With the MS20 Kit, Korg have taken the Ikea approach to selling synths — by insisting you build it yourself!
You have to admit that Korg have made some daring decisions recently. Take their announcement of the resurrection of the ARP Odyssey. Anyone who tells you that they saw that coming is lying. Then there's its collaboration with littleBits, providing valuable 'know‑how' to create modules such as the i32 filter. And what about the MS20 Mini? Sure, its mini‑keys are about as much use as a chocolate teapot to some players, but there's no doubt that it's been a success, leading to widespread speculation that a genuine recreation of the MS20 costing just a few quid more must be following not far behind. This has now proved to be both right and wrong. I have before me a sealed box containing a full‑sized MS20. But it costs considerably more than a few quid more, and I have to build it myself. What on earth is going on?
I started by confirming that all was present and correct, with welcome spares of all the screws, nuts and washers. Next, I read the instructions from beginning to end. I realise that this flies in the face of all that is British and male, but live with it... I read manuals. Having satisfied myself that construction was going to be straightforward, I laid everything out on a white sheet on my dining room table (it's much easier to find the tiny washer that you just dropped on a white sheet than on any other surface) then put on a pair of thin latex gloves to protect both the synth and my fingers from nicks and scratches, and began.
Disassembling and reassembling an original MS20 can be a bit fiddly but, thanks to a redesign of the case and board mountings, building the Kit turned out to be a doddle. First, I fixed the electromagnetic shielding foil to the insides of both end cheeks and mounted the mod wheel to the left-hand cheek. Next, I attached the rubber feet to the underside of the base plate (which, to my surprise, had a couple of tiny nicks in the paintwork), mounted the keyboard and I/O board on top of it, and then attached the cheeks. Moving on, I screwed the two L-shaped brackets to the rear of the front panel and, being careful to ensure that the LEDs were correctly positioned, mounted the three main circuit boards as well as the on/off/volume control to it. This proved to be the most time‑consuming task because every knob and socket had to be fixed using the right type of nut and, where appropriate, washer. Once that was done, I mounted the front panel within the cheeks, and connected the various wiring harnesses, which are designed so that you can't stick the wrong male thingy into the wrong female thingy. Finally, I attached the rear panel and the hook for the power cord, and then placed the 36 knobs on their shafts. Et voilà... a Korg MS20!
Let's face it, assembly was hardly rocket science (although neither is brain surgery) and I'm confident that almost anyone who has ever played with Meccano or even Lego could complete the Kit successfully. The only time that I had to retrace my steps was when I noticed that I had folded one of the hum shields when attaching the front panel. Removing a couple of screws and pushing the shield back into place sorted this out in a minute or less. The only thing that I found tricky was inserting the little rubber grommets that cover the trimmers into their holes in the control panel. So, thinking that I might need to calibrate something later in the review (which I didn't) I left them off, pretending to myself that I had intended to do so all along.
The whole operation took under two hours and I suspect that, had I chosen, I could have assembled it in half as long. But I'm glad that I took the time, not just because it ensured that I got everything right, but because I thoroughly enjoyed myself. The truth is, I think that Korg has pulled off a masterstroke here. Building the MS20 Kit made me feel that I had put something of myself into the resulting instrument and, while the assembly process is superficial, I nonetheless appreciated the resulting instrument more than if I had pulled a finished unit out of its box. I suspect that many others will feel the same.
Do you remember that curious statement from Korg that, "it would be devaluing to make [the MS20 Mini] exactly the same size as the original”. I thought that it was daft at the time, and I still do. Which leads me to the size of the MS20 Kit, which isn't physically identical to the original, although I expect that it was meant to be. I placed it next to a vintage MS20 and found that the Kit's keyboard is about 5mm wider, and that the case is both a few millimetres longer and deeper than the original. It doesn't make any difference, but you have to wonder if there was a reason for this.
Internally, there have been far too many changes in the laws pertaining to electronic components for the Kit to be remotely the same as the vintage synth, and Korg's engineers would probably be banged up at his Emperor's pleasure if they tried to build an exact clone. So, where the original synth had chunky discrete components and chunky DIL chips, the new one has the tiny discrete components and the tiny surface‑mount chips that are legal in 2014. Likewise, the chances of the Kit meeting safety regulations with the original power supply would have been zero, so it was entirely reasonable for Korg to power it using an external 12V PSU. The blurb promises that the voltages supplied to the power rails inside the synth are the same as before, and that the change should have no effect on the sound. So the $64,000 dollar question is... does the assembled Kit sound the same as a vintage MS20?
I started to check this by comparing the oscillators on the Kit against my MS20 Blackboard and a standard MS20. I was impressed. VCO1's square/pulse and sawtooth waves were all but identical, as were VCO2's square and sawtooth waves. VCO1's noise generator was slightly bluer and VCO2's pulse wave was slightly wider, but the differences were nothing to worry about. Impressively, VCO2's Ring setting (which isn't ring modulation, despite its name) was also very close to that of the Blackboard but, given the similarities of the square/pulse waves from which it's derived, that shouldn't be a surprise. The only waveform that was noticeably different was VCO1's triangle wave, which was duller and, therefore, more accurate than the vintage synths'.
All previous MS20s have had two filters in the primary audio signal path: a 6dB/oct HPF followed by a 12dB/oct LPF. Early models used a filter called the Korg35, while later ones used a more conventional OTA‑based device that isn't held in as great esteem by aficionados. The MS20 Kit is the first version of the MS20 to offer both versions, although not simultaneously, and you can select the default filter by setting a jumper on the main board, or choose either by pressing the appropriate key combination when switching on the synth. Unfortunately, the manual tells you about this jumper after you finish building the Kit, so reading it beforehand had indeed proved to be useful!
Comparing all three filters (the HPF and both LPFs) on the Kit to the HPF and LPF filters on the vintage synths revealed that all of the maximum cut‑off frequencies were a bit higher on the originals. The differences weren't significant in my view, and may even have been the consequence of something as trivial as trimmer settings. I found a more noticeable difference when I checked the waveforms generated when the filters were self‑oscillating. These were brighter when generated by the Kit, but sweeping the self‑oscillating filters while playing a rich waveform through them produced almost identical results on all three synths. This was impressive.
There was, however, one thing that concerned me. The background noise generated by the Kit when in Korg35 mode was considerably higher than the noise generated by either of the vintage synths and, with just VCO1 generating a triangle wave, the mixer set to around '5' and the filters wide open, it was as loud as the wanted signal. Interestingly, the noise level in OTA mode was almost negligible. So I contacted Korg, and received an explanation from the General Manager of the Product Planning Department in Tokyo confirming my findings. He explained that the signal/noise ratio in Korg35 mode could have been improved, either by changing the transistors in the filter or by increasing the output levels of the oscillators, but when his engineers tried the first of these, they were unable to find alternative components that reduced the noise while retaining the sound of the original. The second approach was dismissed because the filter input level has a significant effect on the timbre, and making this 'hotter' would have changed the sound of the synth. In his words, "We faced a hard‑to‑decide trade‑off, and in the end we chose to prioritise the sound of the filter. Replicating the sound characteristics of one of the most iconic synthesizers in music history was our prime focus.” Fortunately, the noise isn't a problem with louder patches (for example, using square or sawtooth waves with higher mixer settings), but you may want to stick to the OTA filter for gentler sounds.
Next, I turned to the HASDR contour generator. The blurb quotes maximum times for the A, D and R stages as 10 seconds in all cases, but I measured them to be nine, 40 and 40 seconds respectively. I wasn't able to obtain the slow crescendos on the Kit that I can on my Blackboard (which has a maximum Attack of nearly 20 seconds), but the other figures were very much in line with the vintage synth. More relevant, perhaps, is that the key on/off thump generated when the Attack and Release are set to zero and the HASDR is directed to the filters can be somewhat louder on the Kit, accentuating percussive sounds in ways that you may or may not like. Happily, there was no evidence of the voltage calibration error that I discovered on the contour generator of the MS20 Mini review unit, and the voltages output by both the contour generators were very close to their nominal values.
Continuing my comparisons, both of individual facilities and of patches, I found that adjusting things carefully could eliminate most differences between the synths. So, for example, when I found that the CVs on my Blackboard were 'hotter' than on the Kit, I could often compensate for this by adjusting the appropriate level control knobs. Even modulated patches with bucketfuls of filter resonance could be made to respond similarly, and quirks such as the wheel‑to‑VCO2 behaviour that I described in my MS20 Mini review were also recreated accurately. Of course, really complex patches on one synth weren't always recreated perfectly on the others, but that was as true when comparing the two vintage synths to each other as it was when comparing either of them to the new one. Ultimately, these were just three different flavours of MS20; one with a Korg35 filter, one with an OTA filter, and one with both.
Given the above, you're probably wondering what the differences between the built kit and a vintage MS20 are. Ignoring the physical changes, the dual filter options and the external PSU, there's really only one: the Kit offers a five‑pin MIDI In and MIDI In/Out on USB. Communication is limited to note on/off on Channel 1, but it's still very welcome, and you can always direct CVs from a suitable MIDI/CV converter to the patch panel to make patches velocity and pressure sensitive, and to control modulation using MIDI CCs. Perhaps it's more informative to note the major differences between the Kit and the MS20 Mini. Again ignoring the dual filter options, there are three of these. Firstly, I'm delighted to report that the action of the full‑sized 37‑note keyboard on the Kit is a million miles from the squidgy mini‑keys on the Mini. Secondly, the pots on the Kit are solid, whereas those on the Mini can be a touch wobbly. Thirdly, the quarter‑inch sockets on the Kit make rock-solid connections. There's none of the wiggling sometimes required to ensure a connection when using 3.5mm sockets, and I suspect that you'll still be able to patch the Kit reliably in 25 years, whereas the Mini will, I fear, be at best intermittent. Where keyboards, potentiometers and patch connections are concerned, size really does matter.
Rather than plunging in headlong, Korg have spent the past few years dipping their toes back into the waters of analogue synthesis. Small analogue boxes and a synth with miniature keys, no matter how successful, do not suggest a plunge to me. But the MS20 Kit is a much more significant statement; perhaps the purest reinvention of a classic analogue synth since... well, since forever. I suspect that it would have required a significant amount of effort and cost to convert the original design to modern electronics while retaining the sound and quirks of the original to the remarkable degree that has been achieved. (With both filters on the same board, it wasn't just a question of making the MS20 Mini board a bit larger.) Nonetheless, some people will complain about the price. At around £1100 in the UK$1400 it's not cheap, but, with prices of second‑hand units peaking at around £1600 on this side of the pond, it's still a lot more affordable than a vintage MS20. When you consider that the £425 I paid for mine in 1978 is equivalent to well over £2000 today, it suggests that the Kit really isn't excessively expensive.
But what of the MS20 Mini? Cynics might suggest that the Kit is just 16 percent more synth for around 120 percent more money, but that misses the point. For many players, the additional cost will be more than justified by its full‑size keyboard, twin filters, quarter‑inch sockets and solid knobs. Sure, you can buy all manner of alternatives for less, with enough change left over for a digital effects unit to go with them, but that's still not the point. The MS20 is the favourite instrument of a significant number of analogue synth fanatics, and many of them will already be raiding their piggy banks to snap up one of the 1000 units that Korg are manufacturing. If recreations of any of the other models in the range (the MS10, MS50, VC10 and SQ10) are announced, I fear that I may be among them. In the meantime, it's no longer possible to doubt the company's commitment to analogue synthesis, and who knows what monophonic, duophonic or even polyphonic goodies might be around the corner.
Despite what you'll read elsewhere, the MS20 is not a modular synthesizer because you can't insert or remove modules. In fact, it only just deserves the description 'semi‑modular' because, unlike (for example) an ARP 2600, you can't tap the signals from the oscillators, nor from the signal path between the filters, nor from before the audio VCA. The patch panel looks beguiling, but it only allows you to affect the modulation applied to the signal, not the audio signal path itself.
Consequently, I'm stunned that Korg have made no mention in their blurb of the 13 additional patch points on the Kit's main board, with a further two on its patch board. These are not trivial; indeed, they not only seem to provide access to the audio signal at every point in the signal chain but also promise to add oscillator sync and PWM, which would hugely increase the range of sounds obtainable. If Korg doesn't build an MS20 Expander to take advantage of these, I'm going to fly to Tokyo and bang on the factory door until somebody does. In the meantime, I can see members of the Synth‑DIY community getting very jiggy with their soldering irons.
- It's an MS20. Really.
- It offers the low‑pass filters from both the early MS20s and the later models.
- It has quarter‑inch sockets, sturdy knobs, a full‑sized keyboard and MIDI.
- It will appeal to players who'll appreciate building 'their' synthesizer.
- The Korg35 filter mode is too noisy for quiet patches.
- To some, it will seem pricey.
The MS20 Kit is an excellent recreation of the original MS20. Many people will wonder what all the fuss is about. Others will sell their grannies to be able to afford one.
Korg UK +44 (0)1908 304600
Korg USA +1 631 390 6800.