Magazine Archive

Home -> Magazines -> Issues -> Articles in this issue -> View

MIDI Matters (Part 3)

Jay Chapman continues his 'back to basics' description of MIDI with a light-hearted pot-pourri of handy hints and tips.

Having re-read last month's instalment I have come to the conclusion that life was getting a bit too technical, diy and tedious, so this month's opus is a little more general and even on occasions quite light-hearted! We'll pick up on Real-Time messages and so on next time. What follows is in no particular order...

Do you realise that every time you talk about (or I write about) the 'MIDI interface' we are talking (writing) rubbish! If you expand the acronym 'MIDI', in the phrase 'MIDI interface', you get 'Musical Instrument Digital Interface interface' - what's an 'Interface interface' then?!

MIDI is basically concerned with communicating information between various pieces of musical equipment. Some of the information is to do with control and the rest is to do with the creation of sounds. Control information might represent the playing of notes on a keyboard, the movement of a pitch wheel or the pressing of a button on the front panel of a synthesizer. Sound information might include the numbers which define a sample or the 'map' of the current connections inside a synthesizer which correspond to the old idea of a 'patch'.

Generally, it's not a good idea to connect the MIDI Out to the In (or the Thru) on the same piece of equipment (unless you're investigating 'Local Off' when Out to In makes very good sense).

"MIDI is the specification of the interface hardware and message protocol for a simplex digital communications link between items of musical equipment allowing the creation of a complex network with real-time capabilities." Memorise, and amaze your friends at parties by trotting it out with a straight face!

MIDI is all to do with the microprocessors inside your equipment talking to each other. The more common messages, such as note-on or note-off messages, consist of three bytes and take about 1/1000th of a second to get from one device to the next. During this time the microprocessors can process hundreds or even thousands of instructions, which is why some microcomputer-based sequencers can handle fancy graphics displays at the same time as playing your music.

There is still only one major league microcomputer system (as far as I am aware - corrections to SOS please!) that features a built-in MIDI interface - and even then it's non-standard (MIDI Out and Thru are wired into the same DIN socket). If you have to buy an interface for your micro, then make sure it is one recognised by all of the software you want to buy: check in the adverts or phone the suppliers. It's not unknown for an interface manufacturer to keep all the details of his interface to himself so that you are locked into the set of software he supplies!

Midi hi-fi systems have nothing to do with our MIDI - their 'midi' simply refers to their size! MIDI also has nothing to do with the south of France or skirt lengths... Is a MIDIot a person with a MIDI interface and a 6-pin DIN plug? Is a MIDIator someone who handles negotiations between synthesizers that refuse to work? Sorry - brain went out for a walk without me...

All a MIDI sequencer really has to do is suck in the messages arriving on its MIDI In and store each message alongside the time that it arrived. Playback just involves counting out the same lengths of time (ie. between events) and then sending the previously recorded message out via MIDI Out. The amount of money you pay for a sequencer depends on how easy it is to convince the sequencer of your true desires, ie. how easy it is to edit the mess you managed to input into the masterpiece you want it to output!

Don't forget that playing one note (on then off) only generates two MIDI messages but tweaking a continuous controller (pitch-bend wheel, modulation wheel, volume pedal, etc) can get you into the hundreds of messages ballpark. Your sequencer may well not need to suck, store and blow all of these messages for you to still hear the effect you were trying to create. It is rather nice, therefore, if your sequencer software can make some sensible attempt at filtering out some of the superfluous controller messages.

Not all MIDI software is created equal! In fact, MIDI is not too difficult to programme for but the all-important 'user interface' between the computer and the musician... now that is hard to do well. If in doubt, try the software in the shop for half an hour. If you can't easily record and play back in that time, forget it! Find another software package/computer/shop(assistant)!

Who needs 46-track recording studios? You can produce perfect stereo masters by letting your sequencer control an orchestra of MIDI instruments in real-time: no chance of signal degradation - the sounds are always of the original quality. Of course, if you insist on using a vocalist (or a violin section) that hasn't had a MIDI retrofit...

MIDI delays are not usually the problem you have been told they are - especially if you own exactly two synths and one MIDI cable! When you own 27 different MIDI-equipped devices, you can afford to solve the problem anyway. The MIDI 1.0 Specification does consider daisy-chains of more than three devices to be long chains, however... maybe you should too!

In an emergency, normal hi-fi leads of the male, 5-pin DIN, 180 degree, straight through connection variety will work just fine (even though they don't meet the MIDI Specification). Having said that, you should endeavour to buy the proper cables, of course. I mean, compare the cost of the proper cable (shielded, twisted pair, etc) to the cost of the two devices it is connecting... not to mention that one hour of studio time wasted because of cable problems which will dent your bank balance a lot more than a good quality, purpose designed cable.

The MIDI 1.0 Specification is quite clear that you must connect one output to one and only one input. This means that you are not allowed to start soldering up neat little Y-cable 'splitters' like you do for audio or even for trigger pulses. If you need this facility, then buy one of the proper MIDI boxes that advertise their abilities in this area.

MIDI uses a serial bit stream to carry information. This is exactly what the RS232C or RS423 interface you may have on your microcomputer does too. Don't let some computer salesman tell you that since MIDI is "the same as RS232C" you can use the RS232C port as a MIDI interface, you can't. Look him straight in the eye and tell him that MIDI is 'current loop' and 'opto-isolated' and watch him shrivel up before your eyes!

MIDI messages are like sugar in coffee: if you need three lumps (bytes) then two just won't do the job! Which brings me on to the subject of error-correction (what?!). Unlike many computer networks - and MIDI is a computer network whether you like it or not - MIDI has no error-correction capability. In fact, the only error-detection capability provided is that specified by some of the manufacturers in their own System Exclusive messages.

What this means is quite simple: in general, if a single bit in your message (or, heaven forbid, a whole byte) gets corrupted, gets lost, goes walkabout or whatever, then you are in trouble. Something nasty is (probably) going to happen! Two common symptoms of this phenomenon are (a) drone notes, and (b) voice data which was saved to disk but just refuses to load back into the sampler or synth.

Don't despair if moving a control (such as the sustain pedal, for example) on your master synthesizer doesn't cause anything to happen on the slave. You will have to read both manuals carefully to find out which controller number is being sent and which one is expected. You may well be able to change one or the other...

Not all of the MIDI messages work on the basis of channels where different devices, or sub-devices, can be 'targeted' to a specific instrument by including a MIDI channel number in the message. Where targeting is possible but doesn't seem to be working... don't panic! Check that both the intended transmitting instrument and the intended receiving instrument are set to the same MIDI channel (manuals time again, I'm afraid). If other devices apart from the intended receiver respond when they shouldn't, make sure that they are not set to the same channel and that they aren't in Omni mode.

Oh yes, somewhat on the same subject (sub-devices): if you have a synthesizer module that is multitimbral, then you have a duty to investigate MIDI Channel Mode 4 (Omni Off/Mono). This is how you get to control the multitimbres (pardon?) separately, ie. a separate track on your sequencer can play each set of voice circuits independently. I know it's hard work but it will be worth all the effort in the end!

Another piece of work that may save you some time: investigate what defaults each piece of equipment has (ie. the parameters that are set up automatically when you first switch the device on, for example). It may well be that you can set and store these defaults so that the device powers up in the correct mode, listening and transmitting on the correct channels, etc.

Find out if both your sequencer and your drum machine can handle MIDI Song Position Pointer (SPP) messages; if either can't, skip the rest of this paragraph. If they both can, you have now found a way of recording drop-ins without either playing back right from the start of the song (to get to the right place on the drum machine) or not having the drum machine switched on at all! If this doesn't make sense, read up in both manuals about Song Position Pointers.

MIDI is not fixed forever (written on tablets of stone...). Conferences, often held via electronic mail over international communications networks, are held and various consultations are ongoing to discuss the future of MIDI, its expansion and enhancement. A more comprehensive timecode facility (related to SMPTE) and a standard file format for dumping sample information are already in evidence.

Many, but not all, musicians see MIDI as a great 'enabling technology'. Some decry its event-driven structure as placing a constraint on free musical expression... They also see a problem with the fact that MIDI isn't microtonal - however, this should be less of a problem now that at least one instrument, the Yamaha DX7II (and related modules), deals with microtonality itself.

Okay folks, that's enough for this month. I'll see you next time when we will return to some more of the detail of MIDI's innermost secrets. See you then.


Read the next part in this series:
MIDI Matters (Part 4)

Previous Article in this issue

Casio HZ-600 synthesizer

Next article in this issue

Inside Views: J L Cooper

Sound On Sound - Copyright: SOS Publications Ltd.
The contents of this magazine are re-published here with the kind permission of SOS Publications Ltd.


Sound On Sound - Oct 1987




MIDI Basics

Part 1 | Part 2 | Part 3 (Viewing) | Part 4 | Part 5

Feature by Jay Chapman

Previous article in this issue:

> Casio HZ-600 synthesizer

Next article in this issue:

> Inside Views: J L Cooper

Help Support The Things You Love

mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.

If you value this resource, you can support this project - it really helps!

Donations for November 2020
Issues donated this month: 2

New issues that have been donated or scanned for us this month.

Funds donated this month: £43.00

All donations and support are gratefully appreciated - thank you.

Please Contribute to mu:zines by supplying magazines, scanning or donating funds. Thanks!

Monetary donations go towards site running costs, and the occasional coffee for me if there's anything left over!

Small Print

Terms of usePrivacy