Experimental raw data for density plots in Fig 2. Each .csv contains an array, where 1st row corresponds to x_axis (mechanical frequency in MHz for panels 1,2,3,4) and first column the y_axis (optical frequency in THz for panel 1, modulation frequency in MHz for panels 2,3,4). First nonzero component is the 2nd for each array. Remaining array elements contain the z values (Thermomechanical noise spectral for panel 1, Amplitude of driven responses for panels 2,3,4). An illustrative example of plotting in an ipython notebook follows: %pylab inline A= genfromtxt('Fig2_data_modVolt=0mV_experiment.csv', delimiter=',') x = A[0,1:] y = A[1:,0] z = A[1:,1:] imshow(z,aspect='auto',vmin=z.min(),vmax=z.max(),extent=[x.min(),x.max(),y.min(),y.max()],cmap='magma') Theoretical data for panel 4 in Fig 2, stored in a .csv with the same structure as previous. Raw experimental data for upper panels in Fig 3. Each .csv contains an array where 1st row corresponds to x_axis (modulation phase) and first column the y_axis (optical frequency in THz). Z values contain the experimental signal proportional to the Y optical quadrature of the transferred mode. Theoretical data for lower panels in Fig 3, stored in a .csv with the same structure as previous. Jupyter notebook to produce and plot typical data for Fig 4: phononic amplitude averaged over 100 disorder realizations, normalized to the maximum value (*extra_dependencies: Kwant Python library: https://kwant-project.org/).