Overview

In many genres of music, drums are rightfully a center piece of the overall arrangement.  After all, drums are percussive elements that establish beat and tone, convey intensity, and define levels of swing, so to me and the music I like writing it’s natural to place as much importance on them as the musical arrangements of a song.  Throughout the history of electronically assisted commercial music creation and production there have been many styles, techniques, and practices developed.  Some of the most incredible music out there bends and blends these together to avoid pitfalls of stereotypes and “same-y” sounding productions.  I’ve talked with a lot of producers about production and have received a lot of great advice and ideas.  The purpose of this entry is to go over them along with my take of how I utilize them in my own productions.  As always, in my opinion there aren’t really any wrong decisions in music, but there are definitely better decisions and these are the ones I tend to employ most often.

 

Frequency Is Everything

To some, frequency with relation to audio could simply denote the stale laboratory definition of the range of human hearing (20Hz to 20KHz).  In application, I find that it’s better to talk about pockets of frequencies. A very talented producer I know, Grand Daddy Frost, recently spoke about frequency in a video and it really got me thinking about how important the presentation of frequency is.  If music is a series of tonal frequencies, then the presentation of those frequencies can suffer if they are not properly played back and cleared up of distraction.

Have you ever been in a room where only a drum is playing a full drum set?  It’s an amazing experience.  You get a full barrage of frequencies thrown at you.  The higher frequencies are a soup of cymbals and hihats aided in their presentation by the reverberation (or robbery of reverberation in the case of treated rooms).  The upper mid frequencies clue you into the materials used for the percussive elements: the snares and toms have a thin layer the drum stick strikes, while the kick is usually noticeably more thump oriented, with the hihats and cymbals giving away their metallic natures by the high pitch of the initial hit.  In the lower mid range, we feel heart beat of a drummer’s performance where the driving nature of a kick and share push the rhythm.  In the lower end frequencies, we’re met with a curious aspect of human hearing in that while we can’t hear the fullness of these frequencies directly (much like how a person can’t hear a dog whistle, those here at the opposite end of the spectrum), we are very much able to both hear and feel the atmosphere being moved around us by their wide frequency waves of around 20Hz and lower.

From the various eras of music production I’ve listened to, there have been three main approaches utilizing these pockets of frequencies: 1) attempting to reproduce faithfully, 2) attempting to hyper-accentuate, or 3) attempting to disrupt the ordinary pocket ranges entirely.

And it doesn’t matter if you’re recording a live drummer, arranging real, synthesized, or sampled instruments, in a production one of these three techniques will be employed, and sometimes combinations of the three, or all three at once for layering concepts.

Let’s take a look at approach #1, attempting to reproduce faithfully.  Because frequency pocket response acts differently in the same space different effects can be applied to various parts of the drums to re-engineer room response that wasn’t picked up on the way into recording.  In the most common approach, this generally meant putting a mic inside the kick drum, putting mics near snares and toms, and then an overhead mic along with two mics on either side of cymbals and hihats.  Afterwards some light reverb will be given to the hihats and cymbals, and depending on style varying degrees of reverb on the snare, with the kick drum usually being left untreated.

In approach #2, attempting to hyper-accentuate, we see some of the discoveries and stylistic touches made popular in the 80s and perfected in the 90s and 00s.  Namely, the following: Utilizing EQ to bring out higher frequencies in the kicks (among other things, making kicks more pronounced in formats like radio where lower responses might not always be heard).  This was popular in everything from Metallica to Peter Gabriel.  Utilizing compression and attack shaping of the kicks and snares to accentuate the chief drivers of rhythm (the SPL Transient designer is a great tool for this).  Gating a snare with a thick reverb, where the tail end of a reverb is cut off giving a quick “breath” of reverb while another signal controls its play and release envelopes.  That’s certainly not all of the stylistic touches, but it gives an idea to what hyper-accentuating means with relation to drums; they refine the sound one might hear in a room alone with a drummer and a drum set and focus more on bringing out the presentation in ways that might otherwise be unnatural in a regular acoustic room.

Finally, in approach #3, attempting to disrupt the ordinary pocket ranges entirely, we are introduced to the concept of utilizing percussive and rhythmical elements in ways they weren’t originally intended and in ways that are impossible to recreate without being aided by electronic or analog effects.  One might notices that in many delay and reverb processors today, there are filter controls one might usually find on a synthesizer, namely cutoff and resonate controls.  These controls allow the effects to be synthetically altered from a natural response, where delay output of hi hats might have lower frequencies where as naturally they would simply dissipate in volume, high and low end all simultaneously (think yelling into the Grand Canyon but you have effects further altering the response).  Where this approach differs from simply hyper-accentuating is that it also often involves altering the sounds of the percussive elements themselves.  Maybe you’re passing a kick or 808 sub kick through a distortion processor and cutting off the lows.  Maybe you’re layering a high pitched kick or lower pitched snare along with a clap and manipulating the stereo image of the output.  Maybe you’re sending drums through a synthesizer’s input processing feed such as those available in a Korg MS-20 or any various modular setups.  Maybe you’re not even using drums or drum machine at all and are simply establishing a rhythm with entirely synthetic components such as a synthesizer.  A Roland TR-808 is largely a synthesizer attempting to emulate their acoustic counterparts, but you could also use a modular synthesizer to play a rhythm that doesn’t resemble a drum set in the slightest.

 

My Application

I grew up listening to everything from Digital Underground to Massive Attack to The Prodigy to The B-52s.  I don’t claim to have a masterful ear in producing and engineering drum elements, but I have heard a lot of good things that tend to sound good to me.  Over the years I’ve developed a system to employ this regardless of the source of the drums, be they sampled, synthesized or whatever.  My method of mixing drums falls between reproducing faithfully and hyper-accentuation, though with hihats I definitely try to do weird things because they’re more present throughout a mix usually existing in between kick and snares and therefore more apt for usage with a mostly dry phaser or flanger to keep the drums sounding a bit interesting overall.

Reproducing A Little Imperfection

A chief element I tend to spend time on is attempting to reproduce a little imperfection.  I tend to rely on delays for this, aided by very selective filtering.  Previously, I talked about how some recordings simply trying to reproduce what it sounded like to be in a room with a drummer and a drum set.  If you’ve ever heard a drummer play, you’ll quickly pick up on the curious effect of even when they mess up or get slightly off a beat, it can still sound really good.  It’s ridiculously hard to re-engineer this if you’re starring at a sequencer.  Sure, you can apply things like swing, but that still sounds mathematically cold.  With a delay and a filter on hihat and cymbal elements, you can produce and almost misfire that doesn’t take away from the rhythm you’ve captured or programmed in.  Curiously enough, it wasn’t even electronic music producers who discovered this phenomenon and how implement it artificially, it was a subgenre of raggae called dub (not dubstep).

A great example of this would be the Mad Professor Vs. Massive Attack track entitled “Radiation Ruling the Nation (No Protection)”:

If you notice in this song, the hihats are achieved by applying a delay effect to a hihat rhythm presumably established first on an Akai sampler or MPC.  The entirely of the hihats have no frequency response below upper mids and largely exist in the higher frequencies, that’s their pocket for this production.

I remember someone once struggling with hihats on a beat he was making on an Akai MPC 2000XL.  Essentially, he wasn’t utilizing an 8out board to a mixer, so all elements were going through the main L/R out.  He had cleverly normalized all of his samples, but that also meant his hihats were at the same levels of kicks and snares.  Keep in mind, this doesn’t really satisfy any of the 3 methods of production I mentioned above, so in this case I would consider his attempt simply unproduced as it is.  And that’s fine when writing music.  You can go down some rabbit holes trying to mix and write at the same time (to the point where some people tell you just not to do that at all).

Pocket Isolation

You gotta keep em seperated!  Obviously a good sounding drum mix will have overlap because each element of a drum kit doesn’t exist in a vacuum, nor do drum machines frequency isolate each instrument.  But separating the elements will lead to a generally more pleasant sound.  Simple isolation doesn’t take much effort to set up either so it’s something you can set up early in your writing process if you’re writing entirely in a DAW or partially in a DAW and come back to mix properly once your instrumentation and vocals are recorded as well.  In fact, I wouldn’t break out *any* mixing gear or plugins when you’re just working on a beat.  Aside from the drum mix needing more tweaking once everything else in place, mixing and mastering plugins can sometimes introduce a ton of lag in your DAW playback response making it difficult to record a live part later on.  Aside from delay and reverb, focus on your DAW’s native filtering and EQ effects as they are likely very CPU optimized.   A glue compressor of some sort between a kick and snare can make their frequency separation sound a bit more natural or pleasant.

 

The Common Situation

So let’s put ourselves into a common situation (especially for people new to sampling and production): you’ve fired up an MPC, sampled records or your drum collection, and you made a sweet groove.  It doesn’t sound polished, and it’s maybe blowing out your headphones or speakers with the bass response.  As is, if you were to simply take it and attempt to add vocals or additional instrumentation, there likely wouldn’t be any place for those elements leaving one with what many refer to as a “muddy” mix.   The problem with relying strictly on a sampler or MPC or drum machine in production is that there’s often not a way to place elements in their own pockets for a cleaner mix.

Keep in mind that all of the material you’ve heard from the 70s up to the 90s was fed into an analog console.  Each channel on the console had built in effects like compressors, HPF, LPF, and EQ.  More than that, each signal passed through a channel went through some form of additional coloration from discrete hardware due to the nature of tube components.  If what you hear coming out of the sequencer or MPC doesn’t sound professional, it’s not the end of the world!  Until things are mixed, pretty much all projects sound pretty terrible as far as the mix goes.

 

How To Make It Better

So we have the sweet groove, and we’ve either wired the multiple outputs of an outboard sampler or MPC or we have a plugin sampler and we’ve wired their busses into dedicated audio tracks in our DAW.  I commonly approach my drum I/O like this:

  • Channel 1 (mono): Sub Kick – On vinyl, it was common for mastering engineers to reduce the stereo width of the bass image/pocket range because disparate frequencies in the bass stereo image could cause the needle to inadvertently skip.  These days, it’s no longer technically necessary with digital playback.  However, a pleasant side effect of narrowing the bass stereo image is that it made the final mix much cleaner sounding.  I think the science behind this is that we feel bass more than we hear it.  Hearing a single or narrow bass image is more natural and easier on the overall sound than a bunch of data going on down below.  I always sometimes slightly overdrive a sub kick with a tube driven micpre to give it an element of slight saturation.  This also helps make sub bass kicks audible in higher harmonic frequencies on devices like phones that don’t have subwoofers:  you won’t hear the sub bass, but you’re hear a higher harmonic of them along with a litter grit and character.  I usually find myself chopping off everything below 18Hz because no one can hear it despite it adding to the overall headroom available within a mix.
  • Channel 2 (mono): Kick –  I primarily produce electronic music, which is known for strong beat driving kicks.  However I’ve found that the best sounding kicks are actually layered kicks containing a sub kick and its harmonics along with the upper frequencies of another kick, usually sampled or an drum set kick heavily processed.  I keep the lows and mid lows out of this channel because that’s usually where I want to place my synthesizer or bass guitar parts.
  • Channel 3/4 (stereo): Snare or ClapsLike my Kick channel #2, I like to keep lows and mid lows out of these channel.  Additionally, I find it sounds pleasant to add a slight reverb to snares and claps.  You can achieve an 80s gated sound when isolating these channels in this manner. 
  • Channel 5 (mono): HiHats – I probably apply more effects to hihats than any other instrument in a drum beat, and my general rule is that spatial effects tend to sound better with a mono input.  So I start with a mono HiHats channel and feed that out to my stereo reverb and delay effects.  I like to cut off everything but the upper mid and high frequencies and bring down the volume.  This approaches the hyper-accentuate method I talked about earlier because in the absence of super loud hihats the kicks and snares are much more pronounced.  This is probably the most common issue I see with people producing beats:  they don’t manage their hihat levels in relation to everything else going on.  Sometimes it can sound good.  WuTang Clan is famous for having a striking hihat line come through in a mix, but they also very cleverly EQ’d and mixed their snares and kicks along with it to gel really well with their sampled musical arrangements.
  • Channel 6+7 (stereo): Cymbals – I generally process my cymbal crashes and reverse cymbals much like how I do HiHats, however I give them a bit more upper mid frequency pocket owing to they’re being physically bigger and more pronounced in real life.  Along with also reducing their dB in the mixer like the hihats, I also do purposefully different reverb and delay effect chains to draw them out and bring attention to them.

Some import things to note:  First, always mix with your ears.  Just because something looks right through a plugin doesn’t always mean it’s the way it sounds best.  Be mindful of clipping digitally, but also don’t get too attached to customs and habits.  Sometimes I bring HiHats far down in the spectrum, sometimes I allow kicks to exist in the 25Hz range alongside a sub bass kick.  Ultimately, what mixing drums in this general manner does is allow me to add other elements to a song without crowding an overall mix.  Also, I know there’s a lot of different schools of thought about stereo mixing and drums, but I tend to want to keep my drums and bass in the center as much as possible.  This frees me up to use the wider stereo image for synthesizer pads, the outputs of effects processors taking in mono audio, and other elements.  If the kick, snare, bass, lead, and vocal can’t fit in the center, then I’m usually of the mind that the pockets containing each need more attention and differentiation.

 

An Example

I’ve made a short example of the manner in which I usually mix so that ultimately if it’s something you’d like to take and implement in your own productions, you can see how I do it.  You can mix drums like this in any DAW.  I work primarily out of Pro Tools, but I also use Ableton Live quite a bit and seeing as how that’s a much more popular DAW I’ve chosen to implement it there.

Below is a project folder containing all the samples of a short arrangement of drums I’ve mixed together.   It follows AB testing where A is unmixed and B is mixed.  In this example, the entire beat is made out of one hit samples from my collection.

Mixing a Beat Project (download)

If you do not have Ableton Live Suite 10, and would simply like to listen to the example, here is the audio file.

 

Conclusion

This isn’t the only way of mixing, it’s not the best way of mixing, but I’ve found it works best for me.  I hope at the very least it’s given you an idea that you can snag and make your own.  But remember that the best tool when mixing is reduction.  In placing drums (and instrumentation and voices) in their own pockets in a mix, you keep them from trampling over each other.  Sometimes I follow my own technique and sometimes I jettison it if it’s not working.  Notice I didn’t make all the elements mono that I usually do in the project example above.  Cheers!