The first and easiest distinction is voicedness. Make a ‘t’ sound. Not ‘tee’, but ‘t’. You do this by placing the tip of your tongue somewhere in between your teeth and the back of your alveolar ridge (the bumpy thing behind the teeth), blocking airflow, and then release. Now make a ‘d’ sound. What is the difference? If you put your hand on your throat while you make these sounds, it will be readily apparent. In making a ‘d’ sound, your throat vibrates, whereas in making a ‘t’ sound, it does not. This quality, voicedness, is not an on or off feature. However, to the human (and in fact many other animals) ear, it seems to be interpretted as such.
There is another ‘what if’ question that could be asked. What would a language look like with tripartite categorical distinctions, rather than a binary system? One could imagine a species with two sets of vocal cords. This species would have the possible options of unvoiced, voiced by either of the two, or voiced by both, creating a 4-part categorical distinction, rather than our binary one