MIDI For The Hardware Studio

MIDI is the universal language of the studio — and if you use hardware instruments, you need to understand it!

MIDI 2.0 will undoubtedly be a revolution in seamless and transparent musical connectivity. But until those mystical protocols are negotiating on our behalf, good old‑fashioned MIDI 1.0 is still the method by which most electronic musical instruments communicate. With many people emerging from software studios and encountering hardware setups for the first time, it might be good to brush up on your MIDI skills and make sure you’ve got the basics right.

MIDI In Software?

MIDI reigned supreme for many years, supplanting analogue control voltages with the wit of velocity, the charm of patch changes and the allure of transferable data. But then it seemed to disappear into the computer. As the software studio blossomed, synths and samplers were pulled in and assimilated into the computer, and USB came along to take care of any peripherals. MIDI cables and connections were lost, and those networks of control and data exchange faded from memory. It’s easy to forget that whole generations of composers and producers may never have plugged in a 5‑pin MIDI cable.

DAWs still use MIDI as a data format to represent and manipulate musical notes, but once‑important features such as channel numbers and ports are rarely relevant in software, where you tend to load a separate instance of a software synth for every track. Move to hardware, however, and the workflow is far less fluid. You might be surprised at how often things don’t work quite as you thought.

So, in this guide, we’re going to tackle the fundamentals of MIDI 1.0 from the perspective of moving from DAW to DAWless, or somewhere in between.

What Is MIDI?

MIDI stands for Musical Instrument Digital Interface, and it is a digital protocol that allows you to control one MIDI device from another MIDI device. In practical terms, it means you can play one synthesizer from the keys of another. Before MIDI, you would typically have had a prog‑rock‑sized stand full of individual synthesizers that you would have had to play from their own keyboards, one or two at a time. MIDI allowed you to play multiple synths from a single keyboard, and this led manufacturers to offer synths without the now‑superfluous keyboards. A ’90s studio would thus contain a rack of synth modules, all played from one keyboard. The keyboard itself didn’t need to be anything more than a row of keys that output MIDI, and so, the MIDI controller keyboard was born. It wasn’t long before those hardware synths became software synths, and your entire synthesizer rig was reduced to a laptop and a USB‑connected MIDI keyboard.

Before MIDI Learn, there were MIDI implementation charts! This table shows how the Korg Volca Drum responds to various MIDI messages. The Control Change section in the middle lists the MIDI CC numbers and the parameters they correspond to. If you wanted to control the pitch of oscillator 1 from an external MIDI controller, for example, you’d set one of its encoders to output MIDI CC 26.MIDI note data tell the receiving device what notes to play, and when they should start and stop, but that’s not all. MIDI can also transfer information about velocity, pitch bend, modulation, sustain and all sorts of parameter values as well as MIDI Clock, Start/Stop commands and tempo. Most importantly, both notes and parameter information can be recorded, edited, stored and played back by a MIDI sequencer or DAW.

There are 128 possible ‘continuous controller’ or ‘CC’ numbers that can travel between devices, and most of these can be mapped to any parameter, knob or control on the destination device. So, when you click on MIDI Learn in a software synth and map knobs from your USB MIDI controller, you are assigning MIDI CC numbers from your controller to the parameters in the synth.

As controller mapping in software is really easy, you may be surprised to find that it can be difficult and fiddly in hardware. It usually involves digging out MIDI implementation charts that are hidden away at the back of manuals and navigating tiny LCD screens and annoying menu systems to set CC numbers manually. But the principle of controlling a synth from a MIDI controller is the same. On the other hand, part of the joy of hardware synthesizers is the uniqueness of their own front panels and workflows. There’s no reason why you have to map controls from a master controller to a synth; you can just use the synth’s own knobs, assuming they are within reach. MIDI gives you lots of possibilities, but you don’t have to take up all of them.

A single stream of MIDI data can contain 16 channels. You may be aware of MIDI channel numbers in your DAW but unless you are using a multi‑channel soft synth like HALion, SampleTank or Kontakt, you’ve probably just ignored them. In hardware terms, those 16 channels mean that one controller can address 16 different devices, down a single cable from a single keyboard. This is where the joy of cabling comes in, which is often the first area where MIDI novices come unstuck.

A single MIDI keyboard can control up to 16 other MIDI instruments.

MIDI Connections

In the heyday of MIDI there was an order to things: a proper way to do connections that we all understood. These days there’s nothing but chaos on the back of synthesizers, which makes talking about it difficult. But in the good old days there were three types of MIDI port: MIDI In, MIDI Out and MIDI Thru. The ports repurposed a popular audio connector at the time of invention called DIN, which I’ve just discovered stands for the Deutsches Institut für Normung. It’s a round and chunky connector with a fan of five pins, although MIDI only actually uses two of these. The ports placed on the back of MIDI devices are the sockets, and the cables carry the pins. The arrangement of the pins meant that the cable could only go in one orientation, which would be the cause of much frustration when trying to reach behind a synth to plug one in.

Unlike USB, traditional 5‑pin MIDI connections are unidirectional — the data only travels one way. If you connect the MIDI Out from a sequencer to the MIDI In of a synth, you’ll be able to control the synth from the sequencer. But if you want the synth to send data back, you need to connect its MIDI Out to the sequencer’s MIDI In.MIDI only travels in one direction along a cable, from the controller (the keyboard or knob controller) to the device (synth or sound module) you’re wanting to control. There’s no handshake or exchange of data like we hope to have in MIDI 2.0, and nothing is automatic or intelligent in the way that things are in DAWs. A MIDI device fires its data out completely blind and instantly, in response to the keys you press and the knobs you turn. To make it work, you connect the MIDI Out port on the controller to the MIDI In port on the synth, and ensure controller and synth are set to transmit and receive respectively on the same MIDI channel. Make the wrong connection, or select the wrong channel, and you won’t hear the synth, because the MIDI data won’t be making it to the other machine. There’s no error message, warning or pop‑up, you just get silence.

It’s important to note here that the sound comes out of the synth you are wanting to play, and so its audio output needs to be routed somewhere where you can hear it. It’s easy to assume that, because the sound from all your software synths comes out of the same computer, the hardware experience ought to be similar, but this isn’t so. If I play a note on a keyboard over here, I should expect the sound to come out of the synth that’s over there. As your collection of synths grows, you may need a hardware mixer to enable you to hear all of them.

Local On/Off

Things gets muddied further if the keyboard you’re using as a controllier also has its own sound engine. In this case, you’ll need to consult the manual to find out how to turn ‘local’ mode on and off. In MIDI parlance, ‘local’ refers to the playing of the sound engine by the keyboard that forms part of the same instrument. You play the keyboard on your synthesizer, and you hear the sounds from that synthesizer — that’s local control. If you are using one synthesizer to play the sounds of another via MIDI, the synth you are playing will sound at the same time as the synth you are controlling. You may not want that. You may also encounter a problem if you are sequencing a synthesizer with a DAW and using it as a MIDI controller. Normally, when using a ‘dumb’ MIDI controller with a DAW, the MIDI data goes in to a track and is routed to your software synth. But we’re wanting to use a hardware synth, so we would set the MIDI output on the track to route back out of the computer and to the synth. As you play the dumb MIDI controller, you will hear the sound of the hardware synth coming from itself — great! However, if you replace the controller with a synthesizer, you will hear the sound of the synth twice: once as you play the keys, and again when it is retriggered by the MIDI going through the DAW and back out to the synth. You really don’t want that.

Both situations can be prevented by turning off local control. Somewhere deep within the synthesizer’s menu you’ll find an option to turn Local to Off. This disconnects the keyboard from its internal synth so that you don’t hear it while controlling another synth, and you don’t get the retriggering when sequencing. Beware, though: forgetting that you’ve turned off local control on a synthesizer can be a remarkable source of frustration when you come back to play it on its own later.

If the keyboard you want to use as your main controller is capable of generating its own sounds, you’ll want to turn ‘local’ off. This breaks the internal connection between the keyboard and the sound generator, so that when you’re playing another instrument, you won’t hear the keyboard as well. To play the keyboard’s own sounds in local off mode, you’ll need to send its MIDI data back to the MIDI In, either via one of the synths you’re controlling or from your DAW or sequencer.

Multitimbrality

Having 16 MIDI channels enables an individual synth to operate more than one sound engine at the same time. This became practical with the advent of digital synthesizers, and is not a feature you often find in an analogue synth. It’s most prevalent in big workstations like the Yamaha Montage or Korg Kronos, as well as home keyboards. I mentioned HALion, SampleTank and Kontakt earlier as multi‑channel synths, and they support different sounds or presets loaded and addressed on different MIDI channels. In hardware this is called multitimbrality.

On a multitimbral workstation synth, you can have different sounds loaded and sequenced all running at the same time. You can be playing a piano sound on the keys while the bass, drums and strings are all taken care of by the sequencer. Everything comes out the back of the synth, and it is a very DAW‑like experience. It also allows for cool performance features like sound layering and keyboard splits. You can do the whole lot on a workstation keyboard without having to look at a computer screen.

Moving to a big workstation synth is actually a really easy transition for a computer‑based musician. However, for many of us, the move towards being DAWless means ending up with a bunch of little synths rather than one big one. Very often these are single‑engine synths that only make one sound at a time. How would you connect those to your MIDI keyboard or sequencer? That’s where MIDI Thru comes in.

MIDI Thru

Your 16 MIDI channels give you the option to play 16 different sounds, all from the one cable. If the destination synth is 16‑part multitimbral then your job is done. If not then you are going to have to chain up your mono‑timbral synths through the cunning use of MIDI Thru. In the days before the chaos (which I’ll deal with in a minute), MIDI synths would all have the three ports on the back: In, Out and Thru. MIDI Thru is a repeater for whatever comes into the MIDI In, allowing you to pass the MIDI data from the keyboard controller or sequencer down a chain of synthesizers that are connected from MIDI Thru to MIDI Tn.

As long as each synth is set to receive on a different MIDI channel, you can address each one independently from your MIDI controller keyboard, DAW, hardware sequencer or whatever you’re using to play them. MIDI Thru is excellent.

In the heyday of MIDI there was an order to things, a proper way to do connections that we all understood. These days there’s nothing but chaos on the back of synthesizers...

The Chaos

Unfortunately, many little synths these days don’t seem to have MIDI Thru. Some don’t appear to have MIDI ports at all, and when they do, they are often on jack cables rather than DIN sockets, and come in two types and at least two sizes. After unpacking your fresh new quirky analogue or hip hybrid desktop synth, you may have no idea if it’s going to fit into your MIDI setup without a serious array of adaptors and a lot of trial and error.

There was a move a couple of years ago to migrate the dated DIN format to the similarly old but more common minijack format. It felt initially like an important and inspirational event that saved manufacturers a load of money on ports and almost ushered in a new era of easy MIDI patching. Unfortunately, a standard wasn’t established and so we ended up with two identical‑looking ports that were wired differently and so were incompatible with each other. They are known as Type‑A and Type‑B, and because compatibility is not assured you invariably end up with a bunch of minijack‑to‑DIN adaptors that never seem to be the right type for the synth you’re trying to cable up. Apparently, a standard type has now been established but no one is quite sure which it is.

Just to toss a further spanner into the works, some synths and many MIDI controllers only handle MIDI over a USB connection. This is probably a throwback to when we thought the computer was always going to stay central to the music making experience.

Anyway, what do you do if you have no MIDI Thru? Well, some synths are able to pass incoming MIDI to their MIDI Out port, if they have one, but many do not. Otherwise, you’ll need a separate MIDI Thru box. This is a very useful chunk of hardware that will mirror a single MIDI In into many MIDI Thru ports for all of your synths to enjoy.

Plugging It All In

If your computer is sticking around as the sequencer for your MIDI synths, you’re going to need a MIDI interface. Often, you’ll find that your audio interface has MIDI ports built in. If not, then you’ll need to add one. At one time you could get rackmounted MIDI interfaces with eight or 16 MIDI ports, but they are a rarity these days. However, you can still get little boxes with a couple of ports to attach to your computer to do the job.

If you are using a hardware sequencer like the Arturia KeyStep Pro or Novation SL Mk3, you might find they have a couple of MIDI Out ports to help make the chaining a little less unweildy. Otherwise, it’s an exercise in cabling and organisation. The position a synth takes in the MIDI Thru chain really doesn’t matter, provided that whatever is doing the control is at the top. All that’s important is that you set them all to different MIDI channels and remember what they are.

So far, we’ve been thinking about this in terms of having a controller or sequencer which is sending MIDI to a bunch of other synths. What happens if you want to send some information back, for example, to record the movement of the filter knob on one of your synths into the sequencer? In a DAW the automation of parameters is dead easy. You can draw it in with a mouse or capture the turning of knobs on your MIDI controller. Hardware synths can be treated the same way, but to capture the movement of knobs from the synth itself, you’ll need a bit more cabling. MIDI, as I mentioned, only travels in one direction down each cable, so you can’t send the data back up the MIDI Thru chain that you were using to control all your synths. Instead, you’ll have to plug the MIDI Out port of your synth into the MIDI In port on the sequencer or DAW.

MIDI merge boxes let you combine multiple MIDI connections and send their data down a single MIDI cable.

That’s all very easy with one synth, but what if you want to do the same with two or more synths? This is where those big eight‑in/eight‑out MIDI interfaces start getting useful. Ideally, every synth needs its own MIDI In and MIDI Out connection to your DAW/sequencer. That way it can be addressed independently by the sequencer with up to 16 channels of multitimbrality, and all the movements and performance elements from the hardware can be recorded directly. However, assuming you have a less comprehensive setup, involving something like the Arturia KeyStep Pro or Korg SQ‑64 as a sequencer and a desktop of little synths in a MIDI Thru chain, what do you do?

The easy answer is that you grab the cable connected to the sequencer’s MIDI In port and swap it between the synths whose data you want to capture. But that’s not great if you are trying to record your performance on a few synths at once. The better answer is another useful box called a MIDI merge, which is essentially the opposite of a MIDI Thru. With MIDI Thru you are running a single generator of control data to multiple synths; with MIDI merge you are combining control data from multiple sources into a single stream for recording. Merging MIDI requires that you keep your wits about you in terms of what data is being recorded to which channel, so that it is sent back out to the right synth. This is one of the reasons why we quite like DAWs.

Clocking

One or more of your desktop MIDI devices might be a drum machine or groovebox. Inside the DAW, you don’t have to worry about such things; everything is running in perfect sync to itself and the master tempo. But now you have different boxes with different tempos in different places; what do we do about that?

Well, MIDI contains a clock signal for just such eventualities. The clock sends out 24 pulses per quarter note (ppqn) at a tempo set by the master device. Most synthesizers have their own clocks for running arpeggiation, sync‑based effects and modulations, but this can be overridden by an external MIDI Clock. Some synths will do this automatically once a clock is detected, but for others, you’ll have to dive back into the depths of a menu system to find the right switch.

In many situations you’ll want the drum machine to play along with the sequencer, or have one groovebox setup to follow another. Be aware of your MIDI chain. Make sure that whatever is dictating the tempo is at the top, or is plugged back into the sequencer in the same way as if you were capturing its parameter movements. You may also want to simply use the sounds of a machine without it running its own rhythms. To enable and disable those sorts of things, close examination of the MIDI settings will get you where you need to go. MIDI can send Start and Stop commands, providing transport as well as tempo sync. Another MIDI feature called Song Position Pointer can tell the connected machines how many 16th notes have elapsed since the start of the song. It can make for a comprehensive sequencing environment but is another one of those reasons why we quite like DAWs.

There are other, non‑MIDI ways to synchronise your devices too... We’ll cover the many ways you can keep your devices playing in time a future article.

Summing Up

The important thing to remember is that MIDI 1.0 is not very clever. To use it successfully requires some thought and fiddling on your part, followed by a healthy dose of trial and error. If it’s not working then it’s probably your fault and you’re going to need to go through all your synth boxes, checking cables and adaptors and methodically digging around in the MIDI settings. Check for local on/off, check the MIDI channels and check that things you don’t want to happen are disabled or filtered out.

The controlling and sequencing of multiple synthesizers outside of a computer environment feels very visceral. You’ve been involved in the creation of a network of musical instruments, you’ve cabled them together, set conditions and mapped parameters through a load of different menu systems and interfaces. You feel invested in the outcome in a way that you don’t when installing a plug‑in. MIDI enables all that. Now, how are you going to save where everything is so that you can work on it again next week?