if you're using 100% sample replacement with a snare that sounds good as is, just comp, then eq, and you're done. if using real drums, gate, low-pass, then do the same. mix the drums first, make a reverb send and send snare+toms to it, then mix the rest, THEN master. people start by putting a limiter on their master bus before they even begin mixing these days. i used to do that. really hard to make stuff sit, guess people do it so the mb fader wont hit the red, but it's harder to make stuff work that way. first of all, you are clipping the snare to make it punch through, you are chopping of the transient peaks off to make the track percievably louder, but you are losing detail. and if the signal to the master is in the red, your mb plugins wont work the way they're intended. that's why people hype about analog sounding so much better, and the digital being harsh. because the gain staging is different. analog operates up to 0 VU, which is like around -20 digital dBFS, before it distorts, and it distorts rather than clipping, so it's still musical. people push their recorded tracks up to -3 to 0 within their DAW, pushing the preamps and converters making it sound boxy and 2D too.
guys like joey sturgis use Gclip over compression because they only work with sampled drums that totally lack dynamic variation for most part, so he just uses it in combination with a transient designer to get the tracks up to par because it is thresholdless. and everyone else is like "HEY, with Gclip i can make the snare audible without making proper headroom for it", but it doesn't work the same way. even joey, uncrowned witchmaster of the digital realm, once wrote that without proper gain staging it's really hard to make the mixes sit, so he uses gain reduction on group channels before the mb (in the 32-bit float realm) so that his comp and stuff can work as intended. lots of people hate doing this, so they go for analog summing instead. also a great solution.
