Choice of low-pass filter influences practical interpretation of ball kicking motions: the effect of a time-frequency filter method

Defining movement strategies in soccer instep kicking using the relationship between pelvis and kick leg rotations

“…Marker trajectories from kicking trials were exported to Visual 3D (V6, C-Motion, USA), where kicking foot and shank markers were low-pass filtered using a timefrequency, fractional Fourier filter (FrFF; Augustus et al, 2020a). The FrFF filters trajectories in consecutive Fourier domains to raise the cut-off frequency near the time of impact and derive valid kinematics during both swing and foot-to-ball contact (Augustus et al, 2020b). The cut-off frequencies were 18 Hz and between 150 -300 Hz for swing and ball contact phases, respectively.…”

Whole-body energy transfer strategies during football instep kicking: implications for training practices

“…Kinematic and ground reaction force (GRFs) data were captured at 1000 Hz using 10camera, 3D motion analysis (T40S, Vicon, UK) and a piezoelectric force platform (9287C, Kistler, UK). Reflective markers determined the position and orientation of bilateral feet, shanks and thighs, and a pelvis, lumbar and thorax in a 6 DOF model (Figure 1; Augustus et al, 2020b). Segments were rigid geometrical volumes scaled to participant height and mass (Hanavan, 1964), inertial characteristics derived from De Leva (1996) (feet, shanks, and thighs) or Pearsall et al (1996) (pelvis, lumbar and thorax), and the mass of the boots added to the feet (0.3 to 0.4 kg).…”

“…Marker trajectories and GRFs were exported to Visual 3D (V6, C-Motion, USA), where kicking foot and shank markers were low-pass filtered using a time-frequency, fractional Fourier filter (FrFF). The FrFF processes markers in consecutive Fourier domains to raise the cut-off frequency near the time of impact, retain high-frequency content owing to physical sources and derive valid kinematics during both swing and foot-to-ball contact (Augustus et al, 2020a(Augustus et al, , 2020b. The cut-off frequency was set as 18 Hz for the swing phase and ranged from 150 to 300 Hz for the contact phase (Augustus et al, 2020b).…”

“…The FrFF processes markers in consecutive Fourier domains to raise the cut-off frequency near the time of impact, retain high-frequency content owing to physical sources and derive valid kinematics during both swing and foot-to-ball contact (Augustus et al, 2020a(Augustus et al, , 2020b. The cut-off frequency was set as 18 Hz for the swing phase and ranged from 150 to 300 Hz for the contact phase (Augustus et al, 2020b). Ball markers were left unfiltered, but all other markers and GRFs were low-pass filtered using a conventional fourth-order, dual-pass Butterworth filter (18 Hz cut-off, determined by residual analysis) to remove erroneous peaks from support leg joint kinetics following SFTD (Bisseling & Hof, 2006).…”

“…Ball markers were left unfiltered, but all other markers and GRFs were low-pass filtered using a conventional fourth-order, dual-pass Butterworth filter (18 Hz cut-off, determined by residual analysis) to remove erroneous peaks from support leg joint kinetics following SFTD (Bisseling & Hof, 2006). Internal joint moments and forces were estimated using Newton-Euler inverse dynamics resolved to the joint co-ordinate system (Derrick et al, 2019) and kick leg kinetic parameters reflected for the final 10 ms before ball contact to remove artefact owing to the collision between foot and ball (Augustus et al, 2020b;Nunome et al, 2006). Foot centre of mass velocity at the instance of ball contact (vector magnitude of X, Y and Z components) and post-strike ball velocity were included as measures of overall kicking performance (Inoue et al, 2014), and ball to foot velocity ratios as an indicator of foot-to-ball impact efficiency.…”