How does MIDI synchronization work

Workshop: synchronizing devices with MIDI clock

MC: The helper for syncing delays, LFOs, drum computers and much more!

Simple, fast, inexpensive and safe - Sync for everyone

© Shutterstock, Turan Ramazanli

MIDI beat clock? Master? Slave? Synchronization? - In this workshop you can learn how to synchronize different devices and software with each other via MIDI clock. Sometimes it is necessary, for example, to run a sequencer / DAW such as Logic, Cubase or Ableton Live together with a drum computer or hardware sequencer; But there are also enormous advantages if external devices such as effects devices or synthesizers are informed of the current song tempo via MIDI clock.

Guitarists For example, a delay synced via MIDI clock can open up new possibilities - and is also more precise than “tapping” with the foot. Drummer can sync your e-drum module or your click to the tempo of the recording system. And if you want to build and edit sounds, you will be pleased about the possibility of being able to couple LFOs for filter movements to the song tempo.

Nice to know: none of this is rocket science and usually done quickly. I limit myself to the much simpler MIDI clock connection (called MC "or" MBC "), which in most cases also makes more sense than the MIDI timecode synchronization via" MTC qfm ".

Brief information on MIDI clock

What is "MIDI Clock"?

  • MIDI Clock is a very simple system with which a "slave" device (or software) is informed of the current song tempo by another device or software ("MIDI Clock Master") via a normal MIDI line.

What can you do with the MIDI clock? What is MIDI clock used for?

  • All parameters in the slave can be controlled in musical beats instead of in time units such as seconds or milliseconds.
  • Playback systems can be connected to one another, drum computers, sequencers etc. can follow the MC master.

Is it difficult to sync with MIDI clock?

  • No, it's pretty easy. The prerequisite for MIDI clock synchronization is that all devices / software provide the master / slave functions.

Is the MIDI cock synchronization safe and stable?

  • In principle, MC-Sync is stable and has little delay.

Are there other terms for MIDI clock?

  • MIDI Clock or MIDI Beat Clock are often abbreviated to MC or MBC.

Is MIDI clock the same as MIDI timecode?

  • No: MIDI Timecode ("MTC") works very differently than MIDI Clock and is usually used differently.
  • The importance of MIDI timecode has also decreased significantly in recent years.

Master and slave: synchronize via MIDI clock

Synchronization via MIDI clock always works with a “boss” device, which tells the other person where to go. This is called the “master”. So if the master claims, for example, “This is a ninety-sixth note.”, Then the connected slaves simply have to believe this and have to obey. Nix democracy: There is not even a great provision for the slaves to be able to report anything back to the master that would be relevant for the synchronization. So it would be clear what we have to look for first: A master needs a MIDI out, a slave a MIDI in. Of course, MIDI-over-USB and even internal computer connections also work in principle (although we have a more extensive system there with ReWire).

Typical setup: The sequencer is the master and supplies its slave with tempo information

MIDI clock: can my device be master or slave?

Unfortunately, MIDI devices (and software!) That could conceptually be able to act as slaves or masters can spoil the soup for you, but still do not offer it. The manual will help you find out. Since manufacturers like to peddle the capabilities of a device, but keep silent about an inability, it is often advisable to take a look at the table of contents or, even better, the register. No entries for “Synchronization”, Sync ”, MIDI Clock”, “MC”, “MIDI Beat Clock” and the like? Then it looks rather bad. The quickest way to be certain is to look at the “MIDI Implementation Chart”. This standardized table is included with almost all MIDI devices (usual place: appendix to the manual!) And provides information about the capabilities and non-capabilities of the device. If you want to know, for example, whether your device can send MIDI Clock, you have to look in the first column (“Function”) in the line “System Realtime - Clock” whether in the column “Transmit” (sometimes just “Tx” ) a small “x” is entered or an “o”. The latter would mean “is implemented”, which in turn means that the device can output the MIDI clock. With “Receive” you can see whether a device is able to receive and convert the clock. If there are any restrictions, it is often already here (for example “LFO2 only”). For some more far-reaching synchronizations you need the “Commands” and the “Song Position Pointer”, as we will see later. If a lot of “x” threatens to ruin the synchronization calculation: Don't give up, a firmware update may help!

MIDI cable in which socket? - Set up a simple MC connection

Apart from the two devices to be connected, you need a MIDI cable, nothing more. This is plugged into the MIDI Out socket of the master and into the MIDI In socket of the slave. If several devices are to run as slaves, there are two options: The simplest is to connect additional devices via MIDI thru. As a rule, problems arise only with a large number and long cables (especially in the rack!). A daisy chain of three or four devices with short cables is usually completely functional. If a device provides several outputs, MIDI clock can of course be sent via these. Usually these are multiport interfaces that are connected to computers.

MIDI Clock: Master Settings

The only thing that needs to be set on the master is that the clock should be sent. Typically this setting can be found in one of the global menus, often under “System - MIDI”, “Utility - MIDI Settings” or the like. Unfortunately, this is not standardized. If in doubt, the manual is useful to find it again. Once you have found the parameter on your device, everything is very simple: “MIDI Clock: OFF” can be changed to “MIDI Clock: ON”. Sometimes it is called a little differently, but most of the time the function is self-explanatory. The setting is also quite easy to find with sequencer programs / DAWs such as Logic, Cubase, Live, Reaper, StudioOne and the like: It is usually a setting that is saved with the document ("Session", "Song" etc.) , not a global program setting.

To be found in the document settings: MC-Settings at Logic. (Click into the picture to enlarge)

Perhaps you are surprised that sometimes a “delay” can be set, which has a negative (!) Value and allows the relevant data to be sent ahead of time. It is not necessary to compensate for the transmission time (this takes less than 300 microseconds per command!), Rather it can react to somewhat sluggish recipients. Most of the time you will not need this function.

Slave settings

First of all, it is important to distinguish the purpose for which a device should receive the MIDI clock. For example, if it is a drum computer, an external workstation or any other system in which a sequencer should run as a slave to the clock master, everything is usually quite simple: At one point in the menu, for example, “Transport” or “Sync Settings” means (sometimes at the same point at which a device sends the MIDI clock instead of receiving it) the control of the playback can be changed. Usually you have to “Sync” from “Int.” (ie “internal”) to “Ext.” or just toggle “MIDI”. Especially great: That's about it! As soon as the master is running, the slave should already be running. In reality, most masters send additional commands that are necessary for synchronization (especially “SPP”, the “Song Position Pointer”, which tells the slave about the startplace informed) without having to activate this separately. You will gradually learn here what exactly is going on. Most of the time, you can only find out about the commands used by looking at the MIDI implementation table. If something does not work again expecting it, it may be due to a lack of support for this type of command, but usually the hare is elsewhere:

The right settings are of no use if the filters are switched on

MIDI filters can be set in the receiver that block synchronization commands globally. A frequent source of error is the control software of a MIDI interface or audio interfaces with MIDI functionality.

Some DAWs are too good

There are always some reasons ... Technically it is not a problem to run large sequencer / DAW ships as MIDI clock slaves, but some manufacturers seem to be too sensitive to let their valuable programs become low followers. No, seriously: In terms of timing stability, digital audio can be a bit rougher in slave mode than in master mode, but there are examples of how this works properly in spite of everything. If a program nevertheless allows it to be subordinate to an external MIDI sequencer or comparable system, the settings do not differ much from those required for hardware.

Ableton Live does it in an exemplary manner! Clock slave settings simply by activating "Sync" on the MIDI input via which the MIDI clock master sends the commands - done!

Sync parameters in the slave

MIDI clock is very popular for adapting delay times to song tempos. Delays can actually be set in one way or another on all devices, so why synchronization?

Let's think about it: at a speed of 120 bpm, you can quickly calculate which delay time you have to set for a rhythmic delay of a quarter note duration: at 120 bpm this is 0.5 seconds.

The general rule is: 60 / bpm = quarter note spacing in seconds - provided that it is a 4/4 time signature (or a comparable one).

It's all doable, but how long is a dotted eighth at 141 bpm? It usually takes a little time to get to 0.319 seconds. Tapping with the hand or foot works, but it is often shockingly inaccurate - and therefore only recommended if you are not playing at the click. And 141 bpm doesn't have to be exactly the same speed on one device as on another! You see: the synchronization to a sequencer is extremely useful.

To synchronize effects, it is sometimes sufficient to send the effects device the MIDI clock - often enough the delay display then jumps from milliseconds to musical time units such as 1/4, 1 / 8T (an eighth triplet) or 1/16. (a dotted sixteenth note) around. Sometimes you have to activate the “clock-bondage” separately or the parameter has a double scale: In this case, times are set in milliseconds on one side, and musical divisions are available on the other. Often you have to set a “global tempo” for the device elsewhere, in case you work unsynced. For troubleshooting, the following also applies here: Perhaps global filters are set (e.g. “Clock Receive: OFF” or “MIDI Clock: Don't Listen”).

Behind the scenes: How the MIDI clock works

It couldn't be simpler: via MIDI clock not the song tempo in a format like “126 bpm”, because for this all connected devices would have to be able to refer to the same time reference - and what one device lasts for one second is perhaps only 0.9978 seconds for another. Such clocks also like to change with the operating temperature.

With the MIDI clock, the "Metronome principle”For use: The information that is important for the musician playing the click is not the numerical display of the speed that was set for the metronome, but only the distance between the two impulses. The metronome makes a “click” and from the next “click” everyone can play along, regardless of whether they know or recognize the speed indeed is. When three musicians play to the click, one of them might think it's 127 bpm, another 130 bpm, another 125 bpm, while the metronome is actually set to 126.5 bpm - but who cares if it works? MIDI Clock does the same thing: It is a simple command that very banally consists of a fixed sequence of binary numbers (this one: 11111000) and is never varied - i.e. it cannot transport different information. Just like a metronome sound or a clap of hands, such a command only means “Click” or “Now”!

However, the distances are significantly higher than with conventional metronomes, which are usually set in quarter notes, less often half or eighth notes. The resolution of the MIDI clock is “24 PPQN”, which stands for “24 pulses per quarter note”. So 24 of these commands are sent between two quarter notes, which corresponds to a grid of 24 x 4 = 1/96. Conversely, that also means that speedchanges can be implemented at the master within this time.

The MIDI clock works that easy! Every little dot is a MIDI clock command. And all of these commands are as identical as clapping your hands or a "click" sound from the metronome!

You can imagine that fluctuations when sending MIDI clock can cause problems just as much as a sluggish metronome. Fortunately, the commands belonging to the common messages are assigned one of the highest priorities - and with a total of ten bits in length, they are extremely short! You don't need to be afraid that an activated clock will really have a negative effect on the further performance of the MIDI transmission: At 120 bpm, only 1.5% of the transmission capacity is occupied by the clock.

The advantages are clear: If I have set an eighth note as the delay time in my delay, then remains the delay is an eighth note, even if the song tempo changes changes. After all, I didn't enter anything in milliseconds, but in musical time values. It is very good for many songs if, for example, they are allowed to record a little more speed in the chorus and a fermata can be set in the stop - without the nice editing options in the DAW making it directly impossible.

Of course, it should be noted that a slave does not know anything about time signature and their resolution due to their principle can. So whether it is a 4/4, a 6/8, a 3/2 or a 5/8 is just as unknown to the recipient as it is to him what a “1” is and what is not.

The principle remains the same as with the effects device: The slave receives the MIDI clock from the master.

Applications for tempo-synchronized parameters on synthesizers and effects devices

Delays are certainly the most frequently used targets for MIDI clock sync, but other parameters also make sense if they can be appended to the song tempo. Just think of auto pans that wander rhythmically in the stereo image or of tremolo or flanger movements. These and some others have in common that the modulation source is a low-frequency oscillator - an LFO. If this is linked to the song tempo, you can create effects that don't somehow drift along indifferently, but can take on a rhythmic task in the mix. If, for example, you don't get overly static surface sounds in the mix or if they “stick in time”, a little rhythmic gymnastics is good: let them wander in the panorama, get louder and quieter and the like. Tip: You can make parts, sometimes even entire songs, interesting by moving the modulation movements in opposite directions, for example a ternary tremolo in a straight 4/4 song.

What you can do with the LFOs in multi-effects devices can of course also be used in synthesizers or samplers. The classic modulation target for a large number of sources is certainly the filter cutoff frequency, especially the cutoff of a low-pass filter. With an LFO in sync, for example, you can achieve that the spectrum of the sound "toggles" in eighth movements without much tinkering.To do this, a low frequency oscillator with the waveform “rectangle” (also: “pulse”) must be set to a quarter note and act on the cutoff frequency. The quarter note is not a typo: the filter is then “open” half an oscillation cycle, half a “closed”, ie the movement is in eighth notes.

With the Moog Voyager, you simply set the LFO to "MIDI" in order to couple it to the clock.

In the simplest case, the LF oscillator only switches between two extreme values ​​(it is a rectangular shape), but it can also have an ascending or descending sawtooth shape, triangle, sine or more complex shapes (although not all of them are always so easy to sync, especially with Analog synthesizers). The filter function just described can, for example, work with a rising sawtooth oscillator much more precisely than an envelope curve - and re-triggering can be dispensed with.

And of course you can really let off steam here - filter resonances, pulse widths, pitch of individual oscillators, speeds of other LFOs, noise levels, crossmoulation amount, with samplers sample start, loop lengths and much more can serve as modulation targets. "The sky is the limit."

Even an arpeggiator doesn't have to run freely ...

By the way, nobody says that the song tempo only ever has to come from a sequencer! A combination is also conceivable in which a synthesizer is connected to a multi-effects device in a very trivial way. There, for example, the effects device can be the master and the synth can be informed of the tempo settings, so that delays and flangers from the multi-FX and LFO movements in the synthesizer take place at the same tempo. If you consider that it may even be possible to tap the global tempo in the effects device with your foot, this opens up completely new possibilities for live operation or when playing on vinyl decks ...

Two effects devices can also be synced.

Running two or more sequencers together - how does it actually work?

There is a second construction site where the MIDI clock can be used, namely when synchronizing two playback systems, i.e. sequencers and DAWs. A widespread example is the drum computer or the small hardware production workstation with sound generation that should run along with the computer system. Sometimes two computers on which different sound generators are installed (e.g. Mac with Logic and Windows PC with programs not available for Mac) should be used simultaneously. Sure: One of these systems is MC-Master, one or the other run along well. But if you look again at how the MIDI clock works, you are faced with a problem if you go beyond pure pattern playback: After all, it must be possible to run both systems together from bar 25 because you want to edit the transition to the chorus. How does it work?

 

It is pleasant that this is usually done automatically - and it just works. What is behind it: Here the clock receives support, namely through the song position pointer, which tells the slave where the journey is starting from. The accuracy is “only” a sixteenth note. That is not a problem, because the synchronization resolution is still 24 PPQN. It only means that you cannot start from values ​​finer than 1/16 - the slave then only joins with the next sixteenth, which should be tolerable. Incidentally, this is one of the main reasons why the standard grid for sequencer programs is set to 1/16. Furthermore, the SPP does not have to be sent every sixteenth, but only when the master starts and changes its position. Finally, the slave can also count: After six incoming MIDI clock messages, a sixteenth is over (24: 2: 2 = 6). However, the range of SPP values ​​is limited. For technical reasons, “only” 16384 different sixteenth-note positions can be displayed. With a four-quarter time, however, you are still at 1048 beats, which at 120 bpm still last 35 minutes. Should you still have to carry out mammoth projects in sync (such as complete live shows), you should know that you can either no longer start after the last sixteenth - or simply halve or even quarter the tempo nominally (however, this also reduces the clock resolution !). Most systems also work with the very simple commands Start, Stop and Continue (which should not be confused with MIDI Machine Control - that is something else!).

Classically, a DAW sends MIDI clock, song position pointers and commands to a drummy or a keyboard workstation with its own sequencer

MIDI clock: synchronization simple and practical

As you can see, synchronization with MIDI clock is very helpful and, in principle, very simple: the master sends, the slave can "add parameters" to the clock in various ways - or, in the case of sequencers, they can run as slaves. Have fun trying!