Ever wonder why your killer rock tune sounds big and huge while you're mixing it, but when you're finished you're shocked to discover it feels soft compared to other mixes, even though the meters are solidly lit? Think the latest whiz-bang level maximizer plug-in will fix your problem? Think again. Could very well be that your problem isn't a dynamics issue at all.
Broad statement time: the high-pass filter is your friend. Your very good friend!
I generally put a HPF on pretty much most tracks except those that really have business being represented in the LF range (bass and kick drum, perhaps some keyboard pads, and occasionally to allow the lower harmonics of a lead vocal). Most everything else gets filtered. It's a simple as opening a 1-band EQ plugin (like EQ3 that comes with every ProTools system), set it to HFP with a 6dB/octave slope (for more subtle/general use), and sweep while listening to the solo'd track, starting at the lowest frequency, sweeping upward until you hear the timbre of the track change in a marked way. Then back the frequency down a bit to restore the naturalness of the presentation, and move on to the next track (BTW sweep without looking at the screen... just use your ears. You'd be shocked by what your eyes make you hear!). There are a couple of reasons for using the HPF like this:
Firstly, low frequency (LF) content exists in tracks that one wouldn't consider particularly low in tone. ALL electric guitar tracks have (often harmful) LF content, and they would benefit from being filtered this way. Pretty much anything recorded up close with a cardioid microphone suffers similarly (as a result of proximity effect). When a track is solo'd it may sound quite normal without the filtering, but when track after track of unfiltered 'midrange' instruments begin to play together, all of this non-offensive harmonic LF information begins to stack up and 1) robs your track of overall headroom (making the mix feel quiet even when you're maxing out the meters) and 2) the presence of this unneeded/unwanted LF content actually works against the proper presentation of tracks that should play a role in the LF content of the mix.
By removing this unwarranted LF content you reduce the overall energy (amplitude) of your mix (there's a LOT of energy in low frequencies, even subtle ones). This decrease in level allows you to push the faders up overall to make use of the reclaimed headroom (which makes the mix feel louder to the listener.... of all the mistakes I have made in mixing over the years, this is probably my most egregious. It wasn't until just a few of years ago that I fully understood the value of preserving the headroom of the mix in this fashion... it really does make for a louder sensation when listening back).
Also, by removing this information you will find that the true LF tracks (kick, bass, etc.) speak clearly and 'own' the low end, so to speak. They aren't being muddied and smeared by the same frequencies represented in other tracks fighting, interacting and masking the important ones. It makes for more articulation of elements within the mix... with less effort, no less!
You will even find that there are occasions where you can get away with robbing quite a lot of LF from midrange instruments (guitars, bgv's, etc.) to the point that the track will actually sound unnaturally thin when solo'd, but, because of the presence of those attenuated frequencies in other elements within the mix, the perception to the listener is that the filtered tracks still sound quite full within the overall presentation. Again, the more LF you filter the 'louder' you can make your mix, because the less energy you have, the greater your headroom, the higher you can push your faders to reach the same level on the main L/R output meters.