Halfband filter of airspy_set_conversion_filter_int16() #experiment


Kenji Rikitake JJ1BDX
 

I was given a tip here at a tweet https://twitter.com/lambdaprog/status/1101052944974336000 and I tried to put in the tip as follows in airspy_rx:

diff --git a/airspy-tools/src/airspy_rx.c b/airspy-tools/src/airspy_rx.c
index a8ece76..5360ac3 100644
--- a/airspy-tools/src/airspy_rx.c
+++ b/airspy-tools/src/airspy_rx.c
@@ -948,6 +948,17 @@ int main(int argc, char** argv)
                return EXIT_FAILURE;
        }

+       const int16_t halfband_kernel[7] =
+         { -1103, 0, 9291, 16384, 9291, 0, -1103 };
+       result = airspy_set_conversion_filter_int16(device,
+                       halfband_kernel, 7);
+       if( result != AIRSPY_SUCCESS ) {
+               fprintf(stderr, "airspy_set_conversion_filter_int16() failed: %s (%d)\n", airspy_error_name(result), result);
+               airspy_close(device);
+               airspy_exit();
+               return EXIT_FAILURE;
+       }
+
        if (!strcmp(path,"-"))
                fd = stdout;
        else

I observed some changes in the passband behavior when I tried to view the FM broadcast band (in Japan) with SDR# as

./airspy_rx -w -p0 -f85.1 -a10000000 -t2 -g7 -d

but all I could see was a widened passband.

I'd appreciate if anyone can give me an answer about what this conversion filter does.

Regards, Kenji Rikitake


prog
 

As explained on Twitter, that's the effect of aliasing.


Kenji Rikitake JJ1BDX
 

FYI, results are here:
https://github.com/jj1bdx/airspy-fmradion

Kenji Rikitake, JJ1BDX

On Fri, Mar 1, 2019 at 11:45 PM prog <info@sdrsharp.com> wrote:

As explained on Twitter, that's the effect of aliasing.


prog
 

On Thu, Mar 7, 2019 at 08:18 AM, Kenji Rikitake wrote:
FYI, results are here:
https://github.com/jj1bdx/airspy-fmradion

Kenji Rikitake, JJ1BDX
Thanks Kenji. This architechture is recommended when implementing single channel receivers with Airspy One.


Kenji Rikitake JJ1BDX
 

For Airspy HF+ users, I've implemented a prototype as:
https://github.com/jj1bdx/airspyhf-fmradion
(The repository name is with "hf")
The prototype is working fine so far.
Many thanks to prog for all the suggestions!
73
Kenji RIkitake, JJ1BDX

On Thu, Mar 7, 2019 at 4:44 PM prog <info@sdrsharp.com> wrote:

On Thu, Mar 7, 2019 at 08:18 AM, Kenji Rikitake wrote:

FYI, results are here:
https://github.com/jj1bdx/airspy-fmradion

Kenji Rikitake, JJ1BDX

Thanks Kenji. This architechture is recommended when implementing single channel receivers with Airspy One.


Kenji Rikitake JJ1BDX
 

FYI
Airspy HF+ support is merged to
https://github.com/jj1bdx/airspy-fmradion
(v0.3.0 and later)

(airspyhf-fmradion is now in archive mode)
Kenji Rikitake, JJ1BDX

On Fri, Mar 8, 2019 at 9:59 PM Kenji Rikitake <jj1bdx@...> wrote:
For Airspy HF+ users, I've implemented a prototype as:
https://github.com/jj1bdx/airspyhf-fmradion
(The repository name is with "hf")
The prototype is working fine so far.
Many thanks to prog for all the suggestions!
73
Kenji RIkitake, JJ1BDX

On Thu, Mar 7, 2019 at 4:44 PM prog <info@...> wrote:
>
> On Thu, Mar 7, 2019 at 08:18 AM, Kenji Rikitake wrote:
>
> FYI, results are here:
> https://github.com/jj1bdx/airspy-fmradion
>
> Kenji Rikitake, JJ1BDX
>
> Thanks Kenji. This architechture is recommended when implementing single channel receivers with Airspy One.