We study masking of key-click feedback signals on a flat surface for ten-finger touch typing with localized tactile feedback. We hypothesize that people will attribute tactile feedback to the key being pressed, even with global tactile feedback, provided that the tactile signal on other parts of the surface is sufficiently attenuated. To this end, we measure the thresholds at which a tactile signal is barely perceptible to a finger that is resting passively on a surface while another finger actively presses on the surface and receives a key-click feedback signal. Combinations of the index and middle fingers of both hands are tested. The results indicate that the thresholds are independent of the signal amplitude on the active finger. Larger signal attenuation is needed when the index fingers of both hands are involved than when two fingers of the same hand are involved. Future research will extend the current experimental design to ten fingers and typing-based tasks.