Re: Resource issues when trying to record 2 FM broadcast
Trying screen capture. This email will contain the zans FM demod version of receiving 2 broadcast radio stations. based on the project referenced here http://zansprojects.blogspot.com/2019/05/stereo-fm-receiving-with-rtl-sdr-and_23.html https://forums.radioreference.com/threads/multiple-fm-station-receiver.341340/#lg=_xfUid-2-1556650274&slide=0 At 1024Mhz sample rate I was getting about 89% on my 4 core I5. If I went much higher both the wav file and the audio out would glitch. The other message will contain the Marcus Leech version FM_Demod_heir.pdf Description: Adobe PDF document TwoFMDemod.pdf Description: Adobe PDF document
Re: Resource issues when trying to record 2 FM broadcast
Here is the screen capture of the Marcus Leech version of the two FM demodulate as described here https://wiki.gnuradio.org/index.php?title=Frequency_Xlating_FIR_Filter https://github.com/patchvonbraun/wfm_stuff With a samplerate of 2.4M it uses about 60% of the 4 core I5 I am using for this project If anybody wants the grc pm me. The other email contains the Zans version of stereo demod. 2fmtopblock.pdf Description: Adobe PDF document wfm_stereo-heir.pdf Description: Adobe PDF document
Re: Resource issues when trying to record 2 FM broadcast
On 20/05/2023 09:12, Martin Lefkowitz wrote: Here is the screen capture of the Marcus Leech version of the two FM demodulate as described here https://wiki.gnuradio.org/index.php?title=Frequency_Xlating_FIR_Filter https://github.com/patchvonbraun/wfm_stuff With a samplerate of 2.4M it uses about 60% of the 4 core I5 I am using for this project If anybody wants the grc pm me. The other email contains the Zans version of stereo demod. Your taps for the freq-xlating filters aren't properly specified. They must be the output of one of the taps-generator functions--use the Low Pass Filter Taps block. Further, your desired parameters will lead to a very large filter. This is mostly controlled by the transition width parameter, and as that parameter becomes a smaller fraction of the sample-rate, the filter grows non-linearly. Consider a "lazier" transition width, and use an FFT filter instead--the FFT filter structure is more efficient once you have more than perhaps 30-50 taps. I haven't touched the "wfm_stuff" code in a long long time, so no guarantees about anything--including performance.
