i'll throw in my own opinion here...and it is that, just my
opinion.
if you're trying to control something precise, eg the snare drum is too "snappy" or "ring-y", then you'll want to use a high Q. A high Q value means that the frequency-bandwidth affected by the EQ is narrow.
If your track sounds too bright, or not bright enough, you'll want to use a low Q. A low Q means that the frequency-bandwidth affect by the EQ will be wide.
If you're at the stage of mastering and finding you need to do a lot of high-Q Eq'ing, then i suggest you go back and remix the track, as remember when pulling down that 2.3Khz for the ringing in the snare that you'll also pull down 2.3Khz for
everything else, which potentially means you'll make other musical elements, which sound fine in that frequency, sounds quite weird.
Doing low-Q Eq'ing during mastering is normal..but be wary of any amount of high-Q altering.
LemonTree said:
Q = width of the bell curve in simple terms
i hate to be nit-picking, but the Q is actually equal to the center frequency divided by the bandwidth 3dB down(or up) from how much you're boosting (or cutting). this is important to remember, as this means that the bandwidth will actually vary depending on how much you boost or cut, and is NOT constant!! despite the fact that your Q value is remaining the same. allow me to expand:
Q= center freq/(freq 3dB down)
let's imagine you want to boost 1.3Khz by 4dB with a high Q value, let's say a Q of 4.5. your affected frequency-bandwidth is therefore going to be 4.5=1300/(bandwidth 3dB down)
therefore bandwidth=1300/4.5 = ~288hz
now let's assume that the 4dB isn't enough, and that you want say 7dB of gain. let's calculate the freq bandwidth again...
bw=1300/7= ~185hz !!!!!
which is quite a difference than our original value...
just something to keep in mind while EQ'ing!
hope this was helpful...
MD