I am noticing some differences in the data captures that I am making with the tuner’s (AD9361) ADI software when I take them in the Time domain vs when I take them in Frequency domain. If I capture the data in Time, and then do an FFT, I have a significantly higher noise floor, then if I capture the data in the frequency domain and then do my measurements directly. When plotting the data after the FFT that was captured in the Time domain the typical sigma delta ADC noise floor shaping is missing. Do you have any ideas why this might happen, or why it would be better when the ADI software provides the data in the frequency domain directly? I am using the same bin-size and spacing at the ADI provided frequency domain data, so I would expect similar results. FFT example plots are attached.

Please take a look at our source code here:

https://github.com/analogdevicesinc/iio-oscilloscope/blob/multi_plot_osc/osc.c

static void do_fft(Transform *tr)

Are you also applying a Window function?

-Michael