In phonetics, [t] is used to transcribe a "voiceless" consonant phone and [d] is used to transcribe a "voiced" consonant phone.
For English speakers, the difference between the pronunciation of the phonemes /t/ and /d/ at the end of a word is not just a matter of phonetic voicing of the consonant ([t] vs. [d]). The difference between /t/ and /d/ shows up in the length of the preceding vowel. Vowels in syllables ending in phonemically voiceless consonants are shorter than vowels in syllables ending in phonemically voiced consonants. So e.g. "believed" is pronounced with a longer [i] sound than the word "briefed".
Because of effects like this, English speakers hear a difference between words ending in /d/ and words ending in /t/, even when words ending in /d/ are pronounced with a phonetically voiceless consonant sound that you might hear as [t].
A related question: Are "whores" and "horse" homophones?