Over the past few years, samplers have become an essential part of the MIDI recording studio, and it's my guess that the way we use them now is far removed from what was envisioned by their inventors. The first samplers allowed us to capture short bursts of 'real' sound, loop them and then replay them polyphonically from a keyboard -- which was lots of fun at the time. Hearing normal sounds removed from their original pitch was quite a novelty at first, but soon we demanded realism, which meant multisampling, smooth looping, more memory -- and more time spent programming than playing! Along the way, sound-shaping tools were borrowed from the traditional synthesizer, not just amplitude envelopes but resonant filters, modulation and pitch bend. Today there is little difference between a sampler and a synth other than the fact that you can easily load new 'oscillator' sounds into a sampler, which you can't do with most synths.

Of course it didn't take musicians long to realise that if you could sample a single note, you could sample a whole musical bar to create a loop. And if you could sample a bar, why not fit more memory and sample a whole verse or chorus? In fact if you read the Tracks feature in this issue, you'll find that composer David Lowe recorded and mixed virtually all of the recent hit 'Would You?' using just two samplers, augmented only by a couple of synth patches to provide the piano and bass parts.

Many of today's samplers, including the new Akai S5000 and S6000 models reviewed in this issue, incorporate direct-to-disk sampling and playback so that the length of a recorded segment is no longer limited by the amount of RAM fitted -- so what's the difference between a sampler and a multitrack recorder?

The answer depends on how you work. In many instances, the sampler can fill the role of the multitrack recorder, but evolution still has a few more steps to take before the sampler is both jack and master of all its trades. For example, triggering a long sample from disk is all very well, but once triggered, the timing can't generally be sync'ed to an outside source other than via word clock. The longer the recording, therefore, the greater the probability that the timing will drift, especially if the sequencer is locked to an analogue tape machine. What's more, as far as I can tell, none of today's samplers will let you start audio playback part way through a sample -- you have to go back to the beginning, unlike a decent audio/MIDI sequencer, which can start playback from any point in a song. However, it can only be a matter of time before these points are addressed, in which case the sampler will be able to do anything a hard disk recorder can do.

Why is this important? Unlike working with audio in a computer, a sampler allows various sections of music to be mapped to different keys on a keyboard, so the composer has the chance to experiment with the audio in real time. That's how the catchy 'Would You?' trumpet riff was first conceived. In the words of Eric Morecambe as spoken to Andrew Preview, the original recording contained all the right notes, but not necessarily in the right order! A little real-time dabbling using the sampler and keyboard resulted in the final riff, which was later re-recorded by the trumpet player.

If the sampler can combine traditional sampling with hard disk recording, what else can we look forward to? Personally I think there's a lot of exciting stuff to be done using resynthesis and formant manipulation. Imagine an instrument that can take a bottom C from a piano, a middle C from a violin and a top C from a human voice, then automatically create a harmonically correct morph across the keyboard, and you get some idea of what's around the corner. Life is too short for traditional sampling, keyboard mapping and multisample looping, but a system like the one described could give everyone access to both real and abstract sounds with 'point and click' ease. What's more, once a sample can be broken down into its formant and non-formant components, it should be possible to resynthesize all the in-between stages from samples of just the highest and the lowest notes of the instrument, without compromising realism of tone or introducing keygroup discontinuities.

So far I've been theorising as to where the future of the sampler might be heading, but evolution is a parallel process and things haven't been standing still in the area of hard disk recording either. Don't be surprised if the next year brings us MIDI + Audio sequencers that can also handle traditional sampling, and as software tends to evolve faster than hardware, the pace of change could hot up very quickly. Interesting times indeed!

Paul White Editor

 

Printed from the Sound On Sound Website - sospubs.co.uk
The contents of this article are subject to worldwide copyright protection and reproduction in whole or part,
whether mechanical or electronic
are forbidden.