In our previous installment, we tackled the specifics that govern the sample recording process, including sampling rate and sample resolution.
This time around, we’re going to cover how samples are edited and prepared for playback via a software or hardware sampler.
From there, we’ll go over how to make the most of your sampler’s resources.
Once a sound is converted to digital information, the creative possibilities expand exponentially.
At the simple end of the spectrum, you can slice a sample into tinier bits by adjusting the start and end points, then use it percussively or musically – either in its raw form or processed by the sampler’s synthesis tools.
As your skills advance you can apply sample-looping functions to make a sound sustain indefinitely and/or zero in on exotic harmonic spectra that would be difficult to generate using conventional means.
Sound designers and advanced users will eventually explore the possibilities of multisampling, which allows the creation of detailed instruments that can capture the essence of both acoustic and electronic instruments, as well as serving as a method for creating drum kits and the like.
Start and end points
Once you have your sample loaded into your sampler, the first task is to set the start and end points.
These function as expected, with the start point determining where a sample begins playing and the end point determining where it stops.
This enables you to zero in on exactly the material you want to play.
What’s more, you can also create interesting transients by setting the start point in unorthodox ways.
Pictured above is a nifty example: You can add a clicky attack transient to any sound – especially 808 kicks and toms – by setting the start point to the first peak in the waveform.
This works especially well with muted analog sounds, so if you’re after glitchy and/or minimal textures, this is a good place to get your feet wet.
While looping beats and musical recordings is a well-understood process, looping within a sampler is a very different approach.
When samplers first hit the mainstream in the early 80s, RAM was a luxury that few could afford.
Even by the mid 90s, a mere 32MB of sampler RAM was still considered downright extravagant.
In order to have sounds like strings or woodwinds sustain indefinitely, sound designers would use looping to take a fragment of the original recording and loop that section, thus creating the illusion of a sustained note.
If you listen closely to single notes in older workstations and early softsynths, you can actually hear slight imperfections – like bumps or cycling – within the sustaining portion.
These are loops.
The above picture shows a looped sample, with the medium green segment being the part of the sample that plays before the loop and the lightest green segment being the loop itself.
Since most modern samplers include a pre-recorded set of sampled instruments, you no longer need to dive into the complexities of hiring an orchestra and meticulously sampling every note and articulation, then looping it.
But that doesn’t mean that looping isn’t a skill worth mastering.
One nifty looping trick for creating exotic textures can be accomplished within Ableton Live’s basic Simpler instrument.
Take any sound (really, any sound will work for this, even a snare drum or cymbal crash) and set the loop points for just after the beginning of the sound.
Next, create the smallest loop possible.
If you’re doing this correctly, you should hear the initial attack portion of the sound, followed by a buzzy, bright sustaining segment.
Now, slowly increase the length of the loop.
The pitch of the buzzy section should drop in pitch until you start to hear a rhythmic bumpiness.
Now, experiment with different loop lengths until you hear something you like.
Keep in mind that you may also have to adjust the transposition and detuning of the sound in order to keep it in tune with the rest of your track.
Once you’ve tackled the above experiment, you’ll have a better understanding of how looping works.
From there you can begin to explore more advanced looping functions.
Here’s a list of some other looping features and their uses.
Forward looping: This is the default mode for looping samples, with the loop playing forward until it reaches the end point, then starting over from the beginning of the loop.
Bi-directional looping: Instead of starting over at the beginning, the loop reaches the end, then moves backward to the beginning of the sound, then forward to the end and so forth. This is sometimes referred to as back-and-forth looping.
Crossfade looping: This parameter works well with longer loops by blending in a bit of audio from before and after the loop start and end points, which can help to smooth the loop transition for certain sounds.
Snap to zero-crossing: More often than not, the best loop points are located at the point where the waveform crosses above or below the horizontal axis. By turning on “snap” the sampler will auto-detect these crossing points, which often helps to remove clicks and pops from certain types of looped material.
One of the most intricate tools in a sampler is the multisampling function, which allows you to map many different samples to different ranges of a keyboard.
Most commonly, this is used to map different drum samples to various keys so that you can create a single preset for an entire kit.
In professional sound design, multisampling is used to create authentic sounding acoustic instruments.
Well, if you’ve ever tried to sample your friend’s guitar or violin, you’ve probably discovered that sampling a single note doesn’t necessarily translate across the entire keyboard range.
This is because, like vinyl or tape, playing a sample at a higher pitch changes its speed and thus its formant content.
This is great for deep techno vocals or munchkin voices similar to those in Bass Kleph’s “Helium”, but not so great for accurate instrument sampling.
Accordingly, manufacturers have implemented multisampling – sometimes called “keymapping” or “zones” – which allows you to assign multiple samples to specific ranges across the keyboard.
Pictured above is a five zone multisample from Ableton Live’s Sampler instrument.
Most modern samplers also allow you to also create zones for various velocity levels, so different samples are triggered depending on how hard you hit a key.
By sampling different articulations and pitches across the entire range of an acoustic or electric instrument, it’s possible to create convincing pianos, orchestras and so on.
One final tidbit on multisampling.
Each sample that’s assigned to a zone has something called the root key.
If you look closely at the above zone image, you’ll see tiny little r’s in each region; these are the root keys for each zone.
The root key is the key at which a sample plays pack at its original pitch.
Since most samplers default the root key to C3, you will often have to adjust the transposition of this note to reflect the actual key being played.
Just a heads up…
Tips for making a large sample consume less memory
Having a ton of sounds in your sampler – especially if it’s a hardware-based device – frequently requires the management of memory so all your samples can fit into a single program or sequencing session.
Here are some approaches for helping conserve memory.
Truncate or crop unused data. If you’re sampling a tiny bit of audio from a longer recording, there’s probably a ton of data that’s simply not being used. Most samplers allow you to delete the information from before and after the start and end points via tools labeled trim, crop or truncate.
In the above image, the green section represents the range of sound data that is being used, while the black data outside that range can be safely deleted to free up some memory.
Note: To be on the safe side, move your start and end points outward in either direction before cropping, so you have some margin for error when fine tuning your material.
Sum stereo samples to mono. Since a stereo sample consumes twice as much memory as a mono sample, it’s often worth the extra time to convert the sample to mono (usually in a separate audio editor) before loading it into your sampler.
Sample at a lower sampling rate. As discussed in the previous tutorial, sampling at rates higher than 44.1 kHz (or 48 kHz if it’s a video source) really doesn’t yield that much of an improvement unless it’s a pristinely recorded acoustic instrument. It also consumes more CPU and RAM due to the fact that more samples are taken at a faster rate.
Sample at a lower resolution. More bits consume more memory and CPU, so again, unless the sample requires intimate detail, you can usually get away with 16-bit samples – and if you’re into distorted lo-fi samples, feel free to sample at 8-bit if your sampler supports it. Just be aware that once the bit depth is lowered, there’s no going back to a cleaner sample.
Once you have your samples properly edited, you can then use subtractive synthesis tools to further shape the sound.
Next up? The arcane world of FM synthesis.