Samplers - What They Do
& How To Use Them

[Author's note: this article was actually written some years ago, when the technology in use was somewhat more... limited than it is today. One day I'll get around to producing a more up to date version, but in the meanwhile, please read its more quaint references as matters of history rather than current practise !]

The sampler is a much misunderstood item in the recording studio. It is an instrument which, when used imaginatively, is capable of producing all manner of different sounds. Alternatively, if used to the most basic level of its capacity it can do nothing more than be an expensive drum machine, playing back nothing more than a 1 bar pattern repeatedly which has been lifted off a record or CD. In the middle, it can do what until comparatively recently was its most controversial function (indeed, the accusation which has always been levelled at electronic musical instruments since the invention of the synthesiser in the 60s), that of imitating acoustic instruments & supposedly putting other people out of jobs.

The activity which is the subject of most discussion surrounding samplers these days is that of copying; taking small (or some times quite large !) segments of a piece & dropping them wholesale into a new piece. This practise first hit the news a few years ago when The Beastie Boys (I cannot actually remember the specific song) sampled the opening 2 bars (the drum loop) of a Led Zeppelin song (When The Levee Breaks) & included it as part of the backing track over which they rapped. The activity became rife, suddenly every song in the charts had bits which were recognisable from the 'rock classics' of the 60s & 70s; Robert Plant himself even decided 'If you can't beat em, join 'em' for his 'Now & Zen' album of 1987 by sampling parts of songs he wrote 15 - 20 years previously; & the copyright lawsuits flew around the world like there was no tomorrow. Eventually a solution was found, where if you wish to use a section of a piece in a commercial recording, you apply to the copyright owner (usually the record company concerned) & buy a 'license' to use so many seconds of it. Recently it has been made even easier, where you can buy whole CDs full of 2 second bursts of drums, orchestral hits, silly noises, & panting women (!) where the license to use them is granted in the rather high price that some of them cost (50 lbs per CD is not untypical).

At its most basic level, the sampler can be best likened to a digital tape recorder. Using an Analogue to Digital converter, it turns incoming sound into numbers, a series of binary 0s & 1s. The exact amount of numbers it turns a given amount of sound into depends on 2 things - the Sampling Rate & the number of Bits. A higher sampling rate means that higher frequencies can be recorded - the typical untrained human ear can hear sounds up to 15 to 20 Khz (remember, the frequency of an 'A' tuning fork is usually 440 Hz, & the 'A' from an instrument contains other much higher frequencies), & for technical reasons you have to sample at twice the highest frequency you wish to be recorded; early samplers had a maximum sampling rate of 20 Khz, so the highest frequency that was recorded was 10 Khz, leaving a rather dull sound lacking in what sound engineers & hi fi buffs call 'top end'. Modern samplers can record at rates of up to 48 Khz, which is actually higher than the rate CDs are recorded at (44.1 Khz). Unfortunately there is a price to pay for higher sampling rates - more computer memory is required to actually store the samples in. As an example, the Roland S-50 can sample at either 30 Khz or 15 Khz, giving you a maximum of 15 seconds of sampling time at 30 Khz, or 30 seconds at 15 Khz. The other variable which is relevant is the number of bits the sound is converted to; early samplers such as the Ensoniq Mirage or the Fairlight Computer Musical Instrument were 8 bit samplers, which lended the sound a some what rough, buzzy tone in comparison with today's 16 bit samplers; however, studio fashions being odd that they are, some people actually deliberately go for that early sound in some recordings, which it cannot be denied has a certain 'character' !

Once the sound has been converted into numbers, you can then do all kinds of posey things with it; such as looping it, enabling the 1 second of trumpet sound you sampled sound indefinately; defining start & end points, getting rid of the irritating intake of breath because you pressed the button too early; if you have flashy sample editing software such as Steinberg Avalon you can actually alter the sound on the wave form level thus rendering it completely unrecognisable as the sound you originally sampled; & of course you can transpose the sample up & down the keyboard, thus allowing you to play 'Happy Birthday' with the sound of your own Burps.