To celebrate the magazine’s 30th birthday, the SOS team take a trip back in time, finding out what it was really like to work in a mid-’80s home studio.
Nostalgia is a kind of selective memory. From our 21st-century perspective, the past stretches out like a golden landscape of great music, classic gear and bizarre haircuts, and we tend to overlook the fact that there were plenty of dreadful records in the ’70s and ’80s — and that many of the great ones were made despite, not because of, the equipment then available. It’s also easy to forget that the gap between home and professional studio environments was once a gaping chasm. Much of the equipment that is revered as ‘vintage’ today would have been available only in professional studios in 1985, and far too expensive for most home-studio operators.
With that in mind, we thought that a fitting way to mark SOS’s 30th anniversary would be to revisit home recording as it really was in the mid-’80s. We set out to assemble a setup that would include only equipment that was realistically within the reach of the average SOS reader, and use it to program, record and mix a complete pop song — in an ’80s musical style, naturally!
Sound On Sound was launched in response to a revolution in project-studio music recording. The advent of the MIDI protocol, MIDI sequencers, timecode-based synchronisers and narrow-gauge multitrack, all in the first half of the 1980s, meant it was becoming practical for hobbyist musicians to create complete tracks. Suddenly, eight- and 16-track recording was no longer solely the province of professional studios; but perhaps just as significantly, sequenced MIDI synth sounds and drums could both augment track count and maintain audio quality by being run into the final mix as synchronised live sources, rather than being recorded to multitrack.
Instruments such as the Yamaha DX7, Roland Juno 106 and Korg Polysix were bringing polyphonic synthesis to a mass market; Akai and Emu were doing the same with sampling, and with a little imagination it was possible to extract plausible rhythm tracks from a new generation of drum machines. There was a good selection of relatively affordable mixers to tie everything together, and even luxuries such as digital reverb were starting to become a realistic prospect for home-studio operators. For the first time ever, talented amateurs could create sophisticated recordings without the help of other musicians or engineers.
Yet, at the same time, many other technologies we now take for granted still lay in the future. Affordable sample-based synths and workstation keyboards would not appear until the later part of the decade. Steinberg had yet to create the ubiquitous Arrange Page, the graphical template to which today’s sequencing software adheres. Digital recording was in its infancy, so not only were luxuries such as non-linear editing and plug-in processing unknown, but unless you were lucky enough to own a Sony PCM F1 converter, master mixes had to be recorded to quarter-inch analogue or cassette tape. And entire market sectors that thrive today, such as active monitors, acoustic treatment and affordable capacitor microphones, had yet to be created at the launch of SOS.
Of those SOS staff based in our offices near Cambridge (a list that does not include founders Ian and Paul Gilby, Editor in Chief Paul White or Technical Editor Hugh Robjohns), only Editorial Director Dave Lockwood worked in ’80s home studios first time around. His recollection of the typical home-studio setups of the time — described in the ‘What Would You Find In An ’80s Home Studio?’ box — became our template in setting up the studio. His attic also proved an invaluable source of period-correct equipment, including the centrepiece of the entire setup: an Atari 1040ST home computer, complete with MIDI synchroniser and a copy of C-Lab’s Notator sequencer.
Other team members were able to beg, steal or borrow a few other typical home-studio items: a 16-channel Soundcraft 200SR mixer, a pair of Yamaha REV7 digital multi-effects units and a Fostex M80 quarter-inch eight-track tape machine. However, although we could muster various ‘vintage’ synths between us, what none of us had was a suitable selection of early MIDI instruments. With that in mind, we turned to leading UK hire firm FX Rentals, who very kindly lent us several key pieces from their vast treasure trove of music gear: an original Yamaha DX7, a Roland Juno 106 polysynth and TR707 drum machine, and an Akai S950 rackmounting sampler.
Assembling this lot brought home how unwieldy the home studios of yore were. Our setup might have been a typical ‘bedroom studio’, but anyone who managed to fit that lot into his or her bedroom would struggle to do any actual going to bed too! Even in SOS’s large meeting room, it was a challenge to arrange it all in such a way that all of our cables reached, and everything was accessible, while any changes to the setup would plunge us into feverish sessions of patching and re-patching. Since the Soundcraft mixer was a live-sound variant with no tape returns and only four groups, this had to be done on a regular basis. In some cases it was also a headache to find appropriate cables: the main outputs of the mixer, for example, appeared only on male XLRs, which needed somehow to be connected to RCA phono inputs on our master recorders and be paralleled to our monitors.
Our first challenge was to find a suitable song to work on. We toyed with the idea of covering an actual ’80s hit, but this would have posed copyright problems, and we didn’t want to be tied to slavishly recreating existing sounds. Nor did we want to begin with a song that had already been demoed or arranged using modern equipment and techniques. We also felt it would be nice to have a lyrical theme that tipped its hat to the decade in question. Reviews Editor Chris Korff supplied that with his genius title ‘Love Is A Two Player Game’, and from there it was a relatively straightforward process to weld some vaguely suggestive ’80s video-game references to cheesy extended chords of the kind favoured in ’80s white soul. I made a deliberately rough demo with vocal and acoustic guitar, leaving lots of space that could be filled in later with characteristic ’80s arrangement ideas.
As we started the project before our hire synths arrived, Dave Lockwood and I decided to try to get a head start by hooking the Atari up to a Yamaha MU50 General MIDI sound module, and roughing out the MIDI arrangement.
Dave picks up the story: “I have no idea why I’ve kept an Atari computer in my attic for the last 25 years, but having dug it out for this project, I can’t say I really expected it to power up, never mind to find all the necessary peripherals — you know how there’s always a vital part missing when you revisit old technology. But, there I was with a 1040, its dedicated, small monochrome monitor, a mouse and copies of C-Lab’s Creator and Notator on floppy disk, complete with the all-important hardware dongle. The only dongle I could find was for Notator, so we would have to use that, and my Version 1.0 disks seemed to have all been updated to Version 3.0, but that really didn’t matter, as we wouldn’t be using anything that wasn’t in Version 1.0.
“The software fired up first time. Amazing. It wouldn’t always do that when it was new. Then I installed my E-Magic Unitor SMPTE sync box, causing Notator to hang at the end of loading. Cleaning the tarnished connectors worked its usual magic and we were up and running. Now all I had to do was remember how to work it all.
“To be strictly ‘period correct’ it should probably have been a less sophisticated SMPTE sync box — something like an XRI Systems unit, maybe — and Steinberg’s pioneering Pro 24 doing the sequencing, but we had the Unitor and knew that it worked. Creator/Notator was also what I spent many years using in the mid-to-late ’80s, so even though I hadn’t looked at it for the best part of 25 years, I reckoned it would all come back pretty quickly.
“Notator (Creator is the same sequencer without musical notation) is a pattern-based sequencer with a very powerful Arrange mode for song construction. Becoming reacquainted with it for this project reminded me just what a powerful tool that mode is when building the early shell of a MIDI arrangement. No modern DAW is quite so fast at changing the length of whole sections or adding repeats with small variations. Huge structural changes can be made with a single numeric variable.
“Tempo was established using the system click, and then a simple shell drum part, step-programmed with kick, snare and hi-hat. In recognition of the number of sequencer-based ‘80s hits that were between 120 and 125 bpm, 4/4 time, with drums and keyboards hard-quantised to 16ths, we created a song based on exactly that, settling on a tempo of 123bpm.
“Looking at the 16 empty tracks of Pattern 1, I could remember that step writing was done in the track Edit window and, eventually, that notes were inserted by dragging the Note icon (pleasingly, actually the word ‘Note’) to the work area where they appear with the usual parameter set: note pitch, MIDI channel, velocity and position.”
“With a pattern-based sequencer, positional data always relates to position within the pattern, not the arrangement, so the start of your pattern is always 126.96.36.199 and halfway through an eight-bar pattern is therefore always 188.8.131.52. This makes step writing easy, as you always know where you are in the pattern, even if you have no idea what position that would be in the overall arrangement. Segment Copy took our two bars of basic drums up to eight bars, which was the longest individual pattern length required by the song.
“The vintage keyboards and drum machine that we were going to be using had not arrived in time for our programming session, so we were temporarily using a hardware General MIDI module for sounds. Nonetheless, we were careful to choose sounds that were as close as possible to the ones we would actually be using. Our GM module had a DX-style Rhodes and an analogue synth-like pad that would cover the work of the DX7 and Juno 106 that we were awaiting. The note allocations for the drums, which were to come from a Roland TR707, were reliably standard, although I miscalculated both the number of toms and crash cymbals, and forgot that the actual 707 ride cymbal would sound like a trash can being hit with... probably another trash can. We would rethink that one in the later programming.
“At this stage, we still planned to use real bass guitar, so a very basic synth bass playing eighth notes gave us our basic feel for the programming work.
“With Arrange mode turned on, patterns are placed in a list that is played out sequentially, so our basic working method was to create a pattern for each of our major sections: intro, verse, chorus, and so on. The simplified harmonic shape of the piece was outlined with the pad sound, and the Rhodes-like voice used to replicate the distinctive DADGAD-tuned acoustic guitar part Sam had used to write the song. Some of those voicings were then simplified in deference to parts that we thought would be added later.
“Notator’s Arrange mode has a set of track mutes that are associated only with Arrange entries, not the patterns themselves. This means you can use the same pattern a number of times whilst muting and unmuting different tracks each time it is played — perfect for a simple song structure where you might want to vary just the drum fill into the next part. Nine patterns covered all the sections, given the variations that could be generated by the Arrange mutes, and in an afternoon we were able to create a complete framework for the song that could be used to program more detailed MIDI parts and ultimately sync to tape for audio overdubs.”
Despite occasional frustrations, we got things done surprisingly quickly, and one day’s work was enough for us to generate a near-complete MIDI arrangement of the song. This, I think, was mainly down to the simplicity of the Notator system compared with a modern DAW. The lack of a graphical arrange page, tempo track or piano-roll editor closed off some of the options we might otherwise have explored in a modern DAW; there’s a limit to what you can do, but what’s possible can be achieved with a speed and immediacy that is lacking in today’s vastly more complex music software. Likewise, the knowledge that most aspects of the arrangement have to be fixed before you can record any audio forces you to make decisions rather than put them off until mixdown.
Even on the much-maligned General MIDI sound palette, our draft arrangement sounded pretty good, with appropriately brash drums, lush keyboard pads and punchy synth bass. We couldn’t wait to hear the effect of translating it onto period-correct, sought-after vintage instruments. With a sense of real anticipation, we unpacked the DX7, Juno and TR707 from their flight cases, dug out yet more MIDI cables, selected the classic Rhodes piano and string pad patches and pressed Play.
It was awful.
The richness and density of our arrangement was completely lost. Classic synths though they are, the DX and Juno sounded thin and weedy in isolation. Compared with the MU50, the DX seemed to have an exaggerated response to note velocity, which made even a simple keyboard part sound horribly uneven, while the Juno simply wouldn’t play even a basic melody part without randomly dropping notes. And although the TR707’s chunky kick and snare sounds held up well, its MIDI spec was about as basic as it could be, more or less responding only to Note On messages and nothing else.
Although we did have access to an original Korg MS20, we didn’t have a MIDI-CV converter handy, so we cheated slightly and pressed the reissue MS20M into service on bass. As if to show solidarity with the instruments of 30 years ago, it too turned out to have an incomprehensible MIDI spec, and it took us at least half an hour to figure out how to change the channel on which it received notes.
In a modern context, it would have been the work of moments to shove a compressor plug-in over the DX piano, but that wasn’t an option for home studios of the ’80s. Happily, Notator actually features quite a comprehensive set of real-time MIDI processors and effects, so although some manual editing of note velocity was necessary, adding some MIDI compression took care of the worst of it. Meanwhile, it turned out that the one thing that the Juno would reliably respond to was program change messages, so we were able to program a couple of slightly different patches for different sections of the song. Once we’d fired up the REV7s and added swathes of cheap reverb to everything, it was beginning to come back together again, though we eventually ended up keeping a few of the more incidental GM sounds, simply because they sounded so gloriously ’80s.
With our MIDI arrangement almost complete, it was time to think about recording some audio. We had a tape machine, we had tape, we had looms and isopropyl alcohol for cleaning the heads. What could possibly go wrong? Pretty much everything, as it turned out, as Dave recalls: “We started with a Foxtex M80 quarter-inch eight-track, found that it wasn’t working properly, went to a B16 that we also soon found didn’t work properly, nearly bought a Tascam 38 on eBay, tried to borrow an R8, and then fixed the M80 that we started with. Of course, it didn’t stay fixed: tracks one and eight ceased recording as soon as we got everything plumbed in, but we decided this was not too atypical of the ‘80s home-studio experience, and agreed to press on regardless. With track eight inoperative, track seven became our new ‘edge track’ to be used for SMPTE timecode, which left five working tracks for audio.”
If there was one positive that surprised us about the tape experience, though, it was the resilience of the synchroniser. Despite the indignities to which our timecode track was subjected by the tape machine, we never once lost sync, and the Atari would resume MIDI playback almost as soon as the recorder got up to speed.
We had always planned to run the MIDI instruments live at mixdown, but even then, given that we had hoped to record doubled lead and backing vocals, at least two electric guitar tracks, acoustic guitar and fretless bass, it wasn’t obvious how we were going to fit everything in. Putting off that dilemma for the time being, we started by recording electric guitar.
When we think of the sounds that define the ’80s, it’s perhaps not guitar tones that spring immediately to mind. If anything, it’s probably remembered as a decade when guitar music was pushed out of the mainstream by synth-pop. Nevertheless, there were some guitar sounds that are very characteristic of the era. Having worked as a session guitarist as well as a producer/engineer, Dave Lockwood was in the perfect position to recapture these.
“If pushed to think of a typically ’80s guitar sound, many of us would probably identify ‘clean, compressed, ‘clicky’ DI’d Strat’ as a candidate. The ‘clean’ setting of a Tom Scholtz Rockman box is probably responsible for popularising this sound, but a nicer version of it, with less noise, can be achieved with a suitably squashy compressor pedal straight into a DI box. For this we used a ’70s Strat, on the ‘inbetween’ setting between bridge and middle pickups, through a clone of the MXR Dynacomp compressor, which kills the transients to just the right extent, feeding a 1980 BSS active DI box with a very useful 8kHz low-pass filter setting.
“Even in the smaller home and project studios of the ‘80s, most players generally preferred to mic an amp for any distorted guitar parts, whenever possible. Acceptable DI solutions simply weren’t available, perhaps until the advent of the Tom Scholtz Rockman — although some might argue the assertion that the Rockman’s distorted sound approached acceptability!
“A slightly later, and better, DI solution was the original SansAmp pedal. I discovered you can make one of those sound a bit like a Rockman if you dial it in really badly, but you wouldn’t have thanked me for treading all over the track with that, and in the end I settled for something fairly generically amp-like from the SansAmp using its Mesa Boogie settings. The same sound was used for the crunchy chording and the second half of the solo.”
Dave’s electric guitar parts occupied two of our five available tape tracks, albeit with occasional gaps. Next up, we decided, would be acoustic guitar: our first foray into recording an acoustic source. These days, we are spoilt for choice regarding studio microphones, but prior to the influx of affordable Chinese-built capacitor mics in the ’90s, it was a very different state of affairs, and it would be a rare home studio that owned a range of mics for different purposes. Much more likely was that a single half-decent mic would be pressed into service on any source, and we decided that the only mic available to our time-capsule home studio would be an AKG C451 with the CK1 cardioid capsule. This combination is still widely used today on acoustic guitar, but probably isn’t something most of us would reach for on male or female vocals!
Today, many home-studio owners probably have at least one fancy outboard mic preamp, or perhaps a ‘voice channel’ with compression and EQ built in, but such indulgences were unknown in 1985. You recorded through your mixer, and if, like ours, that mixer was a Soundcraft 200SR, your processing options were limited to fixed, four-band EQ. We did allow ourselves a pop shield, since small-diaphragm capacitor mics are very susceptible to plosives, but although our improvised studio space was quite boxy, we decided that acoustic screens such as the SE Reflexion Filter were a definite no-no for the time.
The dynamic range available on even the most basic of today’s computer-based recording systems typically exceeds 100dB, meaning that you can pretty much leave as much headroom as you like when recording quiet or dynamic instruments; the limitation on the dynamic range in the recording will be due to ambient noise in the recording space rather than system noise. Things are a bit different when recording to narrow-band tape on a poorly maintained, 30-year-old eight-track machine, and it took some juggling of gain pots and a firm strumming hand to get a decent level to tape. The C451’s fierce high-frequency lift, which can be a hindrance with digital recording, helped to combat the inevitable loss of top end that comes with the recording format.
In 1985, as now, the most important part of any pop track was the vocals, but the number of products targeted specially at recording and mixing the human voice has mushroomed since the ’80s. As well as the aforementioned ‘voice channels’, most of us probably have access to pitch- and time-correction, de-essing and other vocal-oriented effects, not to mention unlimited recording tracks with advanced editing and comping tools. So it was an interesting experience recording vocals to a single track, with options limited only to ‘keep it’ and ‘do it again’. In practice, though, we quickly got used to the approach of working through the song, recording a verse at a time and performing drop-ins to correct any dodgy lines. Once again, it meant committing to a take straight away rather than putting off the decision until mixdown, but although my final vocal performance had plenty of ropey bits in it, that was more to do with lack of time and an over-ambitious harmony part than with the limitations of the recording process.
The ’80s was arguably the golden age of the pop duet — who can forget such gems as ‘Don’t Give Up’, ‘Up Where We Belong’ or, er, ‘Especially For You’? — and our song lent itself fairly naturally to this treatment. The two vocalists were, however, recorded separately: not an uncommon procedure when a home studio may perhaps have had only one good microphone.
Nell Glasper, Sound On Sound’s Production Editor, is a fine singer of traditional and contemporary folk music, but soon locked on to the necessary ‘detached’ vocal style required of the female lead here. With Nell’s main vocal down, we were faced with a dilemma. We wanted the female vocal to be doubled, but we had run out of working tracks! So we took the bold decision to erase the acoustic guitar, which seemed a bit less crucial to the arrangement once the vocals were in place. It was a decision we never regretted: Nell’s accurate double perfectly captured the full ‘80s ‘disinterested vocalist’ effect. The only place where the acoustic guitar was really missed was in the middle eight, where one of Dave’s electric parts did not play, so with some nifty drop-in work we were able to reincorporate a new take just for this section.
That chorus-y double-tracked sound is very characteristic of ’80s pop production, so we decided to try the same trick on the bass. Elsewhere in the song, the synth bass was a pretty dull pedal-note affair, but we’d come up with quite a nice melodic line for the middle eight, and there was a gap in the male vocal part at this point, so we supplied Chris Korff with a fretless MusicMan and set him to work.
With a bit of imagination, it’s surprising how much you can fit on to a handful of tape tracks, and had we been working on the song for longer, I’m sure we’d have come up with more ideas to squeeze into unused corners of the arrangement. With a deadline looming, though, we decided it was time to integrate the one piece of ’80s music technology we hadn’t yet exploited: FX’s Akai S950 sampler. Although we had a decent supply of 3.5-inch floppy disks, what we didn’t have was the factory library or indeed any sounds for the unit, so Reviews Editor David Glasper was assigned the unenviable task of programming something from scratch. There wasn’t a pressing need for additional musical elements in the song, so he chose to reinforce the lyrical theme by sampling and manipulating some classic ’80s video-game sounds (see box), which we integrated mainly into the fade at the end.
Finally, it was News Editor Will Betts who supplied the finishing touches to our track, with a couple of magnificent orchestral stabs and a virtuoso General MIDI helicopter solo — anachronistic perhaps, but very, very ’80s in sound. It was time to mix.
The very idea of a separate mix stage has become unfamiliar in the DAW world, and it’s perhaps in mixing above all that the difference between old and new emerged most clearly. In a computer-based system we have complete freedom to EQ, compress, effect and automate every last shake of a tambourine or pluck of a string, and often enough, we start to do this as soon as the tracks are laid down, safe in the knowledge that everything can be ripped out and changed at a moment’s notice.
In our ’80s home studio, by contrast, we had a pair of Yamaha REV7 reverbs (although we used only one of them), a guitar pedal set up as a global mono delay, and a four-channel Aphex compressor. The Soundcraft desk has fixed four-band EQ on every channel, but no master EQ at all; it does at least have master bus inserts, though, so we were able to set two bands of the Aphex up as a bus compressor, while a third did its best to iron out the inconsistencies in my vocal performance.
In the DAW world, everything we do at the mix can be automated, edited and stored for later recall. Mixing on an analogue desk is much more like a musical performance, and the same recording of the same song will come out differently every time. Luckily, we had plenty of hands available to man the faders, and our MIDI arrangement was sufficiently developed that few fader moves were required in that department. However, two of the TR707 sounds we were using — the handclaps and the tambourine — shared an output, and one seemed much louder than the other, so Nell was deployed to adjust the tiny onboard slider that controlled their level. As the sources most in need of fader riding, the vocals each got their own operator, while Chris Korff was deputed to add spacey delays to the video-game samples.
After a number of rehearsal passes, we had come up with a fairly detailed map of the song with levels for each section. Even then, we had to do four or five ‘takes’ of the final mix before we got one that we were satisfied with; the difference between something’s sitting at the right level and its being noticeably too loud or quiet could be as little as a couple of millimetres of fader travel. Knowing that the song would eventually need to end up in the digital domain if anyone was to hear it, we cheated slightly and recorded the master to a Mac as a 24-bit WAV, as well as to DAT and cassette, for the sake of authenticity. The latter sounded a lot better than a lo-res MP3.
Between us, the SOS staff who participated in this project can look back on more than a century of home-studio recording experience. Yet apart from Dave Lockwood, none of us have first-hand experience of the studio world into which SOS was born three decades ago. If there’s one thing that our attempt to recreate this world brought home, it’s that music technology has always been the tail that wags the dog. Working with an ’80s studio makes crystal clear why so much ’80s music sounds as it does! Machine-like, hard-quantised backing tracks; arrangements given life by MIDI mutes rather than human performance variation; unconvincing synthetic recreations of real instruments; digital reverb as a sticking plaster to cover up all sorts of evils: so many of the characteristic qualities of ’80s records came about not because they were intrinsically desirable, but because they represented the path of least resistance given the technology of the time. It would have been far harder to use our ’80s home studio to make music that didn’t recall that decade.
Today’s music technology is infinitely more powerful, and the ways in which it shapes our music are perhaps more subtle — but are we any closer to being its masters than was the case 30 years ago? Or does today’s technology merely push us in different directions?
Thanks to FX Rentals for supplying the instruments used in this feature.
I started my recording career with a Studer 24-track in a pro studio, but I did my fair share of home and project studio recording, too. In my experience, a typical mid-to-late ’80s home studio would be based around a multitrack tape machine — quarter-inch eight-track, half-inch eight-track or half-inch 16-track, according to budget — and a computer running MIDI sequencing software. The computer would almost certainly have been an Atari 1040, with the most popular sequencers being Steinberg’s Pro 24 and later Cubase, and Creator/Notator from C-Lab (later called E-Magic). A SMPTE timecode synchroniser would have been used to lock the sequencer to the tape machine, so all the MIDI sources could remain ‘live’ through to the final mix.
In the earlier years, synths would often only run a single polyphonic voice, so you would need a few of them for a full production, leading to the popularity of keyboardless voice modules like Yamaha’s TX7, which offered almost all the functionality of a DX7 in a compact format. Samplers, like Akai’s popular S900, were inherently multitimbral instruments with a variety of applications and would often be used in place of a dedicated drum machine, with every voice split out to its own output.
Combining these sources at mixdown called for an inconveniently large number of inputs, often with every aux and tape monitor return pressed into service alongside the input channels. Affordable digital reverb from companies like Alesis and Yamaha added still further to the pressure on line inputs — somehow there were never enough.
In a typical setup of this period, there would be a modest amount of outboard processing: perhaps one decent stereo compressor for mixdown, which would also have been used for tracking vocals. Mic choice too would often have been quite limited: Shure SM58s and 57s were often pressed into service, or AKG’s C1000 and C451 capacitor models. AKG’s D190 and the rather nicer D202 and D224 were certainly around in home studios, and maybe something affordable from Sennheiser or Beyer, but if you had a large-diaphragm, multi-pattern model back then, you’d be considered practically a pro.
Mixdown would generally have been to quarter-inch analogue tape or possibly even cassette, for those who had already spent too much on the rest of their system. The DAT format of compact digital audio cassettes was still a few years off, but for the really flush there was always the option of a Sony PCM F1 and Betamax recorder. The F1 was a digital converter that stored its bitstream as a video signal onto Betamax video cassettes. It was a bit finnicky to use, but sounded OK, and I mastered a fair few albums with it until DAT came along. Dave Lockwood
When we were drawing up our list of must-have ’80s equipment, an Akai sampler was near the top. This really was cutting-edge technology at the time: when Akai released the S612 in 1985 it was arguably the first sampler that could be considered anywhere near affordable. It came with 128kB of internal memory, with which you could sample up to eight seconds of audio at the lowest 4kHz resolution (a single second was available at the maximum 32kHz), and an optional 2.8-inch Quick Disk drive was necessary if you wished to preserve your samples for posterity. Still, you could hope to own one without selling a grandmother (an Emulator 2 cost approximately nine grandmothers back in 1985) and sampling suddenly became feasible for the project studio owner.
S612s are pretty thin on the ground these days, and 2.8-inch QD drives even more so, and when we discovered that the extremely helpful people at FX Rentals had an S950 we could borrow, we realised we had to compromise slightly on period authenticity. Nonetheless, the S950 is not too far removed from its forebear, being an updated version of 1986’s S900. It’s still a 12-bit machine, but offers sampling rates of up to 48kHz and an internal memory that’s expandable to a whopping 2.5MB. It also has a standard 3.5-inch floppy drive, and as we never throw anything away at SOS, finding a few disks for it wasn’t going to be a problem.
So, faced with a song with lyrics about video games and an insatiable appetite for as many corny ‘80s clichés as we could throw at it, it was obviously necessary to sample some eight-bit computer-game sounds and use them to decorate the song’s intro and outro. And as the onetime owner of an S950, it fell to me to play sampler tech...
I began by firing up the machine and staring at it uncomprehendingly for a while. It’s been at least 13 years since I last used an S950, and I quickly realised I’d forgotten everything I ever knew about programming one. Actually, I didn’t stare uncomprehendingly for long, because the 950 is a very simple machine to use, and has a number of helpful instructions printed on its front panel. With these instructions as a crib and a little exploration of the menus, one can begin recording and editing samples almost immediately. In fact the only time I needed to refer to the manual I’d so diligently downloaded beforehand was to find out how to add another keygroup. (The manual, by the way, is charmingly written and is a refreshing change from the usual dodgy translation from the Japanese typical of the period. Bravo to whoever was responsible.)
I didn’t attempt any Fatboy Slim-style Akai stunt programming, and simply sampled a dozen or so sounds from a popular eight-bit computer game (also, coincidentally, released in 1985), which I reckoned gave a fair selection of the kind of crunches, bloops and beeooow sounds we were looking for. I then set about programming them across the keyboard.
This was where the complete lack of a graphical interface really made itself felt. Instead of drawing or dragging anything with a mouse, it was a case of manually selecting a sample, setting its pitch and then setting a maximum pitch range, all with a grid of buttons and a data wheel. Easy enough with a dozen samples, perhaps, but I hate to think how long it would have taken to program a properly multisampled instrument. (That said, I used to own a Roland S550 with a green and black CRT monitor that gave you a waveform display, amongst other things, and that used to give me actual nightmares, so perhaps a GUI isn’t everything.)
Once I’d relearned how it worked, I actually began to enjoy using the S950. It certainly has a sound of its own, and once you get used to the lack of any modern editing facilities, it becomes rather fun to use. It also produces far more unexpected results than its modern, computer-based equivalents, perhaps as a result of my less-than-expert programming, and I found that firing random notes at it from an Arturia Beatstep Pro created all sorts of interesting rhythmic effects, which I in turn sampled back into my computer (taking a fraction of the time to do so, I might add, and with no worries about running out of memory). It also has an endearing habit of defaulting to ‘Omni On’ (in fact the front panel’s ‘On’ and ‘Off’ buttons seem to exist solely to adjust this control), which caused all sorts of bonkers noises to issue forth every time we switched it on.
For conventional sampling work there’s very little to recommend a sampler of this sort, but for creative serendipity or sheer grit they’re hard to beat. We’ve certainly gained a lot in terms of usability and speed in the last 30 years, but perhaps we’ve also lost something of what made sampling so interesting in the first place. David Glasper
Point your browser at https://sosm.ag/nov15media to hear our ’80s mix in all its glory — including versions mastered to DAT, cassette and MP3!
To mark our 30th anniversary, the SOS team chose some highlights from the history of modern music technology. Click the link to download our Timeline Of Tech! timeline-tech-1115.pdf