Creative Control Strategies: MIDI to CV

Connecting the Old and the New

Naomi M · 07/21/21

For many years, the dominant control method for synthesizers was the paradigm of CV (control voltage) and gate (on/off messages), an analog standard that directly translates analog voltages to control parameters within a synthesizer. The specifications of the control scheme of analog synthesizers varied between many manufacturers, such as in Moog's 1V/oct, Korg's Hz/V, or Buchla's 1.2V/oct pitch scaling schemes, as well as in their differing voltage range and gate/pulse formats. With the advent of digital synthesizers and more specifically polyphonic digital synthesizers, manufacturers sought to overcome the inherent incompatibilities between their products by developing a universally compatible communication standard for new instruments. Out of this need arose a new digital communication standard: MIDI, or Musical Instrument Digital Interface. For a deeper dive into the history of MIDI, please see the A Brief History of MIDI article.

I'm going to assume you have some knowledge of MIDI, but I will summarize briefly. MIDI is a protocol that allows instruments to communicate with each other or a computer in order to transmit data. This data typically includes information pertaining to note start/stop (on/off), pitch, and velocity (volume), but may also contain other information such as pitch bend and the BPM of the source instrument or project. Typically MIDI messages are fairly low-resolution, featuring a range of integer-only values from 0–127. The pitch range of MIDI goes from C-1 (0) to G9 (127), with middle C, also known as C4, placed at the note value 60. Additional information may be sent over continuous controller (CC) messages which send a range of values from 0–127 to preset or assignable destinations.

One of MIDI's great advantages is that it allows transmission of a large amount of simultaneous control data over a single cable. Compared to the one-cable-per-parameter approach of modular synthesizers and early analog keyboard synths, this offers a great reduction of clutter, and makes it much simpler to create complex, recallable control structures between multiple instruments. While this has distinct advantages over standard CV/Gate-based approaches, though, modular synthesizers weren't made obsolete by MIDI: and today, it is relatively common for studios and performance setups to employ a combination of MIDI and CV-capable gear. The trick, then, is about how to get these different types of gear talking with one another—and one of the most common approaches involves translating MIDI data into CV. (The opposite approach is also possible—and we'll discuss it in more detail in a forthcoming article.)

What is a MIDI to CV Converter?

MIDI to CV converters take MIDI messages and convert them into voltages a modular synthesizer understands. MIDI to CV converters cover a wide range of features, and different converters may work better or worse for different applications. Some of the simplest options convert MIDI note information to a pair of Pitch CV and gate outputs, great for using a MIDI keyboard or sequencer to provide straightforward pitch and timing information to an analog synth voice. Others are better-suited for sending many CC messages as CV signals. Some occupy a small form factor for when space is limited, while others work as a drum machine sequencer with many gates or trigger outputs for initializing envelopes, drum modules, or samples. Knowing a bit more about what is possible, let's take a look at current and historic MIDI to CV converters that fit a wide range of applications, and maybe assist in finding the right one for you.

MIDI to CV Converters of the Past

While the popularity of modular and analog synthesizers waned in the 1980s and 90s, the diehard users stayed on and wanted to integrate their analog synthesizers with MIDI controllers.

One early device for converting MIDI to CV came from the company that would go on to start the Eurorack modular format, Doepfer. In the mid-1980s, Doepfer developed the MCV8, an eight-voice MIDI to CV interface. Originally, it featured a single mode, eight-voice polyphonic, with little in the way of customization and no memory. Version two added additional modes, including 2 x 4 voice and 4 x 2 voice. The CV outputs came in the form of a D-SUB connector, necessitating a specialized snake cable to convert the signals to 1/4" connections. Version three (whose spectacularly minimal design can be observed in the image above, pulled from its spectacularly minimal user manual), a rackmount version released in 1990, broke out the connectivity on 16 1/4" connections, eight for CV and eight for gates. It also provided additional controls for parameters such as legato, pitch bend, gate polarity, and a selection between 1V/oct and Hz/V. It included six standard modes of operation and memory slots for up to six user presets. It's pretty wild to consider that this device was always set up to act as a polyphonic MIDI to CV converter—and given the fact that we know that modular polyphony is no easy feat, we have to wonder what sort of systems or other instruments this device was commonly used with.

One of the most interesting classic MIDI to CV converters is the MPU-101 from Roland, a surprisingly advanced device for the time. It features four output channels, each with CV, gate, and dynamics. Additionally, it features individual outputs for pitch bend, modulation, volume, and aftertouch, a feature still not found on many modern MIDI to CV converters. MPU-101 operates in five primary modes: monophonic, two-voice, three-voice, four-voice, and special. In special mode, the highest, latest, and lowest played notes correspond respectively to outputs one, two, and three. This leaves output four, assignable in this mode to any of the three note priorities. Rather than relying on round-robin voice allocation, this mode makes it possible to come up with interesting control schemes in which particular registers are always assigned to particular voices...and again, we have to wonder what this sounds like in use.

The MPU-101's tuning calibration mode outputs a static output voltage in order to properly tune your oscillators. (Roland made the bold choice of defining A4 as 442Hz? Please send in your angry comments by post.) A final tune knob and transpose switch affect all four outputs simultaneously, allowing for a global fine-tuning of approximately +/-100 cents (translating to a transposition of plus or minus one octave). The gate outputs work with either standard polarity, V-trig (positive-going), or Moog's S-trig (negative-going). The MIDI thru and out ports allow for two MPU-101s to be used in unison, allowing for eight voices of MIDI to CV goodness.

Other early MIDI to CV converters include the MV8 made by PAiA and the short-lived Serge/STS CV8.

Considerations: Tuning and Resolution

The tuning of the destination oscillator is extremely important when tracking with MIDI. Generally speaking, you want to tune your oscillator to the lowest note in the sequence. Send a MIDI note of a C1 to your oscillator and then reference a tuner or hertz counter on an oscilloscope and match the oscillator's pitch to the incoming MIDI note. The standard concert pitch in the UK and US is A440, meaning that the A note above middle C (A4) is tuned to 440Hz. This is one way of tuning your oscillators, but by no means the perfect or only way to tune them. For example, as we saw above, the MPU-101 tunes A4 to 442Hz, a slight variation that I'm sure some people can distinguish. This type of fine-tuning is typically not required for parameters other than pitch—our ears are, of course, very sensitive to changes in pitch, but not as sensitive to changes in loudness, timbre, etc. So if you're using MIDI to control pitch of an analog device, expect to take a little bit of time tuning things up.

Due to the fact that MIDI is a digital protocol with limited resolution, fast CC changes may incur some amount of aliasing and stair-stepping as they convert into the analog realm. An easy way to eliminate this is with an external slew generator, which will smooth out the hard digital edges to ensure smooth voltage changes. Again, the success of this approach depends on the relationship between the slew rate and the rate of change of the CC value—so you may need to do some tweaking in order to get completely smooth parameter changes.

Physical Considerations

MIDI originally employed a 5-pin DIN connection, but two of the five pins, pins 1 and 3, were unused. DIN remained a popular standard for decades (and is still common on all sorts of electronic music gear), and eventually, with the development of USB, MIDI became a standard part of the USB specification. As such, USB has also become a fairly standard way of communicating MIDI between multiple devices (more thoughts on that below!).

DIN connectors are relatively large, and the unused pins mean a certain amount of wasted real in a world where electronics have grown more and more compact, DIN MIDI connections have started to seem a bit clunky. Because of this, many manufacturers have begun to use another common connector type for MIDI communication: 3.5mm TRS. This three-conductor connection standard is commonly found on stereo headphone jacks and other audio equipment, so the parts are both readily available and occupy a small footprint. While the same size as Eurorack connections, note that TRS MIDI does not work if plugged directly into a 3.5mm CV/gate input. In order to convert MIDI information into something that your Eurorack modular or semi-modular device understands, you still need a dedicated MIDI to CV converter.

TRS MIDI comes in two distinct forms, Type A and Type B. Companies such as Korg and Make Noise use Type A, adopted as the MIDI standard in 2018. Although not the official standard, Type B still retains some popularity, including its use by Arturia and 1010 Music. Devices from all these manufacturers typically come with a dedicated TRS to 5-pin DIN adapter cables, allowing the device to connect to devices with the older standard 5-pin DIN MIDI jacks. Using a pair of TRS to DIN converters allows you to convert from A to B and vice versa.

The XOXO Modular MIDIXO takes the hassle out of MIDI TRS types, offering a simple solution to switch between the two standards. On top of that, it works with Expert Sleepers devices such as the FH-2 and Disting Mk4, adding TRS MIDI connections to them for general-purpose interfacing and uncovering some under-the-hood features. For more information on Expert Sleepers' MIDI offerings, see the recent Signal article on them: Which Expert Sleepers Interface is Right For You?.

Another consequence of the aforementioned adoption of MIDI as part of USB protocol was the development of USB MIDI controllers—in the present day, it has become common for MIDI controllers to offer any combination of DIN, TRS, and USB MIDI connections. In fact, many MIDI controllers intended to be used with computers now feature only USB connections. In order to use a MIDI to CV converter with such devices, a separate MIDI host is required, such as a computer or a device such as the USB MIDI Host from Kenton. For more info on USB Hosts, please see the article What is a MIDI USB Host?. Devices like the Expert Sleepers FH-2, Shuttle Control, and Polyend Poly2 feature built-in USB hosting, allowing you to plug your external controller directly in and use it with no hassle. (Coincidentally, these are some of the most interesting and multi-functional MIDI to CV converters altogether.)

Because MIDI is a 40-year-old technological standard, some improvements and extensions have been made over the years. MPE, or MIDI Polyphonic Expression, expands polyphonic expression by allowing control over the parameters of individual voices instead of the global control common in most MIDI-capable instruments and controllers. This includes individual pitch bend and timbre per-note, allowing for improved dynamic and expressive playing when using a compatible controller and instrument. At the January 2020 NAMM show, a proposal for an official MIDI 2.0 protocol brought together many manufacturers as they worked to introduce a new interaction of this nearly universal standard. One of the most important features of MIDI 2.0 is backward compatibility, allowing new MIDI instruments and old ones to communicate seamlessly. While we have yet to see a MIDI 2.0 to CV converter, it is only a matter of time. MPE-compatible MIDI to CV converters are relatively common, however: again, Polyend's Poly2, the Shuttle Control, and Expert Sleepers FH-2 are some of our strongest recommendations.

Standalone MIDI-CV Options

Of course, not everyone has or wants to have a Eurorack system. Over the last few years, we've seen the proliferation of standalone synthesizers that include some sort of CV/gate integration (and for that matter, many vintage monosynths also offered CV/Gate functionality in the days before MIDI). For out-of-rack MIDI to CV conversion, devices like the Kenton Pro Solo Mk3 and the Doepfer Dark Link provide standalone conversion of MIDI signals into CV and gate. If your needs extend beyond converting a single monophonic instrument line into CV/Gate, then you could also consider the Kenton Pro 2000, which offers two CV/Gate output pairs, a dedicated clock output, and six assignable auxiliary outputs.

Some of the standalone synthesizers with CV/gate integration offer a MIDI implementation which allows them to act as a MIDI to CV converter without the need for an external one. Instruments such as the ASM Hydrasynth, Moog Mother-32, and Make Noise 0-Coast all feature the ability to route incoming MIDI data to the CV outputs. Many times, standalone semi-modular synthesizers come recommended to new users looking to get into modular synthesis. Whether they intend to expand their gear collection or not, integrated MIDI to CV conversion allows for control possibilities that extend beyond the self-contained instrument alone.

Eurorack MIDI to CV Converters

Of course, if you're trying to add MIDI control to your modular system, an in-rack solution may be the most effective solution. Luckily, there are plenty of options out there.

The Polyend Poly2, as mentioned above, features a USB host, allowing it to function with standalone USB MIDI controllers without the need for a dedicated host, such as a computer. It includes four pairs of pitch and gate outputs, plus 12 assignable CV outputs. An OLED gives direct visual feedback on the parameters and presets. One thing that sets it apart from many other MIDI to CV converters is the support for MPE using the many assignable CV outputs. The assignable output voltage standard and gate polarity allows it to work with just about any synthesizer that accepts CV and gate. Using the SmartThru option, multiple Poly2s can work in tandem to expand the total number of outputs. Shuttle Control and Expert Sleepers FH-2, mentioned elsewhere in this article, provide similar flexibility to the Poly2. These modules have completely arbitrarily-definable outputs (each using a browser-based editor)—so all of their outputs could be used for translating any combination of pitch CV, gate, CCs, and much more. They each even include the potential to generate LFOs, random voltages, and more—and even to control these voltages' properties via incoming MIDI messages. Shuttle Control has a fixed 16 outputs, but FH-2 can be expanded using the FHX-8CV and FHX-8GT expanders for adding extra CV and gate outputs, fact, it can be expanded to produce up to 64 simultaneous gate or CV outputs.

Taking another approach, Hexinverter Electronique's Mutant Brain acts as the central nervous system of a modular synthesizer, particularly focused on rhythmic sequencing thanks to its 12 assignable gate/trigger outputs. In addition to the gate outputs, it features four CV outputs. Hexinverter noted the proliferation of inexpensive MIDI controllers allows for sequencing possibilities that would require massive real estate if ported to a Eurorack module. By separating the sequencer from the rack, users have the flexibility to use whatever MIDI source works best for them. A web-based companion tool allows for the configuration of all 16 outputs. The four CV outputs can correspond to MIDI notes, pitch bend, velocity, aftertouch, fixed voltage levels, or MIDI CC messages. The 12 gate outputs feature a range of behavior from gates, retriggers, transport messages (start, stop, run, reset), or divisions of the MIDI clock. Hexinverter's line of Mutant Drums took the topology of classic drum sounds and reinvented them to create altogether new drums. The Mutant Brain ties all of them together, providing extensive sequencing options.

Bastl's 1983 is another MIDI to CV interface that packs many features up its sleeves. One unique feature of the four-channel module is a dedicated "listen" input for each set of CV and gates. The four channels can function in a number of ways, from four-voice polyphony to eight channels of CC messages. These 'listen' to simple waveforms (triangle, sawtooth, pulse, or sine), and based on the source oscillator's frequency, can adjust the MIDI tracking to keep everything in tune over its seven-octave range. It also supports microtuning, allowing users to nudge the semitones of a 12TET scale up or down minutely. Not only does it have CV outputs, but four CV inputs also allow for control over MIDI note allocation, note transposition, triggered note updates, and portamento to add slide between notes. In quantizer mode, these CV inputs act as the source inputs for the quantizer.

Integrating your MIDI and modular gear has never been easier. The plethora of converters that marry these two formats bring out the best parts of both. It's the marriage of (nearly) universal communication and free-form signal routing that make MIDI to CV converters a useful way to unify your studio.