Re: Mids and mixing
Either additive or subtractive EQ will work.
The main considerations when EQ'ing are:
1. Gain. When you use additive EQ you are adding a significant amount of gain; you have to be careful that the EQ'd track doesn't end up louder than the original, so some trim will probably have to be applied. That being said, don't be afraid to get extreme with the EQ to make something fit. Try to avoid listening to tracks solo'd when EQ'ing; that way, madness lies...
2. Phase. When using either additive or subtractive EQ, there is usually a bump in the frequency response at the cutoff point; this can create phasing artifacts ... you will want to set the EQ point accordingly. (You could also use a 'linear phase' EQ to eliminate this from happening)
3. Q. Generally speaking, you want to use a wide Q for adding and a narrow Q for subtracting. This applies to hi/low pass filtering as well.
4. Masking. If you are having trouble getting track A to fit, you can try finding the main frequency of that track and subtract it from every other track.
5. "Stuck Fader". A sure sign that you need EQ is when you have a track which has good level but in the mix it simultaneously sounds too loud and too quiet.
6. HPF. Use high pass filtering on every track. Yes, every track.
6. Know Your Room. Use reference tracks often to make sure that what you are hearing as an EQ issue isn't your crappy room causing an issue.
As to which is one is actually needed for a mix, a cool trick is to try the 'pink noise' mixing method. If you are having trouble fitting the track onto the pink noise, you may have to add a couple of dB of EQ. Subjectively speaking, and especially with frequency masking, subtractive EQ is not necessarily the best way to get something to fit.