > clip #2
                                      > >
                                      > >
                                      > >
                                      > >
                                      > >
                                      > > Jurgen Bartels Suellwarden, N. Germany
                                      > > http://dx.3sdesign.de/tv_logs_2008.htm
                                      > >
                                      > > http://nordsee-ferienwohnung-burhave.de - Vacation home at
                                      Northsea
                                      > >
                                      >
                                    • Roelof Bakker
                                      Hi, ... Seems to me a bit masochistic at the moment! Best regards, Roelof Bakker, pa0rdt Netherlands
                                      Message 18 of 27 , Jun 14, 2008
                                      • 0 Attachment
                                        Hi,

                                        >> (yes, I like to watch the European Soccer Championship at TV from time
                                        >> to time),

                                        Seems to me a bit masochistic at the moment!

                                        Best regards,
                                        Roelof Bakker, pa0rdt
                                        Netherlands
                                      • Nico iv3nwv
                                        ... Yes, it is :-))
                                        Message 19 of 27 , Jun 14, 2008
                                        • 0 Attachment
                                          > Seems to me a bit masochistic at the moment!

                                          Yes, it is :-))
                                        • Nico iv3nwv
                                          ... This can t be made in the software. For this purpose it is necessary a little bit of hardware to remove from his shack the old man who is overdriving his
                                          Message 20 of 27 , Jun 14, 2008
                                          • 0 Attachment
                                            Nelu wrote:

                                            > I wish it were effective in removing SSB splatter too....

                                            This can't be made in the software.
                                            For this purpose it is necessary a little bit of hardware to remove
                                            from his shack the old man who is overdriving his linear (??!?) amplifier.

                                            73,
                                            Nico
                                          • Guy Atkins
                                            I m really looking forward to this experimental feature, Nico! I will be at the Oregon USA coast next month for a six day vacation, and Perseus will be coming
                                            Message 21 of 27 , Jun 14, 2008
                                            • 0 Attachment
                                              I'm really looking forward to this experimental feature, Nico! I will
                                              be at the Oregon USA coast next month for a six day vacation, and
                                              Perseus will be coming along (oh, and my family will also be there :^)

                                              It will be a good opportunity to try the AM splatter killer on
                                              domestic MW stations that interfere with adjacent & co-channel DU DX
                                              from Australia and New Zealand during our local summer.

                                              Keep up the innovative work, Nico!

                                              Best DX,

                                              Guy Atkins
                                              Puyallup, WA USA
                                              www.perseus-sdr.blogspot.com


                                              --- In perseus_SDR@yahoogroups.com, "Nico iv3nwv" <nicopal@...> wrote:
                                              > I've added a AM splatter rejection block to the Perseus software
                                              too.
                                              >
                                              > I hope to release this feature soon, maybe in a publicly available
                                              > beta version release so that everybody can try it.

                                              <SNIP>
                                            • Walter Salmaniw
                                              ... Nico, I wish to second that! On the west coast of North America, an AM splatter rejection mechanism for trans Pacific and trans Atlantic MW reception
                                              Message 22 of 27 , Jun 14, 2008
                                              • 0 Attachment
                                                At 11:07 AM 6/14/2008, you wrote:


                                                >I'm really looking forward to this experimental feature, Nico! I will
                                                >be at the Oregon USA coast next month for a six day vacation, and
                                                >Perseus will be coming along (oh, and my family will also be there :^)
                                                >
                                                >It will be a good opportunity to try the AM splatter killer on
                                                >domestic MW stations that interfere with adjacent & co-channel DU DX
                                                >from Australia and New Zealand during our local summer.
                                                >
                                                >Keep up the innovative work, Nico!
                                                >
                                                >Best DX,
                                                >
                                                >Guy Atkins
                                                >Puyallup, WA USA
                                                >www.perseus-sdr.blogspot.com
                                                >
                                                >--- In <mailto:perseus_SDR%40yahoogroups.com>perseus_SDR@yahoogroups.com, "Nico iv3nwv" <nicopal@...> wrote:
                                                >> I've added a AM splatter rejection block to the Perseus software
                                                >too.
                                                >>
                                                >> I hope to release this feature soon, maybe in a publicly available
                                                >> beta version release so that everybody can try it.
                                                >
                                                ><SNIP>


                                                Nico, I wish to second that! On the west coast of North America, an AM splatter rejection mechanism for trans Pacific and trans Atlantic MW reception would be an absolute godsend! Thank you! Also, what about a built in timer(s), please!!!! .......Walt Salmaniw, Victoria, BC.
                                              • CESARE
                                                Hi Between a Perseus and the next one... http://iv3mur.dyndns.org/cucina.html IV3MUR Cesare
                                                Message 23 of 27 , Oct 13, 2011
                                                • 0 Attachment
                                                  Hi

                                                  Between a Perseus and the next one...


                                                  http://iv3mur.dyndns.org/cucina.html


                                                  IV3MUR Cesare
                                                • Ron
                                                  Cesare, Looks delicious! The kitchen and barbeque is also my second nature. Love prepairing slow cooked food in the winter. Sorry for this off topic reply. 73
                                                  Message 24 of 27 , Oct 13, 2011
                                                  • 0 Attachment
                                                    Cesare,

                                                    Looks delicious! The kitchen and barbeque is also my second nature. Love prepairing slow cooked food in the winter.

                                                    Sorry for this off topic reply.

                                                    73' Ron
                                                    ON2RON - DXA64
                                                  • CESARE
                                                    Hi Ron Every once in a while I really appreciate talking about good cuisine. Cesare
                                                    Message 25 of 27 , Oct 13, 2011
                                                    • 0 Attachment
                                                      Hi Ron

                                                      Every once in a while I really appreciate talking about good cuisine.

                                                      Cesare
                                                    • Dave
                                                      Interesting to see the dropping of the GD suffix Simon, Are you in England now? Dave ... From: Simon Brown To: perseus_SDR@yahoogroups.com Sent: Wednesday,
                                                      Message 26 of 27 , Mar 26, 2014
                                                      • 0 Attachment
                                                        Interesting to see the dropping of the GD suffix Simon,

                                                        Are you in England now?
                                                         
                                                        Dave
                                                        ----- Original Message -----
                                                        Sent: Wednesday, March 26, 2014 4:53 PM
                                                        Subject: RE: [perseus_SDR] F2 spectrum file 32.x MHz

                                                        Jurgen,

                                                         

                                                        An excellent recording, many thanks!

                                                         

                                                        Simon Brown G4ELI
                                                        http://v2.sdr-radio.com

                                                         

                                                      • Simon Brown
                                                        Hi, Not yet - move in May. Simon Brown G4ELI http://v2.sdr-radio.com From: perseus_SDR@yahoogroups.com [mailto:perseus_SDR@yahoogroups.com] On Behalf Of Dave
                                                        Message 27 of 27 , Mar 26, 2014
                                                        • 0 Attachment

                                                          Hi,

                                                           

                                                          Not yet – move in May.

                                                           

                                                          Simon Brown G4ELI
                                                          http://v2.sdr-radio.com

                                                           

                                                          From: perseus_SDR@yahoogroups.com [mailto:perseus_SDR@yahoogroups.com] On Behalf Of Dave
                                                          Sent: 26 March 2014 23:26
                                                          To: perseus_SDR@yahoogroups.com
                                                          Subject: [perseus_SDR] OT

                                                           

                                                           

                                                          Interesting to see the dropping of the GD suffix Simon,

                                                          Are you in England now?

                                                           

                                                          Dave

                                                          ----- Original Message -----

                                                          Sent: Wednesday, March 26, 2014 4:53 PM

                                                          Subject: RE: [perseus_SDR] F2 spectrum file 32.x MHz

                                                           

                                                          Jurgen,

                                                           

                                                          An excellent recording, many thanks!

                                                           

                                                          Simon Brown G4ELI
                                                          http://v2.sdr-radio.com

                                                           

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