Ambisonics is a sound reproduction technique based on the decomposition of the sound field using spherical harmonics. The truncation in the number of coefficients used to recreate the sound field leads to reproduction artifacts which depend on the frequency and the listener spatial location. In this work, the performance of three different decoding methods (Basic, Max-rE and In-Phase) has been studied and evaluated in the light of the results of experimental measurements. The latter were performed using a spherical array composed of 40 uniformly distributed loudspeakers and a translating 29-channel linear microphone array. An error analysis is presented based on the difference between the desired and synthesized sound pressure and acoustic intensity field. The results indicate that, as expected, the size of the region of accurate sound field reconstruction reduces as frequency increases, but with different trends depending on the type of decoder implemented