simtk.org

I recently received a question about the differences between the Arnold et al. model and gait 2392. I'm starting this thread to address this and any future issues transitioning between the two. From the email:
"I am attempting to use your model, but my files are built around the 2392 model. I am able to get my data to scale with your model, but the motion file is not matching up."

In the new model knee_angle_r and _l range from 0 to 120 deg, where positive is flexion. In the 2392 model the knee angle goes from 10 to -120, where flexion is negative. If there's a mismatch then the model knee_angle will get stuck at 0 deg. I made this change because most people plot knee flexion as positive in results.
To fix this problem, reverse the sign on all the values of knee_angle_l and knee_angle_r in that column of the .mot file.
You should also note that the pelvis has been adjusted in the new model, so hip angle isn't exactly the same in the two models. I recommend re-running IK on your marker data to make sure your joint angles are consistent with the model.