sound_field_analysis example 2 - Visualizing measured data

The recording was done in an anechoic chamber using a cardioid mic, the plane wave is approximated by a speaker at Azimuth=180°, Colatitude=90° at a distance of 3m.

Read in data

array_data is a tuple as defined in io.ArraySignal, which simply bundles three more basic data types:

• .signal - is a io.TimeSignal containing:
  • .signal.signal
  • .signal.fs
• .grid - is a io.SphericalGrid containing:
  • .grid.azimuth
  • .grid.colatitude
  • .grid.radius
  • .grid.weight
• .configuration - is a io.ArrayConfiguration containing:
  • .configuration.array_radius
  • .configuration.array_type
  • .configuration.transducer_type
  • .configuration.scatter_radius
• .temperature - Temperature in °C

The function io.read_miro_struct handles this, only the transducer type has to be set to cardioid manually (default: omni), as it cannot be extraced from the miro file.

Time > Frequency > Spatial domain

The impulse responses are now read into array_data.signal. To bring the time-domain signals to the spatial fourier domain, first the normal Fourier transform (process.FFT) has to be applied.

Spatial domain & radial filters

Now the data is ready for the spatial fourier transform (process.spatFT); the resulting coefficients hold order^2 - 1 rows. The radial filters are also calculated, based on the array configuration, the sampling frequency and the number of FFT bins (NFFT) used.

Both the spatial transform and then radial filters are also order-limited to the same order.

Plot

To visualized the recorded data, the spherical fourier coefficients and the radial filters are passed to plot.makeMTX() along with a specific kr bin. The function returns the sound pressure at a resolution of 1 degree, which then can be visualized using plot.plot3D(). Several styles ('shape', 'sphere', 'flat') exist.

To plot a specific frequency, we use utils.nearest_to_value_IDX() find the index closest to the desired frequency.