MIDI began with competitors agreeing to let their instruments talk

One keyboard can tell another keyboard what to play. It is such an ordinary expectation now that the achievement is easy to miss.

The keys under your hands need not belong to the device making the sound. A performance can be recorded as instructions and played back through another instrument. A small controller can become a piano, a synthesiser or a drum kit without changing its physical keyboard. Musicians build entire working methods around those possibilities.

None of that was inevitable. Manufacturers had their own instruments to sell, their own technical choices and their own reasons to keep customers close. Agreeing on a way for competing products to communicate required them to recognise that a useful musical system could be larger than any one catalogue.

That is the part of MIDI's history I find most interesting. The technology mattered, but so did the decision about what the technology should permit.

A common problem across different instruments

Electronic instruments could be connected before MIDI, but connection did not mean general agreement. Different designs and proprietary systems made it difficult to assume that one manufacturer's equipment would work as expected with another's.

The MIDI Association's account of the years 1981 to 1983 describes discussions involving Dave Smith, Ikutaro Kakehashi, Tom Oberheim and engineers from several manufacturers. It presents a collaboration across companies, with the details developed through meetings and technical work rather than a single moment of inspiration.

That distinction matters. Familiar histories often compress a standard into a photograph of two famous people. The photograph is easier to remember than the negotiations, implementation and correction that make an agreement usable.

By the January 1983 NAMM demonstration, a Sequential Circuits Prophet-600 and a Roland Jupiter-6 could communicate through the developing standard. A competitor's instrument had become part of the demonstration of what your own instrument could do.

Instructions, rather than recorded sound

MIDI communicates musical information. The simplest familiar example is a note being started and stopped. That is different from sending the audio produced by a microphone or instrument.

The Association's overview of MIDI makes that distinction explicitly. Because the information describes a performance rather than storing its sound, the receiving instrument can produce its own result from the instructions.

A practical example helps. Record a short keyboard part as audio and you have the sound of that instrument in that performance. Record the played notes as MIDI and you can send those notes to a different sound source. The rhythm and pitches may be the same while the result changes from an electric piano to a string texture.

Neither is inherently better. They preserve different things. Audio captures a particular sound. MIDI preserves useful performance information that can be interpreted elsewhere. A musician who understands the distinction knows when to keep both.

Separating the performance from the sound changed the workflow

Imagine writing a part that works musically but feels wrong in the arrangement. With performance information available separately, you can investigate another sound without replaying every note immediately.

You can also discover that the new sound reveals a problem in the part. A line that works on a short, percussive patch may become crowded on a sustained one. Changing the receiver does not make arrangement decisions disappear. It makes those decisions easier to hear and revise.

The same goes for timing. Moving notes around on a screen offers control, but it also creates the temptation to tidy away the thing that made the performance feel alive. A grid is useful as a reference. It is less useful as a compulsory destination for every event.

This is one of the more productive lessons in a teaching room: record a short part, make one change and listen to the consequence. The learner can hear the difference between a correction and a loss of character. The technology allows the comparison; the musician decides which version serves the piece.

Compatibility is a creative feature

Equipment discussions often treat compatibility as the dull bit that appears near the bottom of a specification sheet. For a musician, it can determine which ideas are practical.

A controller that works with several instruments lets the player choose a physical interface they like. A sound source that accepts a familiar control method can join a setup without demanding a completely new performance habit. A shared language creates more combinations than any manufacturer could supply alone.

That does not mean every device supports every message or behaves identically. Communication needs specific agreement at both ends. The presence of a familiar socket or a MIDI label is a reason to investigate capability, not proof that an intended setup will work without configuration.

Before buying, describe the actual job. Do you need to play notes, change sounds, synchronise timing or control a particular parameter? Which device will send the information, which will receive it, and how will they connect? These are musical questions expressed through practical details.

A shared standard still needs individual implementation

A standard is not an instrument. It tells makers how to communicate about certain things, while leaving room for different sounds, interfaces and musical purposes.

That room is part of its value. A drum controller and a keyboard do not have to become the same object in order to work within a common system. They need a useful relationship between what the player does and what the receiver understands.

It also explains why a feature can exist in a specification without appearing in every product. Manufacturers make choices about what they implement, and musicians need to read the relevant product documentation. General promises about the standard cannot replace the manual for the two devices on the desk.

For an existing setup, a simple test is often worth more than a long compatibility discussion. Send one note. Confirm the intended sound responds. Check note release. Then test the specific controls required for the performance. Add complexity after the basic path works, so that any failure has a smaller number of possible causes.

The work was recognised, but the reward was already everywhere

Dave Smith and Ikutaro Kakehashi received a Technical Grammy in 2013. The MIDI Association's account of the award describes the significance of cooperation between manufacturers and the widespread use that followed.

Awards are useful markers, but the stronger evidence is less ceremonial. A musician expects to connect equipment without replacing every component with something from the same brand. A student expects a keyboard performance to become material they can edit and arrange. A composer expects an interface to control more than one possible sound.

Those expectations are the accumulated result of a standard becoming ordinary. Successful infrastructure often disappears from the conversation because people spend their attention on what it enables.

The history should therefore leave room for everyone who implemented, maintained and extended the agreement. A celebrated demonstration can open the door. Keeping that door useful over decades takes a different kind of work, repeated across many organisations.

MIDI 2.0 extends the conversation

MIDI did not stop developing with the original specification. The Association describes MIDI 2.0 as an extension of MIDI 1.0, with compatibility treated as a central concern. Its architecture includes capability inquiry, allowing devices with suitable two-way communication to agree on extended abilities.

The newer protocol provides increased data resolution for channel-voice messages and expanded control of individual notes. Profiles and property exchange address other aspects of how devices establish and describe their behaviour.

For the player, the sensible response is interest rather than automatic replacement. Ask which capability your work would benefit from and whether the entire path supports it. A new controller cannot create an expressive feature in a receiver that does not implement that feature.

Existing equipment does not become musically useless because another version of a standard exists. A part played well on a working instrument remains a good part. Upgrade when the additional control solves a real problem or opens a possibility you intend to use, rather than because a version number has become a status marker.

Keep a performance that can survive the setup changing

There is a practical lesson here for anyone making music with software instruments. A session can contain the notes and still fail to reproduce the sound you remember if the original instrument, patch or processing is unavailable.

Keep clear track names, identify the sound sources and save the information needed to reopen the work. Where a sound is important, an audio render can preserve the result alongside the editable performance. Follow the particular software's documentation for exporting and archiving.

This is ordinary musical housekeeping. It protects the afternoon in which you found the right sound from the afternoon in which a computer changes and you discover that the part was more dependent on its environment than you realised.

The same attitude belongs in a band's gear records: know what you own, what you use and what you need to reconstruct the working setup. A collection of devices is easier to maintain when its relationships are understood.

The larger musical lesson

A standard succeeds when it allows different people to make different things that still have something useful to say to one another. That is a fairly good description of an ensemble as well.

The point of agreement is not to erase individuality. It is to establish enough common ground for individuality to become useful. Players need a shared understanding of time, form and cues precisely so that they can bring different sounds and decisions to the performance.

MIDI's achievement was to make that kind of agreement available to equipment from competing makers. The enduring result is permission to combine, revise and keep using things. Decades after the first demonstration, that remains a much more interesting idea than a cable.