The envelope follower on the Ears probably isn't quite right for the sound level and/or transient behavior. It would probably work better if you had a fully-variable threshold control instead of the jumpered presets on that module. Another idea would be to try a Time Bandit, which is a stompbox-style click-to-trigger audio extractor, provided it can get enough volume from the contact mic alone or if you have some sort of preamp for it.