Emu Systems Emulator II, Part 1 (Retro)

8-bit Sampling Keboard

Reviews : Keyboard

Until the EII came along, Paul Wiffen had no time for sampling as a musical tool. But when he found all the filtering and enveloping he was used to from analogue synthesis, plus a sound fidelity which he hadn't heard before, he was a convert...
 
I am a great believer in heresy. I admire the Salman Rushdies of this world far more than the Ayatollahs, the Martin Luthers more than the Popes or the Archbishops of Canterbury. Those who hang around with me socially eventually get used to my throwaway jibes about the 'all-time greats'. I find nothing jolts people out of their complacent acceptance of supposed greatness as comments along the following lines: "Mozart got as much out of the symphony orchestra as the average Country and Western singer gets out of a 12-string guitar;" "Lennon & McCartney weren't the greatest thing to happen to popular music, although George Martin may have been;" or "a Fender Stratocaster is OK as long as you don't want to offend anybody." They challenge established perceptions, especially amongst those who have never really listened for themselves and at the very least, force people who toe the accepted line to justify their position. Previous Retrozones by this author have contained the odd minor heresy along the lines of "I never thought the Prophet 5 sounded very good" or "if they had spent as much time on the Moog Source's sound as its revolutionary control panel, it might not have sounded worse than any of its Japanese competitors." However, what I'm about to say will be the mother of all blasphemies for generations of electronic musicians. You may want to sit down and take a calming breath before embarking on the next paragraph.

The Fairlight, inspiration from afar to a generation of electronic musicians, was, once you got one in front of you, a load of old tosh, the electronic equivalent of a kazoo, with all the musicality of a vacuum cleaner. Whatever sound you put in, you got the same nasal honk out, and trying to use it like a normal keyboard was the most frustrating thing I ever did. I thought so the first time I heard/played one and nothing has happened since to change my mind (I am of course referring to the pre-Series III machines; the engineers at Fairlight would have been fools and charlatans indeed if they couldn't have made the Series III sound decent and play well after all the progress that had been made by other companies by the time that machine came out...). I am referring to the pile of junk marketed between 1979 and 1985. And when I say "the first time I heard one", I mean really heard one, on its own, not buried in a production with all the craft of experienced engineers and producers disguising its nasal honk. I still vividly remember Ken McAlpine, a former colleague of SOS Publisher Ian Gilby and myself at Electronics & Music Maker, sneaking me into the demo room at Syco in 1983 to hear a Fairlight naked for the first time. I was horrified. Everything sounded like it was coming through a fuzzbox, and a very cheap and nasty fuzzbox at that! I came away from that evening with a profound sense of loss that the supposed greatest electronic musical instrument was such a turkey (though it increased my admiration for all the musicians and engineers who managed to make it sound so good on their records). And sharing a booth with the original Emulator at a Paris show a few weeks later, where the demonstrator did nothing more than play the Marseillaise on an out-of-tune trumpet over a maximum of two octaves, did little to resurrect my faith in sampling.
The Second Coming
All the more appropriate, then, that the place where I lost my faith in the tin god Fairlight should be the very place where, almost exactly a year later, I was to find my faith in sampling reborn. For it was to the same demo room at Syco that I was invited (as an E&MM freelance contributor by then) to do the first UK review of the Emulator II. I must admit I was not that wild on the idea, and was sharpening my quill for a Wiffen hatchet-job, but within the 45 minutes they gave me alone with the machine (not a situation which SOS would accept for doing a review nowadays, but all that was on offer from the sole UK agent at the time), I was completely won over. It answered all of my by-then well-defined objections to sampling in a single instrument -- and that is the most important point to note: it was an instrument. All previous samplers had been at best production tools; the EII was the first I would dignify with the term 'instrument'.

To begin with, when you made a sample, it didn't sound like it had been put through the mangle. Although not in the fidelity league of even the cheapest of today's sampling devices, the sound which came back from an EII was at least recognisable and had fundamentally the same character as what you put in. For me, that was a revelation in itself. Whilst I hadn't been given enough access to any previous sampler to test its sampling fidelity myself, I was pretty sure that the sounds they made had not sounded like that going in (if they had, there were some pretty cloth-eared engineers and musicians out there!).

So where did this quantum leap in fidelity come from? The EII, like the Fairlight and the original Emulator, still stored its samples in 8 bits, and its fixed sample rate of around 27kHz wasn't that much faster than the variable rates on its forebears; so why did it sound more true? Because, instead of linear encoding, the engineers at Emu used a method known as delta/sigma or '8-bit companded'. Without going into a lot of complicated mathematics, the easiest way to understand what this means is to think of it thus: instead of analysing each sample's value over the whole amplitude range (which 8-bit storage is just not up to), the EII would use the 8 bits to make a more accurate measure of the difference in value between the current sample and the previous one (almost always a smaller value, therefore more accurately stored in just eight bits). As a result, for 99 percent of sounds, the EII gave significantly more accurate results than any previous sampler.

Of course, it wasn't perfect on all sounds. Fast attack transients over the entire dynamic range (such as in percussion) could defeat this cunning scheme and, as a result, drums could sound a bit crunchy (although some people loved this effect). Consequently, when Emu released their dedicated percussion sampler, the SP12, the following year, they used 12-bit linear sampling rather than 8-bit delta/sigma. But the overall sample quality of the EII was still streets ahead of all its rivals.

This was only the beginning of the innovations which the EII ushered into the world of sampling. The EII was the first proper multisampler, allowing samples to be allocated at will to user-definable regions of the keyboard (its predecessor had a maximum of two samples, each of which had to occupy half of its four-octave keyboard). So if a sample only worked over four or five notes (as many do), you could assign another sample of the same instrument, as needed, as you went up or down its five-octave keyboard. The number of samples you could have was only limited by the total memory available (17.6 seconds' worth), which meant that if you were careful and used looping properly, you could get a dozen decent samples across the keyboard. Of course, you wouldn't be sequencing tracks and tracks of different multi-instruments, but then this was the early days of sequencing anyway (MIDI was still in the oxygen tent, although the EII did feature both MIDI and a multi-track sequencer).

But for me the thing which made the EII a real instrument was the fact that it was velocity-sensitive, and not just on amplitude. Instruments in the real world don't just get louder as you play them harder, they get brighter as well, because when you put more energy into a system, more high harmonics are generated. If you merely make a sample quieter when you play it more softly, it tends to just sound further away. But if you use the velocity to close a low-pass filter cutoff, then the sound gets duller as well as quieter, which is much more realistic. And this was just what you could do with the EII. As a result of this combined with the multisampling, the EII was the first sampler I ever came across which could get even close to a piano. The Piano disk it came with would not have pleased the average classical piano player, but it was certainly useable in a rock band at full tilt, and it sounded more like an acoustic piano (and weighed a good deal less) than the Yamaha CP70 series, the staple live pianos of most professional keyboard players back then.

Westside Stories

Not having the money to buy an EII myself, I did the next best thing, and started looking around for someone else to buy one for me to play with. As a result of working for Sequential Circuits, I had recently met Geoff Downes (ex-Buggle and all-too-briefly keyboard player for Yes), the man renowned for having more keyboards live on stage than anyone before or since. As I had sold him an OSCar and an Elka Synthex, Geoff had already asked me to get involved with the programming of these instruments on the third Asia album, Aqua. I managed to bring the EII up in conversation with Geoff on my next visit to John Henry's where Asia were rehearsing songs for the new album. "How is it?" he asked. "They've really done quite a good job of turning a sampler into a musical instrument," I replied, not wanting to overplay my hand (Geoff already had a Fairlight, which he had used very successfully to do the cello part on Kate Bush's 'Army Dreamers'). Two weeks later, I found myself in charge of Geoff's brand new EII at Westside Studios where the album was to be recorded. For the next six months, I was actually paid two or three days a week by Geoff to get intimate with this wonderful instrument (oh yes, and develop his library beyond the five disks that came with it). What's more, as the Royal Philharmonic had been hired to play on 'Voice Of America' on that album, Geoff had each section of the orchestra playing individual notes to digital tape long before most classical musicians had any idea of what sampling was. I had mountains of material to sample into the EII. I was in paradise.

This went on throughout the entire recording of the album (whose budget went over £1 million) unless I was actually in Studio 1 tweaking OSCAr, Synthex or Prophet sounds for a part being put to tape or reviewing other keyboards for E&MM (which Geoff let me do there, as he got to check them out without any sales pressure). I sat there sampling the cream of the Royal Philharmonic into the EII and any other weird and wonderful noises that Geoff and I could come up with, inspired by the South Bank Show documentary of Peter Gabriel sampling stuff in a Bath scrapyard some years before. Actually, I didn't feel too guilty about the Royal Philharmonic, because the EII spared their blushes too. When Mike Stone, the producer, listened to the individual recorded tracks after they had gone, the back desk of the second violins could clearly be heard talking, during actual takes, about which horse to back in which race (clearly they were not used to being individually miked!). However, because we had individual notes recorded, I was able to sample the violins into the EII and Geoff used them to replace the parts contaminated by racing-form discussions.

The only time the EII ever let us down was on drums. Having heard how tight the drums were on Duran Duran's 'Wild Boys', which had come out earlier that year, the band, in consultation with Mike Stone, decided that Carl Palmer's somewhat free style of playing might be a bit out of place in 1984. So they decided to sequence the kick and snare drum and then just have Carl play the tom fills and so on over the top. Now Carl's kick and snare had been recorded in the as-yet-unfinished Studio 2 at Westside with bare stone walls, ceiling and floor, and had probably the best drum sound I have ever heard. But when sampled into the EII, this huge sound just crumpled up and died. As a result, the band did some research and, finding out that the Synclavier had been used for the drums on 'Wild Boys', scheduled a demo at Westside. I think the original idea was that the whole band were going to finance the purchase of this 'wonder system' as it was going to 'magically fix the drums', and everything was going according to this plan as the flight-cased racks containing the computer brain were wheeled in.

There were a few problems; the Synclavier's monophonic sample playback restriction meant that when the kick and snare were sequenced together, the kick cut off the ambience of the snare, so in the end, each part had to be put to tape separately. Nevertheless, thanks in part to the Synclavier's 16-bit 50kHz samplin

Only Available In London Few people ever realised that the EII almost didn't see the light of day. When the smiling Californian demonstrators from Emu Systems premiered the prototypes to such enthusiastic reception at the Frankfurt show in February 1984, the company was actually on the brink of financial collapse. They didn't even have the money to build the first production run of machines. As a result, a deal was struck with Syco Systems, the company Peter Gabriel had founded with his cousin Stephen Paine to distribute the Fairlight in the UK (and which also handled Emu), whereby they agreed to pay upfront for the first 25 machines. But as a consequence, some fairly rigid legal restrictions were placed on Emu to make sure that all of the first 25 machines came to the UK. I have heard stories of Santa Cruz lawyers checking serial numbers and units on the production line to make sure that not even one machine slipped out for delivery to US dealers or stars until Emu's obligation to the bacon-saving Syco was fulfilled. As a result, some fairly famous American names were left waiting high and dry for this machine they had heard so much about (the Summer NAMM show in Chicago had let the cat out of the bag in Emu's home country as well). When it was clearly explained to them that the first units had all been shipped to the UK, they did what any self-respecting star would do -- they got on planes and came to London. Which is how I came to see Stevie Wonder coming out of the Syco demo room with his entourage while I was waiting to go in, and (literally) ran into Steve Porcaro and David Paich of Toto coming in as I left. I was keeping fine company indeed, and even though I personally hadn't a snowball's chance in hell of coming up with the 10 grand plus VAT needed to buy my own EII, I still felt like I was walking with giants. Both, I believe, bought EIIs from Syco that day.

g rate (the only system which got anywhere near this back in 1984), the kick and snare drum samples sounded like God himself hitting drums. Everyone was pleased and it looked like Geoff was going to net the Synclavier at the band's expense -- until the other members of Asia noticed that one very important part of the Synclavier system looked an awful lot like a piano keyboard, and began to think that they were being hoodwinked into buying Geoff's keyboards for him. In the end, I think Geoff ended up footing the entire £30,000 bill for the Synclavier system himself. Ah, well, you can't win them all!

Floppy Fun

The first EIIs, weighing in at a shade under £10,000 excluding VAT, came with just five disks: Piano, Marcato Strings, Brass, Drums and Acoustic Guitar. This might seem a little light by today's standards of CD-ROMs packed with great sounds, but it did the job. The Marcato Strings in particular had such depth and power that within months they were appearing on records the world over. Of course, Emu realised that they would need a much bigger library than this in the long run, and over the next two or three years they produced a steady stream of library sounds on those 5.25-inch floppies that seem so seem so huge now. I think at the time EII users may have been paying as much as £50 per disk (plus VAT) for these, but they obviously felt that a good library was worth it. Emu's sample library has justifiably been one of the many attractions to their evolving range of samplers ever since, and became the lynchpin of the Proteus range as they gradually ported this library of sounds from the top of their range to the bottom. They never did this so fast that it devalued the top-of-the-range product, but it always made their cheaper products sound every bit as classy as their most expensive.

But the floppy disks on the EII did not merely contain sounds. As with so many samplers at the time, the operating system did not reside in EPROM (except for the very small 'bootstrap' routine that put a request for a disk in the display). Instead, the operating system was stored on each and every EII floppy disk, along with whatever sounds were also stored there. During programming, when moving between the numerous modules on the EII's front panel which covered everything from sampling and digital processing to analogue filtering and enveloping, the floppy drive would access for a couple of seconds to load the software for that particular module. So if you were doing any more than simply playing the keyboard, it would pay to keep a floppy in the drive the entire time.

The fact that the operating system was on every floppy had its advantages and disadvantages. On the plus side, as long as you had at least one EII floppy disk, you were never without the operating system (which could happen so often on other samplers with a specific operating system disk). On the down side, when you received an update for the EII, you had to sit and run a special routine on each of your library of floppies to get them up to the latest version. If you didn't do this, you could end up with the dreaded 'software mismatch error' in the display during sampling or programming. When Emu eventually introduced a hard disk version of the EII, these problems were solved because you could keep the latest operating system on the hard drive, but the original EIIs had only two floppy drives.

Next time we will look more closely at all the various modules on the EII, which gave it an unprecedented palette of sample-editing tools and real-time analogue processing, as well as some of the more bizarre situations into which being an EII programmer led me (without giving too much away, it involves building bonfires in a gale, a motorcycle, and some fairly bemused chickens). You will also discover how a small strategically placed strip of chinagraph tape can cause a colossal hiccup in the sales of a major sampler. Betcha can't wait!

Published in SOS August 2000