Not satisfied with remaking their own classics, Korg engineers have recreated ARP’s legendary synthesizer, the Odyssey.
When Korg announced that they were going to recreate the ARP Odyssey, there was far less scepticism than one might have expected. Perhaps this was because the company had already been successful in re–releasing their own MS20, so the relevant emotion was surprise at the choice of instrument rather than worries about whether the company was capable of pulling it off. My response was neither sceptical nor concerned. I seem to remember that there was a lot of bouncing around and whooping. Maybe even some hollering too. Then it all went quiet. Time passed. The original launch date came, went, and drifted into memory. Then some more time passed. Suddenly, it was January 2015, and the great and good of the music industry gathered in sunny Anaheim for the NAMM Show. I was there too. And so were all three cosmetic versions of the Korg ARP Odyssey (which, henceforth, I’ll call the KARP to differentiate it from the original Odyssey). Of course, it’s impossible to conduct a sensible test of any instrument in such a chaotic environment, let alone compare it with the memory of the sound of a synth that, at that moment, resided about 5415 miles away (as the bionic crow flies). So I did something different; I set up my favourite trumpet patch as if it were my own Odyssey in front of me, and only then did I put on the headphones and press a key. Sound On Sound can’t print the vulgarities that then issued forth, but I’m confident that passers–by realised that my invitations to sexual congress were statements regarding the accuracy of the sound I was hearing. I know the Odyssey trumpet... I mean, I really KNOW it, and what I was hearing was an Odyssey trumpet.
Five weeks later, back in the dismal English winter, I took delivery of a rather gorgeous black carry case about the size and weight of the hand luggage that you can carry on a reputable airline. The only clue to its contents was the ARP logo on one corner. Opening it revealed a gorgeous, MkIII–styled KARP with the superior case design of the MkI and MkII Odysseys married to the black and orange fascia of the MkIII. Were it not for one enormous, humongous, cannot–be–overstated difference (which we’ll address later), it truly recalled the look and feel of the original. However, it’s not a clone; opening it up revealed a single board for the synth itself, plus a smaller I/O board, both built to modern manufacturing standards, and replete with surface–mount technology. Purists might complain but, if it sounds right, I’m not going to care whether the resistors are the size of microdots.
The KARP’s oscillators appear to be the same as the original synth’s. Indeed, with the exception of a switch that allows you to determine whether portamento (when applied) occurs when you use the transpose lever to switch up or down two octaves, the only thing that appears to be different is... well, nothing. So I started my tests by comparing these with my original Odysseys. Things started well, and the sawtooth waves of the two synths were, in any sensible sense, indistinguishable from one another. The same was not immediately true of the pulse waves. With the PW fader at 50 percent, my Odyssey had the hollow character that one always obtains from a true square wave, whereas the KARP was brighter and richer in harmonics. I hooked both synths up to a signal analyser and, sure enough, the even harmonics were far more prominent on the KARP. So I experimented with the pulse-width control and, a few millimetres north of the 50 percent mark, the correct sound and spectrum emerged. This should be merely a calibration issue, so I hope that Korg will attend to it on retail units. A similar discrepancy existed at the other end of the fader’s travel, too. The KARP was able to output a narrower pulse than the Odyssey, but that’s a good thing because it means that it can offer a wider range of initial waveforms.
I tested the modulation characteristics of the oscillators, the four types of FM (pitch modulation) and PWM. Again, I found that the faders on the KARP offered slightly greater modulation amounts at their maximum settings, but, other than that, all was as before, so it proved to be simple to coax the same sounds from each synth. This was even true of ring-modulated and sync’ed sounds. The only way that you would have known which synth was which while I was setting them up was by the scratchiness of the vintage faders (or, to put in another way, the smooth response of the new ones). As far as the results were concerned, it was like listening to two units of the same vintage synthesizer.
Turning to the noise generator, I found that the KARP’s sounded rather ‘bluer’ than my Odyssey’s, even in its pink noise mode. To turn this observation on its head, the vintage synth’s noise had more of a low–end roar than the new one’s. Interestingly, this didn’t affect the classic ‘Karn Evil 9’-type sounds based on injecting noise into the S&H mixer, so I suspect that the outputs from the two noise generators were similar, and what I was hearing was low–pass filtering in the audio signal path. So that brings us neatly to...
Filters, Amplifier & Contours
As I’m sure that you’re already aware, the KARP claims to offer all three of the major filters types that appeared within the original synth (see box). Trying to describe these subjectively is probably a recipe for disaster but I hope that I’m on safe ground if I tell you that type I is ‘hotter’ and fizzier than type II and type III, and that it allows bass frequencies to pass unimpeded while self–oscillating, which is as it should be. Type II is possibly the ‘roundest’ of the three (whatever that means) and type III is slightly duller than the other two when fully open, which, again, is as it should be. Nonetheless, I wanted to be sure that Korg’s claims stood up to close inspection, so I decided to measure them. Although I didn’t have access to all of my test gear, I was nonetheless able to demonstrate to my satisfaction that, while the KARP’s filters may sound similar to the originals, they are not the same devices.
The original 4023 filter is reputed to have offered a maximum cutoff frequency of around 35kHz, whereas the 4035’s was reputed to be closer to 20kHz. Finally, the 4075 offered a maximum cutoff frequency of under 12kHz, although it’s claimed that this was later increased to 16kHz. By setting the resonance to maximum and measuring the frequency of the self-oscillation, I was able to show that the three filter modes in the KARP have maximum cutoff frequencies of 20.7kHz, >22kHz, and 16.5kHz when using additional CVs to push them beyond what you can obtain from the frequency fader alone. I wouldn’t get too hung up about these numbers — there’s a lot more to a filter than this single attribute — but, given these figures, I think that Korg’s marketing department was perhaps stretching the point a little to claim that, “These distinctive filters have been reproduced just as they originally were.”
Following the low–pass filters (all of which track almost 1:1 across the width of the keyboard) the synths each offer a high–pass filter. However, you can’t control this in any way other than by moving the slider, so there’s little that one can say about it other than that its effect appears to be all but identical on both. So let’s now move on to the contour generators and the VCA.
As you would expect, the dual ADSR and AR contours and their auto/repeat capabilities have been carefully recreated on the KARP. On testing some of the time constants, I found that (for example) the ADSR Decay was slightly longer on the new synth, while the AR Release was slightly longer on the original, but these differences were, in my view, insignificant. However, there’s a big difference in the audio signal paths of the two synths, and it’s an intentional one. Korg have added an overdrive circuit between the VCF and the VCA, and you can use this to add girth to the sound when the signal is of moderate amplitude, and add mild distortion when the signal is already close to (or at) its maximum permissible amplitude. This is a simple addition, but a very welcome one.
If you want to push the synth still further, the addition of the headphone socket on the rear panel makes it possible to do something that, in the 1970s, was rarely discussed in polite circles. While people seemed happy with the idea of feeding back one of the Minimoog’s outputs into its external signal input to fatten up the sound, I don’t remember it being discussed for the Odyssey. Perhaps this was because you needed more than just quarter–inch jack cables to make it work but, for whatever reason, it was considered to be a Moog trick, not an ARP one. That has now changed; just connect the KARP’s headphone output to the external signal input and raise the level and... Ooh, that’s thick. Keep going and, with high signal levels, you’ll reach the point where the synth screams, writhes, and begs you to stop. The results can be remarkable, but perhaps the world wasn’t ready for this in 1975!
Despite offering a keyboard that was neither velocity nor pressure sensitive, despite the lack of conventional performance controls, despite having brittle faders, and despite being impenetrable for most musicians of the era, the ARP Odyssey excelled equally as a lead synth, a bass synth, an orchestral synth and as a sound-effects generator. So, having analysed the KARP in a rather dry and academic manner, I decided that it was now time to set up some exciting sounds on both synths and study the differences between them.
This is what I expected: starting with the same brass patch as in Anaheim, I would set the switches and move the KARP’s silky–smooth faders to the same positions as on my Odyssey and listen to the results. I would then adjust its oscillator levels to obtain the right degree of drive, adjust its filter settings and the amounts by which the contours were modulating its cutoff frequency to obtain the same, brassy response, tweak its contours further to make the two synths ‘speak’ in the same way, and finally adjust its LFO–generated vibrato, all of which would result in two sounds — one from the Odyssey and the other from the KARP — that were almost indistinguishable from one another.
This is what happened: I set the switches and moved the KARP’s silky–smooth faders to the same positions as on my Odyssey and listened to the results. To within a gnat’s wotsit, they were the same. I then performed a fanfare, playing alternate notes on each synth. Seamless. Bloody seamless! So I went on to duplicate many other sounds, and I was gobsmacked at how similar I could make even arcane patches sound. In short, the KARP really is an Odyssey.
Or, rather, it’s a little over 50 percent of an Odyssey.
So that’s the physical volume of the KARP relative to a real Odyssey, but you have to ask yourself why Korg’s engineers would take the time and trouble to design and build such a wonderfully accurate sonic recreation, and then stick a miniature keyboard on it. Come on guys, ’fess up. Whose mind–numbingly daft idea was it to take the world’s second most important soloing synth and make it impossible to solo on it? Whom do I shoot?
Sure, things such as the difference between the KARP’s smooth 42mm faders and the original Odyssey’s scratchy old 45mm faders are irrelevant, and I realise that there are many people for whom the difference between three octaves spanning 443mm rather than the regulation 513mm will be of little consequence. But I’m not one of them, and I don’t feel like being reasonable about this. I love the ARP Odyssey and this is... well, it’s just wrong.
So I did the only thing I could. I took the MIDI Out of a full-sized synth and connected it to the MIDI In of the KARP. Joy! Bliss! Ecstasy! (In a legal way, of course.) Five octaves of sheer early–’70s synthiness. What’s more, the KARP’s duophonic voice structure is still interpreted correctly over MIDI, with VCO1 remaining low-note priority, VCO2 remaining high-note priority, and everything between being ignored. That’s excellent. Unfortunately, the only two things that the KARP understands are MIDI Note On and MIDI Note Off. I didn’t expect velocity or aftertouch (which, of course, are not implemented on the KARP itself), but not even pitch–bend nor modulation are implemented, which seems a bit mean — and philosophically dubious — given that they could be routed to the same places as the outputs of the PPC. Likewise, the only data that the KARP can transmit via USB are Note On and Note Off messages. Not even the performance controls generate MIDI information. I suppose that, given what it’s able to receive, that’s not surprising, but it’s extremely disappointing because it means that you can’t sequence and replay performances that use anything more than just the keys. All your pitch–bends and vibrato from the PPC are lost. Again, whose daft idea was that?
It looks like an Odyssey. It sounds like an Odyssey. In fact, if we ignore the considerable benefits of smooth faders, three filter characteristics, and overdrive, it is an Odyssey in most respects. So let’s give credit where it’s due: Korg’s engineers have done a fabulous job in recreating this ground–breaking synthesizer. But why did they miniaturise it so that the people who love it most can’t play it as nature intended? And why did they omit pitch–bend and modulation from the MIDI specification? I accept that these shortcomings won’t concern everyone but, as someone who grew up soloing on an Odyssey, I don’t get it. I really don’t. From my first Korg 700 in 1974 to the OASYS I use today, I have always had huge respect for Korg, but, right now, I just want to jump on a plane to Tokyo, drive out to the factory and grab the first person I meet by the lapels and cry out, ‘WHY?’
But perhaps there’s hope. When Korg recreated the MS20, they first did so in a miniature format (the MS20 Mini) and only later released the full–sized version (the MS20 Kit) and then, most recently, an enhanced module (the MS20M). Let’s hope that history repeats itself. A full–sized KARP with all of the attributes of this iddy–biddy one will be almost irresistible, as would an enhanced module that takes us further down the road toward the ARP2600. Until then, you just have to ask yourself whether size is important.
The 1970s have become a playground for present–day synth designers, with clones of huge modular synths now available, more variations on the theme of the Minimoog than you can count, accurate recreations of the MS20 and now the Odyssey... plus workstations containing software versions of classic synths, and plug–in soft synths of almost everything that matters. So, in addition to the excellent GForce Oddity 2 soft synth, you would imagine that there would be lots of hardware alternatives to the KARP, wouldn’t you? Strangely, there are not, and I’m unaware of any current alternative that does precisely the same job. But if your criteria centre on the idea of a miniature analogue monosynth rather than the ARP sound itself, alternatives include the Korg MS20 Mini and the Arturia MicroBrute. If you dispense with the mini–keys, you can then add to the list the MS20 Kit, the MiniBrute, the Moog Sub Phatty, the Novation BassStation II, the DSI Mopho and many others.
A Brief History Of Odyssey Filters
Launched in 1972, the earliest Odyssey (the Model 2800) incorporated a 12dB/octave filter (the 4023) with a frequency response that extended out to nearly 35kHz. This allowed the synth to sparkle like some modular synths, and it still makes ‘white faces’ the darlings of Odyssey aficionados.
In 1975, ARP announced a revised Odyssey that claimed to offer many improvements over the original. The Model 2810 replaced the 4023 in favour of the 4035, a 24dB/oct filter that sounded great but trod on the toes of some of Moog’s intellectual property, so this lasted for just one revision and was then replaced on the Model 2811 by ARP’s own 24dB/oct filter, the 4075. In principle, this was an excellent device, offering low noise and low distortion as well as improved stability when self–oscillating, but somebody had calculated the surrounding component values incorrectly, so it had a maximum cutoff frequency of less than 12kHz, making Models 2811 to 2813 sound duller than their predecessors. Fortunately, it was relatively simple to improve matters, either by replacing four resistors or by piggy–backing four resistors onto the existing ones.
When the final major update of the Odyssey arrived in 1978 (the Model 2820), it claimed an improved frequency response of 16kHz, and this remained unchanged until the final version (the Model 2823) was discontinued in 1981.
Connectivity: MIDI, USB, CVs & Gates
ARPOdyssey_02The rear panel of the KARP looks much like that of a MkIII Odyssey. It shares the high and low level outputs of the MkIII, as well as the external audio input, and offers the original inputs for a portamento on/off footswitch and a general pedal that can be routed to the pitch of VCO2 and the filter cutoff frequency, along with the six CV, Gate and Trigger Ins and Outs that first appeared on the MkII. Finally, in the analogue domain, there’s a headphone output with an associated level control. No original Odyssey had this and, as explained in the main text, it’s a more welcome addition than you might imagine.
To the left of the rear panel, where the Odyssey has nothing but a nameplate, the KARP boasts a five–pin MIDI input and a USB type-B socket. Finally, the on/off switch has migrated from the top panel of the original to the rear of the KARP, and the mains cable has been replaced by a 9V DC input. I know that this arrangement makes it much easier to pass electrical safety tests, but I wish that manufacturers wouldn’t do it. Serious equipment deserves an integrated power supply, not a wall–wart.
37 mini–keys, without velocity or aftertouch.
Two (sawtooth and pulse/PWM waveforms) with sync and ring modulation.
One (sine, square).
Sample & Hold
One, with output slew.
Three versions equivalent to MkI (12dB/oct), MkII (24dB/oct) and MkIII (24dB/oct) Odysseys.
Cutoff frequencies: all claimed 16Hz to 16kHz.
All resonant and capable of self–oscillation.
Claimed 16Hz to 16kHz.
80dB dynamic range.
Two (one ADSR, one AR).
Attack 5ms to 5s.
Decay 10ms to 8s.
Release 15ms to 10s (ADSR) and 10ms to 8s (AR).
(These figures are identical to those quoted in the original ARP Odyssey service manual.)
PPC (bend up/down, vibrato), ±2 octave switch, portamento, plus portamento on/off and pedal controller inputs.
Low out, high out, headphones out, external signal in.
CV In/Out (1V/oct), Trigger In/Out, Gate In/Out (10V output).
MIDI In (five–pin DIN), USB type B.
498 x 379 x 120 mm.
Three versions: MkI, MkII and MkIII livery, all with wrap–around vinyl cases and PPCs.
One quarter–inch jack cable, one 3.5mm mini–jack cable, carry case, manuals.
In keeping with its MkIII livery, the KARP has dispensed with the pitch–bend knob of early Odysseys and adopted the three PPC (proportional pitch control) pads of later MkIIs and the MkIII. The idea is a good one; just press harder on the b and # pads to bend the notes down or up, and press on the ~ pad to add vibrato, with the amount of the effect being proportional to the amount of pressure applied. There’s just one problem. As on the original, the pressure that you have to exert is outrageous, and I found myself wrapping my hand around the case to obtain the required leverage, just as I did in the ’70s. I realise that this is authentic, but sometimes designers should learn from their predecessors’ mistakes.