When we want a track to sound powerful, hard-hitting and real, we need the drums to sound authentic. Today we're focusing on the possibilities for recording drums through realistic sequencing. Productions that use drum machines (or samplers) with electronic sounds are probably less affected than those that require large doses of realism from acoustic drums. Obviously, this applies to every instrument, since digital or "made-up" sounds don't have to be faithful to the listener's memory.
Record or sequence?
I think we'd all love to be able to record a good drummer to add some rhythmic color to our songs, but in home studios we run into several problems. The first is space, though it's not the worst. Setting up a drum kit in a small space isn't within everyone's reach, although it's a fairly solvable problem. The second, and worst, is soundproofing. If our studio is in an apartment, it'll be practically impossible to concentrate, since we'll be bothering half the block, at the very least, when we play. Third, we need good mics, know how to place them and have some capacity for simultaneous multitrack recording to capture several channels at once and, logically, separately. Fourth, many of us who record in small or mid-sized home studios are musicians, not audio specialists, and all that EQing, compressing, limiting and so on is something that can be done, but it takes us ages; we often go by intuition and end up with mediocre results (surely some amateurs get heavenly results, but they'll be few and far between, no doubt).
With sequencing and the use of loops, MIDI and samplers, we solve these problems as follows:
1. Space: the space needed on your hard drive, not physical space.
2. Isolation, noise, disturbance. We work by line, so we control the overall volume, and we can even work with headphones.
3. Mics. Obviously, sampler samples (some better than others) are captured by professionals using top-notch mics.
4. Mixing process. Still necessary, but the material is recorded much better than a small studio without the right gear could manage.
5. Separate channels. Most of the latest-generation programs provide each track independently: snare, kick, hi-hat, toms, room, etc. That way, we don't necessarily need an interface with lots of simultaneous recording inputs, and we have all the tracks separate for working on EQ, mixing, dynamics and so on.
Advantages of loops and libraries
Many of them are already described in the previous section. Basically, we can underline quality and variety. If we have a good library, we'll have a range of kicks, snares, hi-hats, toms, cymbals and so on, all from top brands and well recorded.
Quantization. Although by now it's possible to quantize audio, so with the right systems we could correct tempo errors in recordings of a real drummer, quantization processes are becoming more flexible and reliable every day. If you're an expert, you'll only need it to polish minimal deviations. If you're a beginner, the pressing need to have a tight rhythm section will be solved with quantization tools. If your drums are off the grid, it'll be impossible to make a "polished" song, so your track's rhythm section needs to be quantized and perfectly locked to your click.
Achieving realism through drum sequencing
The truth is that for a professional with a good studio, the know-how and the resources, it's almost more convenient and faster to record a drummer. I'm talking about good, believable results. Sequencing a complicated, "humanly real" drum part takes many, many hours of work.
If we sequence a basic (cheesy) kick, snare and hi-hat beat, two bars repeated throughout the whole song, what do we get? An unbearable botch job. For this reason, we need to sequence every detail of the song with its changes, breaks, rhythmic differences and, hardest of all, with the right dynamics on each piece. True, most of the time we can copy a bar and repeat it in several sections of the song. That's the easy part.
Consider that a drummer uses all four limbs and controls the dynamics of each piece at the same time. Bringing this into sequencing, especially in breaks, dynamic changes and so on, is in my view the most tedious part if we do it from a keyboard or similar. When we sequence, we record in parts. It's hard to record all the pieces of the kit at once, so you have to match the dynamics of the various drums. The editing tools are convenient and practical, since we can later correct "velocity," duration and so on, but even so, it's still painstaking work.
For those who don't know, there are tools that control volume (velocity) in the edit windows. That way we can assign a constant velocity to the kick, the snare, the cymbals, etc. This brings a problem: making the volume of the pieces uniform gives a mechanical, unconvincing sound. On the other hand, if we have hits (on the kick, for example) with very different intensities, it'll be impossible to mix well. Imagine all the kick hits are at a medium-low volume but you've got four stray hits at full blast somewhere. That track will have "peaks," and you'll go crazy trying to find where they are. If you turn everything down to avoid the peaks, the medium-low hits will end up too low. A good way to solve this is to fix the velocity of certain pieces in bars that don't have variable dynamics due to our song's "feel." Afterward, we adjust the velocity of the remaining hits "by hand."
Tricks for achieving realism
We need to avoid a mechanical feel. To get a powerful drum track, we should try to give it every nuance possible. Given how crazy it is to sequence the whole song piece by piece, we can do some of the following:
1. Sequence a "pattern" bar with the song's beat, copy it and duplicate it two or three times, and vary the hi-hat or small snare, cymbal or other touches in some of the bars. Use realistic editable loops that fit our song.
2. Once we have a phrase of plenty of bars with some variations, as described in step one, we can move on to duplicating that entire section. This gives us a varied base rhythm that sounds less "machine-like."
3. Before duplicating, it's a good idea to check the velocities carefully so all the bars are perfect in nuance and intent as well. That way we'll get consistency in our track.
4. Now we program the stops, changes and breaks. We insert them into the corresponding bars.
5. Once programming is finished, we can choose which final sounds we want to assign to each piece, although it's most effective to make the final choices during the mixing stage. For example, a snare may seem perfect when you listen to the drums on their own. Once all the tracks are recorded, that snare may turn out to be too fat, thin, harsh, metallic...
6. Effects. Likewise, it's best to add reverbs (or whatever effects you want) at the mixing stage, for the same reasons given in step five.
Multisample
We've already talked about this method on occasion. Multisample libraries have a feature that's especially useful for getting a "human" touch. The same key triggers different sounds from the same piece of the kit.
Taking samples
When the manufacturer (or we ourselves) builds the library, they take a snare, for example, and record different takes: struck softly, medium, hard, in the center of the head, off to the side, hitting rim and head, with overtones, and so on.
So how do we handle all those takes?
Let's imagine the snare is triggered in the first octave, on the note C. The library is set up so that it triggers the soft sound if we play that C softly (between velocity X and Y), the medium sound, the sound with overtones and so on, always depending on the velocity (force) with which we press the C key in that octave (or on the master MIDI instrument).
If you try it, you'll hear the difference. Take a sound and play a snare break with it, same velocity, same sound. The result is a totally artificial break. Now repeat it using different nuances, dynamics and multisamples. You'll find it sounds far more like a real drummer. All that's left is to match each sound to what a real drummer does, and the best way to do that is to watch how drummers do it.
Editable loops
To avoid hours of programming, loops started being used a while back. A series of bars recorded by a drummer. Wonderful: a well-crafted hi-hat, perfect dynamics...
So far, so good. The problem came with how locked-down those loops were. They came set to a specific tempo that was hard to change, and we had to adapt our song to the loops. That meant a lack of freedom and therefore limits on our creativity.
How have those problems been solved?
There's no denying that for doing things quickly, loops are perfect. The first hours or days of programming are already taken care of, and probably with more fidelity and realism than we'd have achieved ourselves.
Various tools appeared to get past the lack of flexibility I mentioned. For example, libraries started providing variations on the same loop. That way, we can choose what best fits our intent and get a varied groove. We also find phrases in loop format that are triggered by one key but also lay out all the kit's sounds separately across the MIDI instrument (a keyboard, for example). Thanks to this breakdown, we can program delicate or special bars in our composition while keeping the loop's original sound. Then we combine the stock loops with the bars we programmed ourselves, and we have our song with the loop's original richness and our composition's own personality.
The Rex and Rex2 formats are also worth mentioning. Thanks to them, a loop will automatically adapt to whatever tempo we've set. This way, the loop adapts to us rather than the other way around.
Sequencing and editing programs include a wide variety of tools for programming work and, in their latest generations, bring valuable new features for working with audio. This lets us "get our hands on" audio loops and even reshape them to our needs without losing exact quantization and tempo.
Master instruments in pad and drum formats
Another way to save programming time is to have a MIDI drum kit, percussion pads and so on: Roland, Yamaha, ddrum, etc.
Besides having this kind of instrument, we (logically) need our drummer, who will surely go on about the advantages of their acoustic kit compared with the pad(s) you hand them. If they're the one playing, they'll nail all the flourishes, variations and hit dynamics for you, and you'll save countless hours at the computer while getting a truly impressive result.
The evolution of this type of instrument is incredible, and problems with flams, stick bounce, false hi-hat triggers and so on are now a thing of the past. It's a perfect way to record drums that will later be triggered via a sampler.
Conclusions
When you spend dozens and dozens of hours programming and then stop to listen to the result, you may be amazed at what you've achieved: realism, punch, a human sound, power, feel... That said, if you hear a good drummer play your song, you'll usually be left slack-jawed, because no matter how many hours you spend programming, there's nothing like a good drummer to bring rhythmic life to any kind of song.
If you want to record demos or sketches and don't have the means to properly record a drummer in the studio, this is the way to go. I'd even say (without meaning to annoy the pro studios) that you can get some really, really nice results. The secret: you can spend as many hours on it as you need, since you're not paying by the hour at home, so you can put more work into the final result.
However, if by some remote chance a good drummer can do the job for you, all nicely recorded on a powerful acoustic kit..., don't think twice.
