Ask ten engineers how to mic a kit and you get ten mic lists. Ask them what actually made the difference on the last record they liked, and the answers converge fast: the kit was tuned, the room was chosen, and the overheads were placed carefully. Everything else is refinement.
Tuning is not an optional first step
A snare that rings for two seconds will ring for two seconds through a thousand-dollar microphone. No amount of gating fixes a badly tuned tom; it only makes the ringing start and stop abruptly. Before anything gets patched, put a hand on each drum, listen to it in the room, and get the head tension even. Detune the lug that sings. If the kick has a dead spot, move the beater slightly off centre and listen again.
This takes twenty minutes and saves an hour of processing later. It also produces a sound that survives being turned down in a mix, which is the real test.
Overheads are the drum sound
Treat the overheads as the recording, and the close mics as the thing you add underneath. If the overhead pair alone sounds like a kit you would want to hear, the session is going to go well. If it sounds like cymbals with some drums behind them, no close mic will rescue it.
Two approaches cover most cases. A spaced pair gives width and a relaxed, open image, at the cost of a centre that can wander. A coincident or near-coincident pair gives a solid centre and predictable mono behaviour, at the cost of some width. Choose based on the song, not on habit. A dense arrangement usually wants the tighter image; a sparse one can carry the width.
Whichever you pick, measure. Run a tape measure from the centre of the snare to each capsule and make the distances equal. If the snare is one hand-width closer to the left mic, it will sit left of centre and no panning will convincingly move it back.
Phase before EQ
Every microphone you add to a kit is another arrival time for the same event. Before reaching for a filter, check what each addition does to the sum. The simplest method is also the most reliable: bring up the overheads, then add one close mic at a time, and listen to the low end as it comes in. If the kick gets thinner when its own microphone joins, that microphone is fighting the overheads, not helping them.
Fixing it is usually physical. Move the mic. Flip the polarity and listen again. Slide the bottom snare mic a few centimetres. These are free and they work on the source. Delay compensation and all-pass filters are real tools, but they are the second answer, not the first.
The close mics are for character, not for level
The temptation with close mics is to use them to make things louder. They are better used to make things specific. A kick microphone inside the drum, near the beater, gives attack. The same microphone outside the resonant head gives weight and less definition. Neither is correct; they answer different questions. If the song needs the pattern to be legible, go for attack. If it needs the drum to be felt, go for weight.
Toms are the most commonly over-miced part of the kit. If the overheads already capture them well and the parts are simple, a single mic between two toms can be cleaner than one per drum, because it halves the number of open channels bleeding cymbals into the mix.
Bleed is information
Engineers new to drums often treat bleed as contamination to be gated away. On a good kit in a good room, bleed is what makes the recording sound like one instrument played by one person rather than several samples stacked. Gating everything produces a kit that is technically clean and emotionally flat.
Where gating earns its place is in fixing a specific problem: a hi-hat that dominates the snare channel, a squeaky pedal, a resonance that only appears on the loudest hits. Use it there. Leave the rest alone.
What to do when the room is wrong
Sometimes there is no good room available. The honest move is to stop trying to record the room and record the kit instead: tighter mic positions, more absorption near reflective surfaces, and a plan to add space later. That is a legitimate choice, made deliberately. What does not work is placing distant mics in a bad room and hoping the result reads as ambience. It reads as a bad room, and it is not removable afterwards.
The order that works, almost every time: tune the kit, pick the spot in the room, place the overheads, check the sum, then add close mics for the specific things the overheads cannot say on their own.