“…It should be noted that in our experiment only three of the four NV axes are used for reconstruction of 𝐵 and the real direction of the detected field in the laboratory frame could not be determined due to the lack of a calibration of the magnetic field. With a method akin to maximum likelihood estimation, [26,32] our numerical calculations implemented a search procedure to match the target experimental resonance frequencies ν 1,2 , ν 3,4 , and ν 5,6 according to the formula…”