Emu Systems Emulator II, Part 2 (Retro)

8-bit Sampling Keyboard, Part 2 (Retro)

Reviews : Keyboard

Paul Wiffen continues his homage to the first 'musical' sampling keyboard, which brought sampling out of the dark ages of simple trigger playback and into the realm of filters, envelopes and performance controls which we take for granted on all modern samplers. The Emulator II was, as he explains, also responsible for other major innovations like on-board hard drives and CD-ROM libraries...
 
Last time I told you about how, despite being deeply cynical about sampling after an initial disappointment, I fell in love with the Emulator II, realising that it wasn't sampling itself, but the implementation of the technology that had been the problem. All previous samplers had contented themselves with allowing you to make a sample, loop it and trigger it from a keyboard. This was OK for certain sounds, particularly if sequenced from something like the Fairlight's Page R, but it hardly allowed you the expression necessary for a musical performance, especially if you had any traditional keyboard skills. For the first time, the Emulator II allowed you to use your performance to control the way the sample played back in real time. It was the first pure sampler with a velocity-sensitive keyboard, which could not only be used to control the loudness of the individual notes but also, thanks to the analogue filters included with each of its eight voices, the brightness of the sample as well. And if you couldn't use a filter to make the change you wanted in a sound, you could use velocity to switch between two different samples which represented the different sounds your target instrument made in different playing situations.

Those of you who own a modern sampler probably think all of this is no big deal; most samplers these days will do this stuff without breathing hard. But for something to pass into the shared common pool of knowledge, someone has to discover it first. When Emu shipped the EII, the common received wisdom was that you didn't put analogue filters on samplers, and that there was no reason why you would want to use velocity to switch between two different samples. 

The Less Successful 'South Coast' Sessions Once Asia had released their Aqua album (on which I spent so much time programming the EII) to worldwide apathy, I moved on to other programming sessions on the EII. I was aided in this by registering with Programming People, an agency run by Karen Clayton, girlfriend of Gary Langan of the Art Of Noise. Due to her aggressive selling, I was going out as a programmer for much higher daily rates than I would ever have dared to ask for myself. Because of this, Geoff Downes asked me if I would get rentals for his EII on sessions I did for other people (once the Asia album was finished, it wasn't getting that much use). As a result, I became the first or second call EII programmer in old London Town (sharing the honours with Nick Magnus, another regular SOS contributor, who was also with the Programming People). When they couldn't get JJ from the Art Of Noise, they would settle for me and Geoff's EII. As a result, I got booked on some very prestigous sessions in 1985 including UFO, John Miles, Gary Moore, Tim Finn, The Bridge (Andrew Gold and Graham Gouldman, ex-10cc) and the soundtrack of Billy the Kid and the Green Baize Vampire (a Phil Daniels musical which disappeared without trace). More often than not, when I got to the sessions, I discovered that what they wanted couldn't be achieved with a sampler at all, and I would often have to delve into the boot of my car for an OSCar (for weird synth noises) or Synthex (for big synth pads). But two of the calls I got led to the most bizarre situations I have ever been in. The first was from Keith Emerson. I had already taken him some new synths to try out for the Emerson, Lake and Powell album but this time he wanted to experiment with sampling. I made a call to Geoff's technician, Pete Lorimer, and we were all set. I took the EII down to Keith's 14th-century manor house in East Sussex and set it up in the barn with all the other keyboards. When Keith appeared, I asked him what he wanted to sample. His fantastic grand piano? The ageing and failing Yamaha GX1? The answer took me completely by surprise. "I thought we'd start with my motorbike," he said, "and then move on to the chickens in the hen-house." Realising he was deadly serious, I set the mic up while Keith fired up 1000cc of noisy armageddon on two wheels. It wasn't long before we had the sound captured in the EII. Then came Keith's first disappointment. He immediately tried to play the sample like a standard synth sound, but because the sound had far more frequencies in it than the standard related harmonics of a musical pitch, all that came out was the blurred sound of half a dozen motobikes being revved at different rates. I tried using the EII's filter resonance to narrow down the frequency band and give it a perceived pitch. It helped a bit, but it was still not a sound anyone was ever going to want to hear on a record. Undaunted, 10 minutes later, Keith was crawling into his hen-house with a microphone (I had been worried that this lot was going to fall to me, but thankfully he was determined to capture the sound for himself). Once again, I sampled the sound into the EII, expecting the worst. I wasn't wrong! Once more, the chickens didn't exactly lend themselves to Keith's lightning keyboard style (though in fairness, I suspect that they would have defeated Gary Numan or Jan Hammer equally). At that point Keith decided that sampling offered nothing he needed.

Filters & Fades

So just what parameters did the EII offer that made it so revolutionary? The same sort of features that were standard on the analogue synths of the time, but which had not previously been supplied on samplers. This meant not only that each voice had an ADSR (Attack, Decay, Sustain and Release) amplitude envelope which controlled the loudness over time, but that each voice had a similar filter envelope. This meant that the low-pass filter could have its cutoff frequency shifted over time as well, and have this filter movement emphasised with Q (aka resonance). But a dedicated filter envelope was not the only way to shift the cutoff frequency: there was also keyboard tracking to make the cutoff dependant on whereabouts on the keyboard you played. But perhaps most importantly, velocity could not only be routed to the cutoff frequency (which lets you make soft keystrikes less bright and hard strikes brighter, imitating the way most physically generated sounds behave), but also to both filter and amplifier envelope attacks and even to the amount of resonance.

As a result, raw samples loaded into the EII could be made to play more expressively than previously possible, with the notes getting brighter and with a faster attack the harder you played. And if you couldn't achieve the required breadth of tone from filtering and enveloping a single sample, then you could always use two and velocity-switch or crossfade them. Velocity-switching was less expensive in terms of polyphony (it meant the EII voice was only ever playing one sample or the other, depending on whether your keystrike was above or below the switching threshold you had set). Velocity crossfading would determine the mix level of two samples from how hard the key was hit. It would, therefore, almost always be playing back two samples, except on really quiet and loud notes, so it took up two voices of polyphony for each note played (and remember there were only eight available to play with). For this reason, I would avoid velocity crossfading as much as possible, except on solo or bass sounds.

The same held true for positional crossfading, which was there to try and smooth the abrupt changes in timbre that could become apparent between the different samples in a multisample as you went up or down the keyboard. The theory was that by fading one sample out towards the upper (and lower) end of its keyboard range and bringing the next one up (or down) in inverse proportion, you would mask the sudden changes that would otherwise occur between keygroups. Remember that as a sample is sped up to achieve higher pitches, it gets brighter and when it meets the transposed-down sample at the bottom end of the next keygroup (which sounds duller because the frequencies present have been lowered), the contrast between timbres is exaggerated. Sometimes positional crossfading worked (although again it used two voices of polyphony for each note played) but more often than not it threw up another problem. Two samples from neighbouring ranges of an instrument would often contain such similar frequencies that there would be cancellation or even phasing. So in general it was better to try and find multisamples that inherently didn't jar too much as you moved from one keygroup to the next, rather than eating up polyphony in positional crossfades and risking unexpected interference.

Velocity and positional crossfading were of limited use for improving realism, then, but they really came into their own when you started to use them for other purposes. Crossfading flutes to pianos or guitars to xylophones (either via velocity or position) produced some amazing results, giving you hybrid instruments not heard before. Many a session was rescued by the inspiration given to the keyboard player by such a previously unheard sound.

The expressive possibilities were not limited to velocity and positional control of the samples. You could also route performance controllers like wheels, pedals and incoming MIDI controllers (the EII was the first sampler to implement MIDI) to previously unheard of destinations (on a sampler) like Filter Cutoff and LFO control of pitch and filter. This meant that introducing vibrato and/or tremolo was a breeze and that brass notes could appear to be 'blown' harder or softer during their sustain phase.

Other unique facilities for the time included the combination of gain stepping and continuously variable level on the input, meaning that it was rare to be fed a signal which you couldn't sample. Sampling could also be initiated automatically by the signal's reaching a threshold level, another feature which we take for granted on today's machines.

Sequencing & Arpeggiation

The EII also boasted a sequencer, but as it acted like a glorified tape recorder, it was strictly for players. There was no quantisation or groove templates back then, although you could cut and paste bars together or repeat them. I did once spend an afternoon with Paul McCartney, laboriously loading the backing track for an entire song into the sequencer, him playing in his sampled bass and drums and me editing the bars into the so

The EIII That Wasn't The most far-reaching effect of my McCartney sessions was not any of the sounds I did for him, but an in-joke that cropped up in an idle moment. Someone noticed that the white tape they used for naming the channels on the mixing console with a chinagraph was almost exactly the same width as the 'I's in the legend 'Emulator II' on the back panel. It was irresistible to stick an additional one on the machine so it now read 'Emulator III'. McCartney thought it was a splendid joke. It would probably have gone no further than the studio if McCartney hadn't decided to take it up to Abbey Road to use in the video for 'Spies Like Us' (title song for the John Landis movie with Chevy Chase and Dan Ackroyd, and the only hit to emerge from that period of McCartney's work). The 'EIII' was used for Chevy Chase to mime the keyboard part in the video, which was then put on heavy rotation on the newly launched MTV in the States. Unbeknown to us, Emu then started getting all sorts of calls from disgruntled EII owners in the US, saying things like "Hey man, I've been using the EII since it came out, how come you've started shipping the EIII in Europe without telling me?" In vain, Emu tried to explain that there was no such thing as an EIII (it would be several years before Emu actually shipped a product with this name). All they would hear back was, "I've seen it on McCartney's video, it's a different colour and shape completely!" (all from the addition of a short strip of chinagraph tape!). Eventually things got so bad that dealers were cancelling their scheduled orders of EIIs, because the customers all 'knew' that the EIII was coming. It ended up causing a major dip in Emu's sales in the US. In England, we were oblivious to all this until I went over to the NAMM show the following January. Mentioning to Service Manager Riley Smith on the Emu booth that I had been working with McCartney, I was dragged off to meet an accountant who wanted to throttle me because I had been involved. Obviously, they must have forgiven me eventually, because when the EIII really did come out a few years later at the NAMM Show they ceremoniously presented me with the first of the promotional T-shirts.

ng structure. I think he felt that if he had paid for it he should try and find some use for it, but the fact that this track never saw the light of day speaks for itself. On top of everything else, it was a pig to sync to tape (remember this is before the days of SMPTE/MIDI converters) and so mainly got used for demo presentations of the machine at international trade shows (it was the only way to use the EII multitimbrally, as the implementation of Multi Mode for this was as yet undreamt of). All in all, the sequencer was just not that useful.
The arpeggiator, on the other hand, (another novelty on a sampler) was an ideal performance device for the dextrously challenged, and the up, down or up/down rippling was particularly useful with harp or guitar samples.

Mac Sound Editing

A fledgling company called Digidesign (I wonder whatever happened to them?) produced a program called Sound Designer which allowed you to transfer EII sounds via SCSI (so the transfers were fairly fast, unlike the later MIDI Sample Dump) and edit them on a Mac computer. This allowed visual editing of the waveform for the first time, and was particularly good for working on loops (especially when you used the Crossfade Loop function). It also allowed you to visualise and edit all the analogue parameters as well, with a neat Sync function which updated the EII in real time. I found this particularly useful with certain clients who always wanted to be sat at the keyboard playing while I was fine-tuning a sound. It was difficult to keep leaning across them to do stuff, so I would do it on the computer instead. Eventually Digidesign produced versions of this program for the Prophet 2000, S900 and Emax and it became a way of transferring samples from one system to the other. And of course, you all know what became of their follow-up, Sound Designer II.

Bonfire Night With A Beatle Another EII session which sticks in my mind came not long after the Keith Emerson debacle and began shrouded in mystery. I got a call from the guy who ran Advanced Sounds, a musical instrument rental place in Bromley, saying that he had heard that I was the guy on the EII. I demurred with my usual modesty, but he was adamant that I came highly recommended. He said he had a great gig for me, the ultimate gig, but he couldn't tell me who with! Did I have a good library of EII sounds? Could I work at short notice? There then followed a succession of phone calls in which we agreed my daily rate which was to remain confidential between us, that I would be ready on a moment's notice to head out to the mystery artist's place which was somewhere south of London, and that this really would be the gig of a lifetime. I began to have misgivings when told that the person in question really was at the top of his field and that he lived down the A22. Keith Emerson lived down the A22 and he was considered by many at the time to be the greatest living keyboard player. Perhaps he felt that the fault had not been with either the sound of the bike or the chickens, but with the programmer, and had decided to try and get a better EII programmer than me to do a better job. I was terrified that when the call came, I would be given directions to Emerson's manor house. But when it did finally come, several weeks later at 10am on November 5th (I had just been out buying fireworks), the directions I was given had me going on past Emerson's village to Hastings. I breathed a sigh of relief and headed out immediately as instructed; they (whoever they were) were all waiting for me. I tore down the A22 as fast as my little jalopy would take me -- but when I got there, the place was deserted. So much for them all being waiting for me! I sat in my little car being buffeted by a gale off the Channel for at least half an hour, convinced that I was going to get blown away. Eventually an engineer I knew from AIR Studios turned up. He finally let me in on the big secret; it was Paul McCartney. It turned out Macca had bought an EII (already having hired one a couple of times from Advanced Sounds for strings and the other four disks that came with it to use on the Press To Play album) but was not really happy with the amount of use he was getting from it, now it was there in the studio. So he had asked for someone to come down and show them how to get more out of it. I was shown into the room with the EII and started to check through the disks I had brought with me, trying to decide which of my sounds to hit the great man with when he arrived. Having settled on some big impressive pads I was ready to impress the hell out of him when he turned up half an hour later. The trouble was that his first remarks after being introduced made me realise that he wasn't really interested in big pads or traditional sounds at all. He had heard the Art Of Noise stuff and had wanted to get into sampling because someone had told him that that was how AON achieved their sound, but he hadn't heard anything like that from the EII. I quickly realised that to impress him I needed completely different sounds and so dug out one of my crunchy drum disks. Now, on the EII the fastest way to load a sound was to put a floppy in the drive and turn the machine off and on again. The EII would load the operating system from the drive (which took a second or two at most) and then load the sound immediately afterwards. So in the interests of getting the sound loaded as fast as possible, this is exactly what I did. Big mistake! As I turned the machine back on, all the lights came on and there was a loud bang, then nothing. The machine was dead; the power supply had blown. Emu used to use the best in-house-designed power supplies for their US machines, but just bought in the cheapest 220-240V power supplies they could find for the export models. I had seen them blow a couple of times before -- but imagine me trying to explain this to McCartney. He's got this hotshot programmer in to show them all the mysteries of the EII, and what's the first thing he does? Blows up his £10,000 instrument! Great. I never had to talk so fast. I explained that if we rang Syco the distributor, they would be able to get us a replacement power supply. So I got on the phone to them asking for Ken McAlpine, head of service, but unfortunately he wasn't around that day. The only person I could get was the admin girl from the service department who told us just to send it in so they could change the power supply, which should take less than a week. Macca wasn't very impressed with this and said something to the effect that if they didn't get a bike down to the South Coast with a replacement power supply double quick then the EII would be coming back to Syco for a refund "with a lemon stuck in the disk drive." Finally, someone in authority there sanctioned the bike and we were told it would be there within an hour and a half. We settled down to discussing childhoods in Liverpool (I grew up there too) until it arrived. On its arrival, one of the engineers quickly fitted the replacement supply and we turned the machine on again. I breathed a sigh of relief as it came on normally and I watched the famous 'This May Take A While' message as my drum sounds loaded. Then suddenly, just as this was nearing completion, the EII's power went off. 'Oh no,' I thought, 'the power supply's gone again.' But in fact the gale that I had been sitting in a few hours earlier had finally blown down the power lines to the studio, and now there was a general power cut. We rang the local electricity board and were told it would be at least a couple of hours before they could get it back on. Four hours there and I still hadn't managed to play a single sound to McCartney, let alone actually do anything like sample! However, he seemed more perturbed by the fact that he had just been about to ask someone to make a cuppa, which now also wouldn't be possible. The next thing I knew we were all outside in the screaming gale (Macca, myself, the engineer and the tape op), stripping branches from trees to make a bonfire so we could boil some water for tea! We even ended up setting off some of the fireworks I had bought that morning because it was Bonfire Night. It was five hours before the power came back on and by that time, McCartney was due to go home. They put me up in a motel along the road and I went back the next day and we finally got something done.

CD-ROM

As well as fitting later machines with an internal 60Mb hard drive in place of one of the floppy drives -- a popular move, especially for stage use -- Emu pioneered an even more innovative data-storage system for EII, which again we take for granted with almost every sampler on the market today. In conjunction with a company called Optical Media, a SCSI CD-ROM drive and library was produced for the EII. The drive came in two versions: a purple and black 2U rack unit (the one I had in return for letting them use some of my sounds on the CDs) and a somewhat uglier one which would sit on the flat top of the floppy/hard drive bays (this is the one shown in the picture).

It was revolutionary at the time (though pricey at around £1500 for the drive and £200 for each of up to 40 CDs), because no-one had had any way to move sound from one sampler to another except via time-consuming floppy transfers. The CD-ROM allowed you to carry hundreds of sounds on a single CD instead of a couple of hundred 5.25-inch disks. It also gave you faster load times, which was invaluable live or with an impatient producer!

All in all, the EII really advanced the state of the sampling art in its four-year lifespan. Before it, multisampling, real-time control of analogue filtering via velocity, and crossfading via velocity and position were all unheard of. By the end of its life, it had seen the introduction of onboard hard drives, CD-ROM libraries and on-screen computer editing via SCSI. Sampling would never be the same again

Published in SOS September 2000