A drum kit is often picked up with 8 or more microphones.  Thus the sound of any one drum is a composite of the signals from all of the microphones.  To hear the snare, you listen to the direct snare sound though the snare microphone but you also hear snare as leakage into cymbal and tom microphones. 

When two microphones at different distances blend to make a sound, there are some frequencies accented and some diminished due to partial phase cancellation.  The possible resulting frequency response is shown in the illustration below:

The impact of the drums has a lot to do with how strong the impact of the hit is.  Some frequencies, of say the snare may be canceled because of leakage of the snare sound into other drum microphones. This will make the snare sound thinner.

The phase reversal switch can change the frequencies that are diminished due to phase interference.  If is often true that switching the phase reversal on the snare microphone can cause the  snare sound to get fatter and fuller - strengthening the impact of the drum in the mix.  To see if using the phase-reversal switch will help, you need to get a mix of the drums in the monitor.  After you do this, try both positions of the phase-reversal switch on the channel that has the direct snare sound.   Leave the phase-reversal switch in the position where the biggest impact can be heard.  To hear the difference you must be listening to all of the drums in a mix.

Some micing situations automatically require the use of the phase-reversal switch.  An example is the popular "Top and Bottom" snare micing technique shown in figure 2, below.  The bottom skin of the drum pushes out when the top skin is pushed in by the stick hitting it.  If you phase-reverse the signal from bottom mic when blending it with the top mic sound, the impact of the mixed snare drum sound will increase dramatically.

The attack of a drum hit is established in the first 2-3 milliseconds of the generated sound wave.  This means that when a drum is picked up by multiple microphones that are spaced apart by more than 2 milliseconds, the attack becomes "blurred" (less definite).  Thus leakage of the snare into cymbal and tom microphones more than 2 feet away will lessen the snare attack.  This same problem would occur for all of the drum kit instruments.

Sound travels at 1130 feet per second.  Thus it takes about 9/10 of a millisecond for sound to travel a one foot distance.  Recording engineers conveniently round this off to the timing formula of 1 ms = 1 foot.

When you put all of the microphones in time with each other, the attack of the drums is astoundingly better.  Ben Blau, an RID instructor, first introduced this idea some 8 years ago when he was managing one of our recording studios.  I had been using delay to compensate for spaced microphones on guitars but had never tried it on the relatively small time differences of the microphones on a drum kit.  I expected somewhat of a difference when Ben explained his idea to me but I was unprepared for how much of a difference it made.

Another technique in getting the killer drum sound is by using a trigger signal.  A trigger signal can be used to activate a pre-recorded drum sound (from a drum machine or sampler). A trigger signal is a microphone pickup of the drum which has been equalized or filtered to accent the attack of the drum.  The trigger signal isn't used in the mix but sent to the "trigger input" of a drum machine.

Capturing the attack of the drums as a separate signal will make a signal that can "turn on" the sound of a drum machine.  Let's use the snare drum as an example.  The final sound is a blend of the "real" drum sound picked up by the snare microphone and the pre-recorded sound from the drum machine.  The drum machine sound often adds body and attack to the real drum.  The drum machine sound is in time with the regular snare sound because it is triggered to play at the same time.

The ideal trigger signal would simply be a short spike that would occur when the drum was hit.  Any lingering of the sound and any leakage of other sounds (like cymbal leakage) could only cause a "mis-trigger" that activates the drum machine sound at the wrong time.  Thus the signal is purposely make to trigger - not to sound good by itself.

When we learned to place microphones on drums, we learned to place them at the edge of the skin to get maximum harmonics and tone out of the drum.  This was critical to the drum sound but not good for making trigger signals.

Very often it is desirable to equalize the trigger signal to remove leakage from other drums and cymbals leaking into the trigger microphone.  An extreme increase around the 3kHz frequency will accent the attack. Filtering any signal out above 3 kHz and below 200 Hz will dramatically reduce leakage from other drums and from cymbals.

It almost goes without saying that you choose what you consider to be a great drum sound (like that "awesome snare sound") to blend with a natural drum sound that normally would be recorded. You can use a little bit of the triggered drum sound to make a very natural sounding drum have a little more impact. The final decision on the blend, however, is yours.

Since it takes more microphones and tracks to provide both a trigger signal, a drum machine triggered sound and a regular drum pickup, you may only have enough inputs or tracks to do one or two of the drums in the kit.  But even helping out only the snare attack will noticeably improve the input of the overall drum sound.