The effects of the manipulation of the sound produced while biting into apple samples, a non-dry food, was investigated. In Experiment 1, participants rated the perceived crispness of flesh cylinders obtained from three apple cultivars differing in their texture profile: ‘Renetta’ (white ‘Renetta Canada’), ‘Golden’ (‘Golden Delicious’), and ‘Fuji’. Participants might hear the veridical sounds they made when biting into an apple cylinder without any frequency adjustment (0 dB filter) or with high frequencies attenuated (either by −12 dB or by −24 dB). Perceived crispness was significantly lower when any of the reductions were applied than when no filter (0 dB) was used. In Experiment 2, new participants rated both crispness and hardness of ‘Renetta’ and ‘Fuji’ cylinders. The sound of the bite could be unfiltered (0 dB), reduced in its high frequencies (−24 dB), or globally reduced (the microphone was switched off). Crispness, again, was perceived as significantly lower with any of the sound reductions. Interestingly, perceived hardness was significantly affected by the sound information as well: Hardness was rated as being significantly lower when a global sound reduction was applied than when the sound was unfiltered. We demonstrated, for the first time, that sound information plays an important role even for the evaluation of hardness, a property believed to be primarily oral/mechanical.