I tried to estimate angular acceleration of a shaft with angular velocity points collected from an incremental encoder but, when I differenciate numerically, errors of estimation became very high due to the delay in capturing flanks (detecting slits). What alternatives do you suggest for my...