Latest / Electronic Music / History Of Samplers

Transcript

History Of Samplers · Electronic Music

• Hello and welcome to the Sound On Sound podcast channel about Electronic Music and all things synth. I'm Oli Freke and in this episode I'll be talking about the history of the sampler, the electronic instrument that revolutionised music production in the early 1980s by making it possible to digitally record and play back any sound imaginable - acoustic instruments, vocals, found sounds and even other previously recorded material. 44.1 kilohertz, the orchestral stab, 8bit, Fairlight, Synclavier, the Nyquist frequency. The sampling revolution has given us iconic terms and concepts and its impact on music production was just as profound as the synthesiser and drum machine revolutions that preceded it. Over the course of this podcast, we'll trace the samplers evolution from early pre-digital efforts to the digital breakthroughs of the late seventies, the explosion of affordable sampling in the mid eighties, the peak of their popularity in the nineties and their gradual demise in the early two thousands with the rise of computer-based production software. Let's dive in. Before the digital revolution of the 1970s, instrument designers were already striving to manipulate and reproduce real world sounds in what we might call a sampler-adjacent manner. While technology could not match their ambitions, the early to mid 20th century saw the development of some fascinating proto samplers. The best known of these is undoubtedly the Mellotron of the 1960s. Pressing a key triggered the playback of a short tape recording, one per key of an instrument or voice. The recordings were limited to around eight seconds, but it was one of the more effective of the pre-digital machines and easily the most widely used. The Mellotron was actually a UK adaptation of the earlier Chamberlin. Invented by Harry Chamberlin in the late 1940s. A rogue employee by the name of Bill Fransen took two Chamberlins to the UK in 1962 without Harry Chamberlin's knowledge or permission, but which led to Fransen forming a partnership with Birmingham based company Bradmatic Limited and the creation of the Mellotron. A licensing agreement between Chamberlin and Bradmatic eventually followed, perhaps a foreshadowing of some of the intellectual property issues that would become common in the history of sampling, as we'll see. But despite its controversial origins, the Mellotron became far more well known than the original Chamberlin, most famously being used for the flute intro of the Beatles Strawberry Fields Forever in 1967. But the story of sampler-like instruments goes back further. Frederick Sammis' singing keyboard of 1936 used the same principle of one recording per key as the Mellotron, but used strips of 35 millimeter film and optical electronics for sound reproduction. It was apparently designed for use in the film industry for playing in sound effects and foley rather than as a musical instrument in its own right, but Sammis also created the fabulously named Radio Organ of a trillion Tones in 1931, which also used strips of film but in this case, to play back hand drawn wave forms of instruments such as horns, violins and the human voice. And even earlier than that, in 1913, American inventor Melvin Severy proposed using magnetic tape or iron filings to record and reproduce instrument waveforms in an organ's tone wheels in his patent for a sound producing device. Perhaps the closest pre-digital machines came to modern sampling was in the experimental musical form of Musique Concrète, pioneered by Pierre Schaeffer, Pierre Henri and others at the Groupe de Recherche de Musique Concrète Studios in Paris during the 1940s and 50s. Beyond the standard tape manipulation techniques of cutting splicing and looping, Schaeffer developed some remarkable machines. The Phonogène Chromatique could play back a single tape recording at variable speeds from a keyboard, a Mellotron that used just one tape recording instead of 35. And even more impressively, the Phonogène Universal used a rotating angled tape head, like a VHS tape recorder called the Springer Temporal Regulator to achieve time stretching and pitch shifting effects, a remarkable achievement for the time. While fascinating, these pre-digital inventions weren't true samplers in the modern sense. That required digital technology, which brings us to the 1970s and the first genuine digital sampling instruments. The key technology needed was, of course, the transistor invented in 1947 at Bell Labs and its subsequent miniaturisation into microchips. These chips containing ever increasing numbers of transistors, enabled digital memory processes and ultimately, software, which could be programmed to do interesting things. Early computers were generally the preserve of government institutions, universities and the military. However in the UK, a company called Electronic Music Studios, EMS, led by Peter Zinovieff, pioneered the creation of a digital music studio in West London in the 1960s using two of these early computers. This groundbreaking studio was dedicated to digital sound synthesis and the creation of some of the earliest music software. Money was always an issue though and to raise funds, EMS released a three oscillator analogue monosynth in 1969 they called the Voltage Controlled Studio Version Three, or VCS3, inadvertently creating a future classic that's still revered to this day. The EMS Digital Studio revolved around two digital equipment corporations, PDP-8 minicomputers and featured several innovative instruments. Of particular interest is their Fast Fourier transform analyser, analogue to digital converter and digital oscillator bank. The 500 band pass filters of the analyser could be used to describe the harmonic profile of any input sound and be used for potential reconstitution by the digital oscillator bank. However, this process seems to have been used more for creative sound design by composers such as Harrison Birtwhistle than something we'd consider a modern sampler. Of course, other institutions were also exploring digital audio territory. For example, the University of Illinois' PLATO computer by the 1970s also had sampling and digital audio storage capabilities. Although these room size systems were expensive, exclusive and relatively inflexible for musical creation, they were vital pioneers easing open the door for what was to come. So let's cut to the chase. The first commercially available digital sampler used on commercial recordings was the Computer Music Melodian of 1976. Like the EMS system, it was based on the PDP-8 minicomputer and was invented by Harry Mendell at the University of California's Annenberg Center for Communication. The Melodian boasted impressive specs for its time, 12 bit sampling at up to 22kHz, a level not commonly seen again until the mid eighties. Stevie Wonder, always at the forefront of music technology, is closely associated with the Melodian. This is due to his use of it on the documentary soundtrack Journey Through the Secret Life of Plants, notably sampling birdsong in one composition from which a melody is created. While the exact number of Melodian's produced or sold remains unclear, Wonder used the system again for the songs for the 1984 film The Woman in Red, including his classic, I Just Called To Say I Love You. Harry Mendell himself even received a credit on Bon Jovi's 1986 album Slippery When Wet. Despite all this, information about the Melodian is scarce. Mendell claimed to have licensed some of its technology to Yamaha for their audio microchip production, suggesting a link to their mid eighties SK series of home keyboards with sampling functionality. I happened to own a Yamaha SK-5, so I took a look inside and couldn't really see any evidence of this, but I'm not a forensic electronics investigator so I can't really comment on the claim with any authority. Mendell himself later left the music technology field contributing to the Unix operating system, amongst other activities. Although the Melodian was clearly an instrument of some capability, it didn't evolve any further and is something of a stub in the history of sampling. Whilst on the subject of interesting stubs, in 1978, the band Steely Dan, somewhat surprisingly, also found themselves in possession of one of the earliest sampler and hard disc recording systems. Whilst in pursuit of their vision of perfection, the driving force behind the band, Walter Becker and Donald Fagan mentioned they'd like to be able to move the drum tracks around independently for greater accuracy. The story goes that their sound engineer, Roger Nichols said, give me $150,000 and I'll do it for you. Six months later he came back with what was essentially a 12 bit sampler built around one of the very early digital computers of the time. They called it Wendel and it was used on Steely Dan's Goucho album of 1979 on the title track and on Hey Nineteen. There were a couple of updates to the Wendel system in the following years, including a 16 bit Wendel-II in 1980 and a sample playback version called the Wendel Jr. in 1984. All quite remarkable for the time and the various incarnations of the machine were used on numerous records by artists as varied as Duran Duran, Supertramp, Paul Simon, Pink Floyd and many more. All very interesting, but not many people beyond the high-end studio world would've had access to it or have even heard of it. So, we shall now turn to one of the most famous samples of all time and with all due respect to the Melodian and the Wendel, the first to make a real impact on music of the late 20th century. I am of course talking about the Fairlight computer musical instrument of 1979, or Fairlight CMI for short. Founded in Sydney, Australia in 1975 by schoolfriends Peter Vogel and Kim Ryrie, Fairlight evolved from a digital synthesizer called the Quasar M8, designed by Tony Furse. The M8 built with two 8 bit Motorola 6800 microprocessors was an additive and FM synthesiser and already used the soon to be iconic light pen for screen input. The company itself was named after the Fairlight Hydrofoil that applied its trade across Sydney Harbor. But by 1978, Vogel and Ryrie were still struggling to make satisfying, realistic sounds on their Quasar M8. Inspiration struck however, when they thought to analyse the harmonics of a piano by recording the sound of it into the M8 and then trying to recreate those harmonics for a more realistic sound. But instead of analysing the harmonics as planned, Vogel found that if he actually just played back the recording at different speeds, it sounded more realistic than any synthesised piano. He said in 2005, "it sounded remarkably like a piano, a real piano. This had never been done before. By today's standards it was a pretty awful piano sound, but at the time it was a million times more like a piano than anything a synthesiser had churned out. So I rapidly realised that we didn't have to bother with all that synthesis stuff. Just take the sounds, whack them in the memory and away you go". This realisation then, became the foundation of the sampling Fairlight available in 1979 for £12,000 and which became an overnight sensation, for those who could afford it, given that that kind of money could literally buy a house in London at the time. But anyone who could afford to buy a seven inch single in the early 1980s could easily hear the results. Peter Gabriel established the company Psyco Systems to import Fairlights to the UK and David Vorhaus, who will come back to shortly, was the lucky recipient of the first unit. John Paul Jones of Led Zeppelin was also one of the earliest purchasers of a Fairlight for use as a modern version of his beloved Mellotron. Peter Gabriel was also the first to release an album to feature the Fairlight, which was his third album released in 1980 and Peter Vogel is credited on the sleeve notes accordingly. Other early adopters include Trevor Horn, who used it extensively with Jeff Downs and The Buggles. Later Trevor Horn worked with Fairlight programmer, JJ Jeczalik on Malcolm McLaren's Duck Rock album of 1983. Anne Dudley also contributed to Duck Rock and some of those sessions ended up on The Art of Noises first album, Who's Afraid of the Art of Noise in 1983, The Art of Noise being a band comprised of initially JJ Jeczalik, Anne Dudley and engineer Gary Langan, produced by Trevor Horn, aided by the conceptual shenanigans of journalist Paul Morley. Their work is notable for its extensive and early exploration of creative sampling. Trevor Horn also produced with Jeczalik and Gary Langan, the album 90125 by Yes in 1983. The song Owner of a Lonely Heart may be the first song to feature a sample of another band's drum break, which in this case was from Funk Inc.'s Kool Is Back. Other notable users included Yellow, Alan Parsons, Richard Wright, Thomas Dolby and later Jean-Michel Jarre, who used it extensively on 1984s Zoolookologie. Continuing the roll call of 80s megastars, other high profile musicians to use it include Duran Duran, Kate Bush, Herbie Hancock, Stevie Wonder and all this from an 8 bit machine with only a quarter of a second sample memory courtesy of its 16 kilobyte memory. Lower sampling rates down to 8kHz could extend this to one second,and Vogel and Ryrie embraced the resulting aliasing as part of the instrument's character and didn't expend too much effort in trying to filter it out. The Fairlight series two costing £30,000 arrived in 1982 and saw the maximum sample rate increased to 32kHz and which also introduced the famous Page R graphical sequencer page, replacing the previously cumbersome music composition language. The upgraded Fairlight II X in 1983 received full MIDI implementation. 1985s Fairlight Series III was the version to offer full professional digital audio of 16 bit, 44.1kHz sampling and upgraded the memory capacity to a full 14 megabytes. The light pen was replaced with a more ergonomic tablet and stylus due to users complaining sore arms, having to hold the light pen against the screen all the time, though it did have a reputation for being less reliable than the previous series II. The film composer Brad Fidel used a series III for his score for Terminator 2 and created the iconic Terminator T-1000 sound by pitching down the library sound Brass Fall. Here's a rough approximation of that process and is maybe the perfect example of how the character of a sound can be radically transformed by the simplest of sampler processes, the pitch shift. Brass Fall was one of the many iconic sounds in the Fairlight library, which also includes arguably the most famous and most used sample of all time, ORCH5, or ORCH2, depending on which version of the library you had. ORCH5 is a sample of the opening chord of the Infernal Dance of the Demonic Magician Kaschei, the fourth movement of Stravinsky's Firebird. The version sampled was from a Classics for Pleasure album played by the Philarmonia Orchestra and conducted by Joseph Cripps in the 1970s. Little did they know that this was to become one of the most heard orchestral moments in recording history and not in the context they might have expected. It was recorded by David Vorhaus, musician and Fairlight programmer extraordinaire at Peter Vogel's house one afternoon in 1980. In those days people frequently contributed their sounds to a kind of informal Fairlight library, which was then distributed to other users and ORCH5 was one such of these. Another sample sound that saw remarkably wide use was ARR1 or SARARR, again, name, depending on the library version. This was a sample of singer Sarah Cohen made in 1980 by Tom Stewart, future Fairlight employee and this one has also been heard on innumerable tracks over the years, including the Art of Noise's Moments In Love, Tears For Fears Shout, Yan Hammer's, Crockett's Theme and many, many more. Despite this impact, by the mid eighties cheaper samplers were emerging thanks to the continuing reduction in costs and increase in power of digital technology and the Fairlight's high price became unsustainable. Fairlight ceased trading in 1989, reportedly never having turned a profit. In 2009, Peter Vogel launched a 30th anniversary edition, the CMI 30A but unfortunately a trademark dispute limited the production run to just 14 units out of the planned 100. Now that we've introduced the world's first influential sampler, let's take a moment to discuss the fundamentals of digital sampling. As you likely know, a digital sampler records sound into computer memory and allows it to be replayed, often at different pitches to recreate existing instruments or to create entirely new sounds. A sample is recorded by passing an audio signal through an analogue to digital converter, an ADC, which transforms it into a string of numbers stored in memory. A digital to analogue converter, a DAC, then reverses this process to reproduce the sound. Two key parameters define the numbers that represent the sample, the sampling rate, the number of samples taken per second, measured in hertz, and the amplitude loudness of each sample represented numerically. The red book CD standard for digital audio specifies a 44.1kHz sampling frequency at a 16 bit depth. 44.1kHz means the audio is sampled 44,100 times a second and a 16 bit depth means that each sample is stored as a 16 digit binary number, i.e. a sequence of 16 zeros or ones. This translates to two to power 16 or 65,536 discrete amplitude levels, providing a dynamic range of approximately 96 decibels. This is sufficient to perfectly reproduce audio for human hearing. 12 bit sampling was common in the mid eighties and offered 4,096 levels with an equivalent 60 decibel dynamic range, whilst 8 bit sampling before that provided only 256 levels and a 49 decibel dynamic range, which was the original sample bit rate of the Fairlight CMI. Equally, the sample frequency in the early days was really as high as 44.1kHz, more usually around 20 to 30kHz and often lowered to increase sample time. To make that more tangible, let's listen to a sample at some different bit rates and sampling frequencies to get a feel for the improvement in quality between the Fairlight of 1979 and something like the Akai S1000 of 1988 with its modern spec of 16 bit, 44.1kHz. This is 8 bit at 10kHz, a low Fairlight I sample rate. This is 8 bit 24kHz, the maximum Fairlight I sample rate. This is 12 bit 24kHz, a fairly standard mid-eighties sample rate. and this is 16 bit 44kHz, the standard set by the Red Book to accurately reproduce a recorded sound. I do realise you're listening to this on an MP3 podcast, which means there's yet another layer of sample rate conversions going on, but I hope the general principle is clear. Many of you listening will know of the Nyquist Chaning sampling theorem, which states that the sampling rate must be at least twice the highest frequency you want to reproduce. Since human hearing extends to a maximum of roughly 20kHz, the industry standard of 44.1kHz provides sufficient headroom. But why such an unusual number? Why not a more computer friendly 48kHz instead of this 44,100? The answer lies in early digital audio storage methods. In the 1970s devices like the Sony PCM F1 stored digital audio on Beta Max videotape. These systems recorded audio by representing binary data as black and white regions on each video line. Given the tape speed, it was possible to write three 16 bits samples per video line, and there were 245 usable lines per frame of video. So the number of samples in one frame was three times 245, which is 735. And given Beta Max has a video frame rate of 60hZ, that means that in one second, 60 frames of 735 samples gives a resulting audio frequency of 44,100. This video based standard became the basis for the CD Red Book specifications, although 48kHz did later emerge as a secondary professional standard. OK, let's move on to another giant of the early sampling world. Like Pepsi and Coke, there had to be another sampler yang to the Fairlight's Ying and indeed there was the New England Digital Synclavier created in 1977 by Professor John Appleton and his team at Dartmouth College, New Hampshire, the Synclavier began as a software tool for pitch recognition, but quickly evolved into a powerful digital synthesiser. Even in its early pre-sampling iterations, the Synclavier was visually distinctive with its illuminated buttons and prominent LED display. At this time, the synthesiser unit was separate from the keyboard and sometimes supplied without one. Powered by five Synclavier synthesiser cards, these early systems were 8 note polyphonic with 8 track sequencing and initially sold to other universities. The Synclavier II released in 1980 continued the digital synthesiser focus. However, recognising the success of the Fairlight, New England Digital introduced sampling capabilities two years later and this was high quality sampling, 16 bit resolution with the option to sample as high as 100kHz. This technology came at a significant price, $25,000 for the base unit, which is approximately $100,000 today, with additional options that could boost the cost to as much as $200,000. The Synclavier also pioneered direct to disc recording, positioning itself as a complete digital recording studio. This bypassed the limitations and expense of RAM, allowing users to record directly to Hard Disc. This configuration was known as the Synclavier Digital Recording Tapeless Studio to users and the Direct to Disc multi-track recording system by New England Digital. Later models were controlled by an Apple Macintosh and had optical drives for accessing large sample libraries. The high price was justified by New England Digital at the time by comparing it to the equivalent cost of hiring an SSL or Neve desk, a 24 track studio tape machine and buying a Bösendorfer piano, which I guess is true up to a point. Direct to disc recording was a great innovation though as a 16 megabyte RAM card cost £14,500 in the UK, even in 1989. By that year the Synclavier system offered 96 note polyphony, 16 track hard disc recording and up to 200 tracks of sequencing on the Mac. It also incorporated advanced features like Additive 3 Synthesis. A full system in 1989, including a Macintosh 2 in peripherals cost around £106,000 or $180,000. Like the Fairlight, the Synclavier found use among wealthy musicians and studios and had a significant impact on music production. Notable uses include the intro notes of Michael Jackson's Beat It, which is almost certainly a note-for-note replay of an original sequence included on a Synclavier II demo disc by Denny Jaeger, though this is one of the machine's additive synth sounds not a sample. Trevor Horn in his productions for Frankie Goes to Hollywood, Grace Jones, Propaganda and Yes was able to make extensive use of a Synclavier courtesy of producer / programmer Steve Lipson and Kraftwerk also showcased the high-fidelity sampling capabilities on their 1986 album Electric Cafe. Suffice to say along with the Fairlight, the Synclavier II played a key role in defining the sound of the early 1980s. These systems, combined with sample-based drum machines, such as the Linn Drum and digital synthesizers, such as the Yamaha DX7, ushered in a significant shift in music production and helped create what we now think of as the glossy sound of the 1980s. However, nothing is constant in the world of technology and like Fairlight, the Synclavier's reign came to an end in 1993 as the rise of cheaper samplers and more affordable digital recording options came onto the market. Despite this, the legacy endures with the Synclavier Go iOS app being created in 2019 and the Synclavier Regen FM synth module of 2023 demonstrating continued interest in this pioneering instrument. While the lucky few enjoyed the capabilities of the Melodian, Fairlight and Synclavier, studio engineers and producers found a more accessible way to incorporate digital samples into their work in the early 1980s, the digital delay line. By recording digital audio in and delaying its playback via RAM for a few fractions or multiples of a second, a new device was created, the digital delay line or DDL. Emerging in the late 1970s, it was soon notice that this opened up possibilities beyond simple delay effects. In 1978, UK firm AMS released the DMX 1580 digital delay line. By 1979, this device had gained pitch shifting capabilities and by 1980 offered stereo delay. Crucially, in 1981, it acquired manual sample triggering, referred to as lock-in. This effectively transformed into a sampler, allowing producers to record audio and trigger it on demand, a technique referred to as spinning it in. Producers like Martin Hannah of New Order and Steve Levine of Culture Club, used this method to trigger drum hits and backing vocals. The French Publison DHM 89 B 2, also released in 1978, offered similar digital delay and pitch shifting functions. It even had an optional keyboard available a couple of years later, the KB2000, which allowed it to function much like a modern sampler, playing back samples chromatically. The Bel BD-80 DDL launched in 1984 provided 16 seconds of memory and CV input for pitch control. Sound On Sound's very own Paul White, writing in Home and Studio Recording magazine in 1984 remarked that the basic two seconds of memory will be more than ample for most users, many of whom will only want to sample percussive sounds which have a relatively short duration anyway. And it's true, no one has needed more than two second sample time since, sorry, only joking Paul. The Japanese band Yellow Magic Orchestra are reported to have used a Toshiba LMD-649 custom made sampler on their 1981 album Technodelic, which would certainly qualify as one of the earliest commercial recordings to use a sampler. Despite the use of rudimentary samplers, these DDLs are often overlooked in sampling history. Perhaps this is because they lack the synthesis features, filters, envelopes, LFOs and so on of later samplers. They were essentially simple sample playback devices, but clever nonetheless. The American company EMU Systems entered the game in 1981 with the Emulator, which bridged the gap between the high-end Fairlight's and Synclavier and the simpler DDLs. The Emulator actually launched before the Synclavier even had sampling capabilities and it offered 8 bit sampling with a 20kHz sampling rate for $8000, which is roughly $23,000 today and it had 120 kilobyte sample memory, allowing for two seconds sample time. This wasn't pocket change, but significantly less than the Fairlight's $30,000 price tag. Consequently, if found favour with a new range of artists such as Tangerine Dream, Tony Banks and New Order - check out New Orders infamous performance of Blue Monday on Top of the Pops in 1984 to see it live on stage. Vangelis also used an Emulator on the Blade Runner soundtrack for the timpani and crash cymbals amongst other things, alongside his much vaunted use of the Yamaha CS80 Polyphonic analogue synthesiser. Stevie Wonder, ever the early adopter, owned an Emulator as well, serial number one of course. However, it had its limitations. The basic model was only duophonic and it only allowed two samples at once, which had to be assigned to separate halves of the keyboard. There were more expensive four voice and eight voice options available and multi-sampling was also only available by another expensive software upgrade. Building on their experience with microchips, EMU addressed these issues with the Emulator II launched in 1984. This iteration boasted a 27kHz sample rate and had resonant analogue filter chips supplied by longtime collaborators SSM. This addition of filters followed modifications made to Emulators by early adopters such as Michael Jackson's keyboard player Michael Boddicker and EMU's Jim Cooper, who offered such hardware modifications to the first version. It also had other synthesiser-like capabilities, including LFOs, envelope generators, velocity sensitivity and true multi-sample support. This was the beginning of samplers exhibiting the sort of sound manipulation capabilities long enjoyed by synth players and despite technically being an 8 bit sampler, the Emulator II employed a clever compounded 8 bit system whereby the difference between adjacent sample values was recorded instead of their absolute values, improving performance and sound quality to something approaching a 12 bit system. Interestingly, the initial 25 Emulator IIs were sold in the UK due to a deal with Peter Gabriel's Syco Systems, which secured the funding required to actually manufacture the instruments and of course, Peter Gabriel's hit song Sledgehammer features the Emulator II's famous Shakuhachi sample. Paul Hardcastle brought us the creative joy of the sample stutter with his use of the Emulator II in 1985. Of course, for copyright reasons I can't play the original song, but 19 was a classic song of the era. Later models came with a more convenient 60 megabyte drive replacing the clunky 5¼" floppies, and by 1987 the Emulator III reached the full professional standard of 16 bit, 44.1kHz. While it cost a not insubstantial $12,000, this was still a fraction of the cost of a Fairlight or Synclavier. EMU didn't just cater to the high-end market, though. The Emax of 1986 and Emax II of 1989 offered 12 bit and 16 bit sampling respectively with the trademark analogue filters at a more accessible price. In 1994, the ESi32 arrived - a 16 bit sampler boasting sophisticated multimode filters, 32 note polyphony and eight outputs. Not content with this range, EMU also released an influential sampling drum machine, the SP1200, in 1987, which also deserves a mention. Its 12 bit grit and ease of use made at a favourite of mid-eighties hip hop producers. These producers actually were responsible for uncovering many of the famous drum loops from old funk and soul records that we are still so familiar with today. The New York DJs of the late 70s had noticed how their dancing audience loved the breakdowns and records that featured drum solos. So these breakdowns with beats or breakbeats then became early targets for sampling. Thus today, such drum breaks as The Incredible Bongo Bands Apache, James Brown's Funky Drummer, Lyn Collins Think About It and of course, of course, the Winston's Amen, Brother are now staple sounds of the modern musical era in genres such as hip hop, big beat, breakbeat, jungle, drum and bass, electronica and beyond. By 1994, EMU launched the fourth generation of Emulator with 128 voices and expandable memory up to 128 megabytes. The series continued until the Emulator IV XT Ultra in the late 1990s, representing the pinnacle of their hardware sampler range and was about as powerful as hardware sampling got. The extensive sound libraries built up over this time found new life in the Proteus sample playback synths that emerged in the early 1990s, but the rise of personal computers in the late nineties ultimately rendered external samplers like the Emulator obsolete leading to the series end in 2002. Okay, let's back up a bit again to the mid-eighties when thanks to the continued reduction of microchip prices, a wave of more affordable samplers emerged, bringing this technology to a wider audience. These instruments, while often compromising on sound quality, frequency response and memory compared to the high-end systems, laid the groundwork for the affordable sampling revolution of the late 80s and early 90s. The first of these, the evocatively named Mirage from American firm Ensoniq released in 1984 was a landmark in accessible sampling. Priced at just $1695, this 8 bit sampler offered a respectable two second sample time at its top rate of 33kHz thanks to its mighty 128 kilobytes of memory. It also boasted eight note polyphony analogue filters using Curtis chips, multi-sampling and multi timbrality. The Mirage's affordability was partly due to its custom chip, the Ensoniq ES5503 digital oscillator chip, also known as the Q-Chip, designed by Robert Yannes. This chip also found use in other Ensoniq instruments like the ESQ1, SDP1 and SQ80, as well as in the Apple IIGS. Yannes had previously designed the iconic SID chip for the Commodore 64, the sound of which is still popular in retro and chip tune music today. Notable recordings featuring the Mirage include Janet Jackson's Rhythm Nation 1814 and Control and Public Enemies Rebel Without a Pause. The Mirage reportedly sold 8000 units in its first year and 30,000 overall over the course of four versions and whose success was built on in 1988 by the Ensoniq EPS, which whilst only 13 bit, did offer 20 voice polyphony. Several other established synthesiser companies also ventured into the sampler market during this mid to late 80s period, though their efforts were generally less impactful and only ran to a few models each, if that. One of the earliest was the PPG Waveterm designed for the PPG Wave and was available from 1982. The Waveterm was a complete computer system that included 8 bit sampling capability. The Waveterm B released in 1982 upgraded this to 12 bits. However, it was more akin to the high cost systems of Fairlight and Synclavier, with a complete system costing upwards of £10,000 and the PPG Wave series remains better known for its revolutionary wave table synthesis than its sampling. In 1986, Korg released the 12 bit DSS1 and DDD1, a drum sampler. Casio introduced the 16 bit FZ-1 in 1987, although its sampling rate topped out at 36kHz, which is not quite the standard 44.1kHz. Casio also released the SK-1 and SK-5 toy samplers in the mid 80s, bringing sampling technology to children even before it was widely accessible to the recording industry. Yamaha's offering was the 12 bit TX16W in 1988, which failed to make a significant impact and sampling was never really a focus for this otherwise dominant Japanese firm. Sequential Circuits released the Prophet 2000 in 1985. This was a 12 bit sampler with an eight note polyphony and up to 30 second sample time with the full 512 kilobyte memory. An upgraded Prophet 3000 was released in 1987 just as Sequential Circuits hit financial difficulties, ceased trading and was purchased by Yamaha who didn't want the Prophet 3000 to compete with their TX16W, so was withdrawn from sale. Also worthy of a mention is a little known British company called Greengate who released a sampler system, the DS3 in 1985. The founders Dave Green and Colin Holgate built the system around an Apple II+ and supplied an audio card and software to run the sampler on. The 8 bit system could sample one second at 16kHz and had a mildly impressive four note polyphony. It was functional and whilst much cheaper than a Synclavier running off an Apple Mac, this small UK firm couldn't compete with the rise of the mid-80s hardware sampler and failed to make much of an impact. So by the late 1980s, the mainstream sampling market was dominated by EMU, Ensoniq, Roland and Akai. We've discussed the American firms EMU and Ensoniq so let's now turn our attention to the Japanese sampling scene dominated by Akai and Roland. Akai was the leading sampler manufacturer from the late 80s into the new millennium, their machines becoming studio staples worldwide. However, the first sampling unit 1985s S612 was a relatively modest 12 bit rackmounted machine with 32kHz sample rate and only one second of sample memory and it held only one sample at a time, which was loaded via a quick disc floppy drive, although it did offer six note polyphony. Interestingly, the S612 was partly designed by David Cockerell, former of EMS and later of Electro Harmonics. At Electro Harmonics. Cockerell had worked on a sampler based on the company's digital delay line pedals called the Instant Replay and Super Replay. When the company faced financial difficulties, he took the idea to Akai. Cockrell continued to work on the Akai S900s, S950 and S1000 samplers and also contributed to the electronic circuit sign of the Akai MPC60 a sampling and compositional workstation whose form factor remains in production today. But it was the Akai S1000 of 1988 that truly brought sampling studios and home setups worldwide and especially in the UK and Europe. The S1000 was the first Akai sampler with 16 bit 44.1kHz stereo sampling at a price that ensured mainstream success. It was ubiquitous in studios from the late 1980s and well into the late 1990s. The S1000 offered a full synth engine with envelopes and LFOs and comprehensive if somewhat complicated MIDI control, multi-sampling, velocity switching and sample cross-fading made it the defacto standard for professional samplers. Its defining time stretching feature was a software update soon after the initial launch and its distinctive design, the raised lower portion, dual rows of buttons beneath the large LCD, data entry wheels and vertical three and a half inch floppy drive is still very familiar to many producers to this day. Following the S1000 success, Akai released several variations and upgrades, including the S1100 in 1990, the S3000 in 1992, and the top of the line S6000 in 1998. However, in what is becoming a familiar narrative, Akai discontinued the S series samplers in the early two thousands as personal computers and digital audio workstations took over. Bucking that trend though, the Akai MPC format remains popular offering a non-computer based workflow that has found a dedicated following among producers from DJ Shadow whose Introducing album was largely created on an MPC to the highly influential work of hiphop Producer J Dilla. By the late 1980s, Roland were also offering high quality samplers. The SM70, their flagship model was released in 1989 and was followed by the S750 and S760 in the years that followed. Whilst these were highly respected and widely used, Roland never quite achieved the same close association with sampling as Akai and EMU. This was possibly due to Rolands broader product range compared with the specialism in samplers of Akai and EMU. A notable feature of Roland's S series samplers was their ability to connect to external monitors, simplifying sample editing and library management. This addressed the limitations of working with small LCD displays. Recognising that precise sample editing on a small display was far from ideal, summarised by the popular observation that editing a sample on a 2x16 LCD was like trying to do heart surgery through a letterbox. The 1990s then were a golden age for hardware samplers. Top of the line machines from EMU, Ensoniq, Akai and Roland dominated the market and each year's new models brought increased sampling time, features and ever more flexibility. However, this couldn't last. As mentioned several times already, the rise of the computer audio workstation signalled the decline of the hardware sampler. The computer gave rise to the software sampler, which often functioned more like a rompler, focusing on the browsing and loading of commercial sample packs. This became the favoured use case rather than creating one's own sounds from scratch and creatively sampling from a wide range of sources, whether they be real world found sounds or liberating an old recording and breathing new life into it. While the possibilities expanded exponentially, it could be argued that some of the art of creative sampling had been lost. Either way during the 2000's, all the major manufacturers ceased production of hardware samplers, with the notable exception of the Akai MPC series. But hold on, did I just say liberating old recordings and breathing new life into them. I believe I did. And that, of course is another whole side to sampling. Unlike synthesisers where the creation of sound is spontaneous, a sampler allows musicians and producers to sample pre-existing recordings, which has been creating legal complexity for decades now. Numerous lawsuits have arisen from the unauthorised use of copyrighted material, starting in the early 1990s as commercially successful records began incorporating obvious samples. An early landmark case in 1989 in the US concerned De La Soul's use of The Turtle's song You Showed Me on their song Transmitting Life From Mars. Though settled out of court, it established a precedent that permission should be sought before using samples. This was firmly reinforced in the US at least in the case of Biz Markie and his use of Gilbert O' Sullivan's Alone Again, Naturally. This case went to court and was found in favour of O' Sullivan, establishing the need for sample clearance. Biz Markie's follow-up album was called All Samples Cleared. However, the legal landscape isn't always so straightforward. In 2016, Madonna was sued for using a horn sample from The Salsoul Orchestra's Ooh, I Love It (Love Break) in her song, Vogue. In this case, the court ruled in Madonna's favour under the de minimis doctrine, stating that the sample was too minimal to require a license. This ruling referenced the earlier case of Newton versus the Beastie Boys, where a six second looped sample of James Newton's flute and voice piece Choir was repeated extensively in the song Pass The Mic. This usage was deemed not infringing and I quote the judge here "Because the average listener would not be able to discern the role played by the part in the new song and as such, its usage is insubstantial". Suffice to say the law regarding sampling is complex, and my non-binding and non-legal advice is don't use uncleared samples on your next mega hit if you want to keep the royalties and not spend years in court and spend a lot of money on lawyers. Whether or not this has a chilling effect on musical creativity and art, I shall let you be the judge. In summary, then, sampling has been a transformative force in music production for nearly 50 years and has been largely driven by the democratisation of technology through increasingly accessible and powerful computing. Once the exclusive domain of high-end producers in the early 1980s, the sampling power that now resides in our laptops and even our mobile phones vastly surpasses the capabilities of those early Fairlights and Synclaviers. This accessibility contributed to the decline of the hardware sampler industry itself as computers gained greater power, memory and affordability. The once ubiquitous racks of hardware, including the iconic Akai S1000, are now largely a thing of the past. Yet the use of samples, whether professionally produced sample packs or personal field recordings, has become pervasive across the musical spectrum from pop and rock to film schools, video game soundtracks and beyond. The hoary old adage, your imagination is the only limit has never been truer, particularly considering subsequent innovations like granular synthesis and other digital audio manipulation techniques that have evolved from sampling. The resurgence of hardware modular synthesis, especially in the Eurorack format over the last 25 years, also offers new avenues for rediscovering the joys of hardware sampling. So we shall see where the current and future generations of creative musical minds takes us with that old new idea of snatching sound waves in the moment from the very air and bending them to their creative purposes. Thank-you for listening and be sure to check out the show notes page for this episode where you will find further information along with additional web links. Before you go, make sure you visit the Sound On Sound podcast page at soundonsound.com/podcasts where you can explore all the other great content playing across the other channels. I'm Oli Freke and this has been a synth evolution production for Sound On Sound.