Sometimes the underground is a surreal place to be.

Thursday, March 31, 2011

Gear Change: Keytek CTS-2000

At the recommendation of more than a few friends (and since I'm not able to upload music from any new rips I might make) I'm switching gears and will post my thoughts/reviews of music hardware that I own or have owned in the past. I've spent the last seven or eight years buying and selling synthesizers on eBay, keeping those that I liked and selling those that I either disliked, couldn't afford to hang on to, or simply needed the money to buy the next synth that caught my fancy. The first of what I hope will be many posts is dedicated to the rarest synth I've ever owned, which also happens to be my current favorite as well as the keyboard which easily wins the award for 'the synthesizer with the least information available,' the Keytek CTS-2000 (second place goes to the Elka EK-22, which has a superb write-up by legowelt and a Vintage Synth Explorer page all it's own).

Keytek was the final moniker of italian synth manufacturer Siel, who produced some of my favorite synthesizers (primarily the DK-70, a beautiful example of the sonic qualities of italian synths as well as of the poor quality which ghettoized them compared to the japanese and american offerings, but to give them their due I'll also mention the Opera 6 and the Mono). Siel was facing bankruptcy in the mid 80s, and their budget synth lines in the later DK double digit series couldn't buoy them up above the ravages of the post DX-7 synth market. Luckily for them, Gibson were looking to expand beyond the guitar markets into keyboards, so along with Oberheim the italians were purchased with an eye towards remaking their brand and exploiting their unreleased synthesizers. In much the same way as Yamaha invested in Korg in order to enable their survival prior to the release of the Wavestation and the M1, Gibson hoped that the engineering division that produced the superb DK-700 polysynth as well as several other extremely successful synths just a few years before (to this day rumors abound that Dave Smith or Bob Moog himself designed one or the other of the Siel line) would pay off.

Unfortunately, Siel's unreleased line didn't quite live up to that hope. Where Yamaha could simply license Korg their FM technology until Korg could exploit the knowledge of the engineers they hired away from Sequential Circuits after their collapse, Gibson had no familiarity with the synth market and Siel apparently had a completely fucking terrible bunch of engineers designing the interfaces for their gear. I mean, just godawful for the time. Not so terrible for us today, in certain ways, kind of.

Siel was rebranded as Keytek under Gibson, and they released a series of horrifically outdated products, with a single exception. They produced the Keytek MDP-40, which I owned for a while as the Siel MDP-40, which was a preset drum machine with poorly sampled 8-bit drum sounds and ultra-generic preset patterns, saved only for it's ability to sync with MIDI clock and ability to play each sample by the non-velocity sensitive pads as well as through MIDI notes (I should mention that I never tested this). Also in the product line was the CTS series, which consisted of the CTS-1000 preset keyboard, the CTS-2000 "Cross-Table Sampled" synth, and the CTS-5000 rackmount piano preset unit. The 1000 is reportedly a Casiotone level preset keyboard with no editability and unknown features (there's zero information about it out there), and the 5000 is a fairly respectable for 1987 electric piano unit aimed at the Roland MKS-20 market. The CTS-2000, however, is a true gem in sound and a prime example of bad design.

The architecture of the CTS-2000 is rather impressive to an analog synth aficionado: two sampled waveform oscillators (OSCs from here on), with two envelope generators (EGs) for tuning and frequency, a deliciously warm 24db Curtis Digitally Controlled (analog) Filter, with two Digitally Controlled Amplifiers and three more independent EGs for the two DCAs as well as the DCF. It also includes internal stereo mixing with level controls and eight placements across the stereo field and is capable of five part multi-timbrality along with it's eight voices. This is both a tremendous point in favor of the 2000, and an hilarious point against.

Before I get to that, I should also mention that the CTS-2000 features 6 faders (not including the volume fader) which are the primary controls of the synth settings, although there is the by-this-point standard up/down increment buttons. The programming works through a menu system on a 2X16 backlit LED, where each setting on a given menu is spelled out above the button on the right side of the keyboard, which matches to one of the 6 sliders on the right side, and to select a given function one must move the slider. Annoying, but slightly more hands on than on a Roland Alpha Juno or JX-3P, especially since everything changes in real time (that includes filter cutoff and resonance) and once you've selected the appropriate menu and placed the sliders to the setting you want, they'll match the existing setting. With the immediacy of the LED readout, it would work quite well for the analog freaks out there.

The bit of hilarity comes in here, once you've figured out how to actually do a little bit of fiddling with the internal presets. The CTS-2000 was designed with the future of MIDI in mind, and multi-timbrality was essential for any new synth in the marketplace of 1986/7. Unfortunately, the engineers of Siel/Keytek either didn't quite grasp how to make this a consumer proposition, were constrained by Sequential Circuits patents on multi-timbral analog equipment, figured that a market which bought 100,000 DX-7s wouldn't have a problem with a ridiculous interface, or were just pressed for time to put out a synth after the Gibson buyout: Either way, the CTS-2000 has one of the most ridiculously idiotic implementations of multi-timbral synthesis ever conceived.

The keyboard is five octaves. Each of the five octaves must be programmed individually. That is, each octave of the CTS-2000 acts as an individual synth which can play a different patch from the other four octaves, with the 8 shared voices able to change for each octave/note demand. It's kind of impressive for the time, if it weren't so incredibly stupid given that each octave must be programmed individually. Thankfully, there's a kind of copy-paste available, although it doesn't work quite right (it's a bit like loading a Roland sample bank on an Akai sampler or vice-versa, there's bits left out that you have to reprogram to get right). The end result, and apparent intent, was that the CTS-2000 acts as five different polysynths, each limited to a single octave and sharing eight voices among them, kind of like some digital version of the Oberheim 4-voice or Korg Mono-Poly , except you can assign each octave of the 2000 to any given octave, and each octave to any given MIDI channel. This brings me to the MIDI implementation.

In his commentary on the Elka EK-22, Legowelt mentions the MIDI implementation in particular. That is, he mentions that the thing drops MIDI notes when stressed too much. Like when you send two messages to it at once. The biggest weakness of the CTS-2000 lies here. The keyboard itself is absolutely wonderful, the action is easily the best I've ever played (and I'm comparing this to every model of the Korg Poly series, all of the Roland Junos, Kawai's K series, Sequential's Sixtrak, Split-Eight, and Prophet-600, even Akai's AX-60 and X7000), it just feels wonderful to play and the velocity response is nice and on point. I imagine this is the same thing people talk about when they say the Elka EK-22 plays like butter. I want to use this as my master keyboard, but that's where I run into problems.

When I set the CTS-2000 up as my master keyboard, I run into issues with the internal capability to keep up with MIDI messages. For example, the preset patch "JAKO-FIFTH" is absolutely going to be the foundation of a future track. The keyboard is divided so that the bottom two octaves are the JAKO solid bass patch and the upper three are a beautiful evolving analog-style fifths pad. When I try to play both patches at once, transmitting MIDI to my sequencer on two MIDI channels, the keyboard misses notes. That is, it literally misses notes that I play in real time, failing to produce sounds when keys are pressed. This explains why there is a MIDI On/Off button, as well as a MIDI Local/Remote button (I have yet to discern the difference between the two, since the MIDI Off setting should do the same thing as the MIDI Local setting does, but I'm not much of a hacker). This is a 'feature' that the CTS-2000 shares with the EK-22: the internal processor is fully capable of handling the various settings the keyboard itself was designed to allow, but nothing beyond that.

Having said all that, I'd like to move into the actual capabilities of the thing. The CTS-2000 has (according to the promo lit) 333 possible waveforms, these provide the basic sound which is processed through the analog filters and is shaped by the various envelope generators. The waveforms are divided into a series of three different sounds which are shared between the two 'oscillators'. The 'oscillators' provide a basic functionality for linking the three transitional waveforms together, a sort of attack/release for the transition between each sound. The resulting sound is comparable only with the implementation of multi-timbrality I described above. It's brilliant in it's own way, but terrible in comparison with similar synthesizers of the same period (namely the PPG and most other wavetable synths). It provides a good level of control over the transition speed between the waveforms, and what I suppose you would call the 'mixing' of the wave transitions is even and allows for very interesting pad sounds which would be impossible to reproduce on any other synthesizer. Add this in to the filtering capabilities and you've got an absolutely amazing source for pad sounds which couldn't be created any other way, even if sampled. I don't know how good the sysex implementation is, but I'd wager it's good-but-strange. If the sysex allows for external control of the waveform transitions as well as the other features (stereo placement control could be amazing), the CTS-2000 would be a monster easily surpassing the similar synths of the time.

The sampled waveforms are about what one would expect for the period, comparable to those found in the Korg DW series or the Kawai K3. Low bitrate, but when processed through the analog filters, very warm and a happy fit for mixing. The bass sounds in particular are phenomenal: the preset bass sounds are extremely usable, though they mostly suffer from being the bottom two octaves of a split with the higher three being some generic-sounding bit obviously aiming the patch towards one-man-band pub type situations (the JAKO-FIFTHS patch being the exception). However, that criticism was leveled at the 303/606 combo in it's time, so take it with a grain of salt. The bass presets are head-and-shoulders above the FM Solid Bass preset of the sought-after Detroit classic DX-100 or the Lately Bass preset of the TX-81z, and would easily hold their own against the M1 basses. The other samples tend towards same-ness, just as the Korg DW waveforms do, but the transitional nature of the design provides for a level of creative interaction which neither of the Korgs nor the Kawai K3 allow. The 'attack' of the sample transitions on the CTS can be set very fast or extremely slow, giving you the ability to provide a weird punchy bass blip or a long evolving lead.

My own taste in synth programming tends towards the moving pad chord style, and I really enjoy programming synths, so I'll advise against the CTS-2000 as a first keyboard. But for the heads? Get one of these if you can find it. I bought mine for $100, and I feel overpaid tremendously given the nonexistent market (I don't price synths based on what I feel they're 'worth', as anyone who has ever bought a synth from me will affirm, I value them based on the market price and essentially only worry about whether something is over-priced. That is, the less I pay, the better the deal, as opposed to "the less I pay, the worse the item.") The week after I bought my CTS from a seller in California, I saw a seller from Queens offering one for $200, which, I might add, sold for that price. I wouldn't pay more than $100 for one, but as I make more posts about synths in the future you will probably notice (especially in light of the above guiding rule) that I lowball prices.

Up next: Siel DK-70, Akai X7000, and possibly XoXBox.


  1. Greetings! I recently picked up a CTS-2000 and was trying to find any information regarding the sysex capabilities of this machine. The only thing I can find is the manufacturers id, but nothing else. Have you had any luck finding any more information out about this?



  2. Thank you, great information!

    i also recently got this synth.. pretty happy so far, i just don't have a cartridge so i can't save the presets. any ideas how to export sysex?

  3. unfortunately I don't have any documentation for the CTS-2000, and I'm not sure that I've gotten around to hooking it up to a MIDI monitor to figure out whether it actually sends sysex messages. From the front panel, there doesn't seem to be any method to do a bulk dump, and I also don't have the cartridge so I can't open it up and figure out how it's designed. That said, I picked up a CTS-1000G rack module last month which I'm eventually going to hook up to the CTS-2000 and see whether or not I can program the 1000 from the 2000, since it appears to have roughly the same presets and same basic hardware architecture. What little information there is about the 1000 indicates that this isn't possible, but I don't know if anyone has even tried.

    1. I have a CTS 2000 and received the printed manual with it. There is no SysEx part in it, so I am pretty sure the instrument does not support SysEx at all.

    2. Do you know where can I find a copy of the manual?

    3. I'd like to know too. Anyone?

  4. This was superbly entertaining, I hope you keep this up. I literally lol:ed when the multi-timbral functionality was revealed.

  5. You dont have to program all octaves separatly... you just have to press the zone/global button to switch from separate octaves(zone) to the full keyboard(global)
    This also works if you want to use it for midi applications.

  6. There is no manual (or at least i've never found one, not that you should need it since everything is spelled out on the front panel), and you do have to program all octaves separately if only to make sure that you don't have five of the same octaves playing when you play the keyboard. idk if yours had a better midi implementation but even in global mine would choke on midi i/o and freeze up each time i tried.

  7. BTW this is Turn, and i've since sold the CTS-2000, so I can't really help out much, but if you need help with one there are few places that talk about this old synth so if you want to post a question, hopefully someone else out there who has found this ancient post will respond. Good luck, synth comrades!

  8. Thanks a lot Turn for the memories - I had all but forgotten the Keytek. Let me add that IMHO the biggest problem was its dull and flat basic sound - and that I found that this was due to the output circuitry: To get rid of the metallic overtones of Lo-Fi samples, they had added simple low-pass filters. (Yes, the very same sound characteristic people try to add to modern digital sounds by bitcrushing.) I removed two capacitors (and thus the filter) with some pliers and liked the Keytek a lot better after that, but not enough that I wasn't extremely happy to sell it and get a Matrix-1000 instead.