No voice transfer between 2 LimeSDR's using .wav file

I have an FM transmitter and receiver with a .wav file for the voice. I am not able to hear the voice at the receiver, but I was able to get rid of the static.

Any advice for hearing the voice at the receiver?

You mixed WBFM and NBFM components. Rational resampler must have reversed settings in order to have identical quadrature rate. Try something like this (default audio in/out is used):

And you should add a “Low-pass Filter Taps” block. And use that for the taps in the “Rational Resampler” block. Every time you use decimation or interpolation you need to filter to remove the images.
For the example from @yt7pwr using NBFM maybe go with something like:
Cutoff Freq 5000
Transition Width 1000

In addition to what has already been said, your rational resampler in the first screenshot has an abnormally high interpolation factor (you probably meant to interpolate by 48 instead of 48k). The resampler filter will run at the sample rate after the interpolation but before decimation, so it would run at an absurdly high sample rate (2M * 48k). Because of this, I’m surprised the flowgraph runs at all.
As a general rule, try to keep the resampler interpolation and decimation factors as low as possible.

1 Like

Thank you so much for your help. I believe I have implemented your suggestions correctly. However, I am getting a burst of low static at the receiver audio sink. I am wondering if it has to do with the LPF or LPF Taps settings?

Here is updated flow graph:

(whith taps inside Rational Resampler TX output is attenuated!)
My GNU Radio is 3.7.13.5 version under Win10, tested in loop mode with LimeSDRMini and RTL_SDR as a monitor.

You could temporarily remove the filter from the “Rational Resampler” blocks and replace with “[ ]”. The images are going to be about 40dB lower than the signal level (~1000 times lower in amplitude), and see the result (if you carefully adjust your gains the unwanted signals could be below the noise floor of the device). Or change the filter characteristics to better match the signal that you want to pass through while still attenuating the the unwanted signals enough. You could make the roll off sharper (lower “Transition Width”), but then there will be more delay because additional taps are needed (which usually mean more memory and more calculations) in the FIR filter. The only reason I suggested adding a filter was to try and minimise, any unnecessary RF emissions. The default signal levels are low enough that it probably does not make much difference, but RF hygiene is a good habit to have. There is only a fixed amount of communal RF spectrum.