For the record, clipping occurs at 0.0 in the digital realm specifically. Since we're working with bits, every sample must have a value between
0000 0000 0000 0000
and
1111 1111 1111 1111
(assuming we're working at 16-bit resolution). 0.0 dB means all the bits are on (all 1s), which is the loudest sample that can be represented digitally. (note: I'm FAIRLY certain this is how it works, but this is really just my taking what I know to a logical conclusion - I'm open to corrections from those better-versed in the technical details).
In the analog realm, things work differently, and clipping occurs somewhere above 0 - I believe it varies from one piece of equipment to another.
Porter, do you really mean -30, or -3.0? For the whole mix, or for individual tracks? If -30 on the mix, aren't you losing a lot of resolution that way?