Play was released by Moby in 1999. It uses a variety of samples from old blues and gospel recordings. Some of the samples contain background noise and are quite ‘grainy’ or ‘lo-fi’. Moby uses this as a creative effect. Play was both a critical success and a commercial phenomenon. The album introduced Moby to a worldwide mainstream audience, not only through a large number of hit singles that helped the album to dominate worldwide charts for two years, but also through unprecedented licensing of his music in films, television, and commercial advertisements. It eventually became the biggest-selling electronica album of all-time, with over 12 million copies sold worldwide. The video below shows some of the samples that were used in the album, both in their original and new contexts.
What is Sampling?
On a basic level, sampling is when you take a part of a song, single note or sound and reuse it in another context. It is common to use a sampler to either record, manipulate or playback one of these pieces of audio material (or any combination of the three).
You can also sample single notes or hits. These can be used to form a sampler instrument, which can recreate the sound of a real instrument very accurately, or create a brand new instrument from a sample.
Originally in the 1960s, samplers used tape loops. To change the pitch on a tape-based recording, the tape was played faster or slower. However, unfortunately, tapes were subject to hiss, wow and flutter, and degradation. The Mellotron was a tape-based sampler with loops attached to each key used by The Beatles. In the video below, Paul McCartney demonstrates the Mellotron that was famously used in Strawberry Fields Forever.
Sampling was later extensively used in hip hop (1970s and 1980s) using turntables Drum machines used early sampling technology because the samples required were short and did not require pitch shifting. Modern day samplers are digital, and often triggered using MIDI as part of a DAW.
In modern day samplers, it is noticeable when a sound is pitch shifted beyond a few tones (when this is done in relation to a keyboard, we call it ‘keyboard tracking’). Multisampling is used to overcome this, ensuring samples are only used over a limited range. Velocity layering is used to change the sample depending on how hard the key is played It is important to ensure samples are edited at a zero-crossing point to avoid a click or a glitch; another solution is to use crossfade looping.
Repeats the sample
Increases the volume to the maximum without distorting
Repeating small parts of the sample
Adding spaces between small parts of the sample
Extending or reducing the time of the sample independently from its pitch
Changing the pitch of a sample independently from its length
Playing the audio data of sample backwards
Flips the waveform so peaks become troughs and vice versa
Sample Rate and Bit Depth
Analogue to digital conversion and recording also involves sampling. As part of this process, we take amplitude measurements of a waveform, and assign each measurement a number or value.
The bit depth gives us the accuracy of the amplitude measurements taken.
The sample rate is the number of amplitude measurements taken per second. It is measured in hertz.
When an analogue signal is sampled, we end up with a stream of numbers which make up the digital signal This is usually measured in binary code (0 & 1). A common bit depth for high quality sampling and CD audio is 16 bit. This means that there are 216 possible amplitude measurements (65,536).
A common sample rate for high quality sampling and CD audio is 44,100Hz This caters to Nyquist’s Theorem, which states that the sample rate is twice the highest frequency captured. Since the human hearing range is between 20Hz and 20,000Hz, this means that high quality sampled audio and CD audio is able to capture all frequencies we can hear with a bit to spare.