First Steps In Multi-timbrality
Parts, channels, timbres, tracks — the terms and techniques of technology-based music can be so baffling to the novice. Is it really that complicated? Not when you know the basics, as Martin Russ explains.
You've just bought a multi-timbral expander. The manual tells you lots that you don't want to know and doesn't seem to answer any of the important questions:
• How do I connect it to my computer?
• How do I use it?
• Is there any way to make all this equipment any easier to use?
• How do you put songs together?
Well, here are some answers...
Connections first. Assuming that you've already got a synthesizer and a computer, you need to connect it up as shown in Figure 1. Notice the 'trick of the trade' in this diagram. The bypass switching looks weird, but is very useful because it enables you to play the synthesizer and connect it to the expander, either when the computer is turned off, or perhaps when the computer is being used to play a game. Despite what many music software manufacturers think, you won't be using their product all the time. If you can't afford the MIDI switches (you did just buy the multi-timbral expander, after all) then you can just wire it up without the switch, as in Figure 2.
Of course, to use the synthesizer to play the expander without the computer you need to do a bit of manual cable changing (shown in grey). Figure 2 shows the idea — you just connect the Out of the synthesizer directly to the In of the expander.
Multi-timbral expanders pack an awful lot into a very small box. Although they may only have one set of MIDI sockets and a stereo pair of audio sockets, inside they contain several independent sound generators, which means that you can make more than one sound at once. Each sound is called a timbre — hence the use of the word multi-timbral. The hardware and software which produce each separate sound can be called a part, partial, or instrument — I reckon that 'part' suits the musical application best, because the box is probably too small to fit any instruments into, even imaginary ones! Music is often referred to as being made up of parts — bass parts, drum parts, and so on, so choosing the same name helps avoid too much technology-talk.
So, the expander behaves like lots of instruments, and these instruments will be playing specific parts of the music. MIDI (Musical Instrument Digital Interface) provides 16 separate channels for controlling separate parts, so all you need to do is assign Parts to channels. Deciding on how to split the music into parts is the tricky bit — you need to work out how many different instruments you are going to use. You then set the Parts in the expander to make the right sounds. A 3-piece rock band might comprise guitar, bass and drums, whilst an orchestra might need strings, percussion, brass, and woodwinds. Dance music would need a backbeat made up of a drum track with an associated bassline, and then another couple of parts for adding sampled sounds: vocal sounds, looped effects, distorted guitar chords, brass stabs, and so on. New age or folk music might have percussion and a bass part, with harp or plucked sounds for arpeggiated chords, and then a flute sound for the melody line.
Most multi-timbral expanders will have a wide selection of sounds, and you should find some which suit your requirements. The very latest multi-timbral expanders tend to have General MIDI (GM) compatibility, indicated by a small logo on the front panel. Amongst other things, General MIDI defines the sound that each Program Change message will select. For example, the basic Acoustic Grand Piano is Program number 1 (also called Patch number 1, or perhaps Memory number 1, depending on your expander). Of course, different manufacturers will have different ideas of exactly what sounds should go with the names given in the General MIDI definition, which means that listening is still the best way to choose sounds. General MIDI sounds are arranged in blocks of eight. By Program Change numbers, these are:
|9-16||Chromatic Percussion (Xylophone, Vibraphone etc)|
|49-56||Ensemble Sounds (Choirs etc.)|
|65-72||Reeds (Sax, Oboe, etc.)|
|73-80||Pipes (Flute, Recorder, etc.)|
|105-112||Ethnic (Sitar, Banjo, etc.)|
General MIDI is useful because the sounds are approximately the same for any GM compatible expander, so you take a sequence written on your sequencer for your GM expander, play it on another sequencer/GM module combination, and it will sound much the same as it does on your own system.
If you want to quickly run through some sounds on the expander without using the computer or the sequencer, then just connect the MIDI Out of the synthesizer to the In of the expander and set the MIDI receive channel of the expander to the same channel number as the MIDI transmit channel of the synthesizer. Note down the sounds which you like for later reference. Most expanders provide a wide range of drum and percussion sounds as a single 'drum set', which you can assign to one part, so you don't need to have separate parts for the bass drum, the snare, the hi-hat, and so on; each percussion sound will be triggered by a different note on the synthesizer keyboard. The manual should have a diagram showing what drum is on which note, but it is useful to play around with the sounds and get the feel of where they are and what they sound like in combination. You may produce some drum patterns that you like the sound of and want to store them away for future use, which is where the sequencer comes in.
Sequencers almost always work like a multitrack tape recorder. Tracks are recorded one at a time, usually starting with some sort of click track, bassline or drum pattern as a timing guide. You then play new tracks of music, which the sequencer records at the same time as playing back previously recorded tracks, so that you can hear the new, track in the context of the whole piece of music.
Each track can be assigned to a separate part inside the multi-timbral expander, although it is often better if several tracks are assigned to just one part. Musical parts can thus be broken up into simpler groups of notes, either to simplify the performance on the keyboard, or else to offer flexibility when editing. If you can't play left and right hand at the same time, then don't. Record the two separately and let the sequencer play them back together. In fact, you may find that you can concentrate better on getting a smaller number of notes exactly right, rather than having to try and do the whole thing in one go.
For example, a bass root note plus a triad chord, with an associated melody note could be recorded as one track, or recorded as three separate tracks (see Figure 3). The one track approach is fast if you get it right first time, but it might be difficult to perform editing on lots of notes all on the same track. You could assign the separate tracks to the same or different parts in the expander by selecting the appropriate MIDI channel(s), and then the instrumentation, transposition and other playback settings of each track can be easily changed — this would be very time-consuming for a single track.
Note that the tracks, parts, and MIDI channels are quite different things — the 16 MIDI channels do not mean that you are restricted to 16 tracks, nor are you restricted to 16 parts. Adding a second MIDI Out socket and getting a second multi-timbral expander can be expensive, which means that most people tend to stay within the 16 MIDI channels and 16 parts at first.
The playback and record facilities of the sequencer are almost always used simultaneously, with 'live' playing on the synthesizer keyboard accompanied by the playback from the sequencer. The computer effectively merges the output of the sequencer with the 'soft thru' which is dealing with the MIDI data from the keyboard. The merged Out is then connected to the Thru box, and so to the MIDI Ins of your expander(s).
The most important aspect of setting up the expander is MIDI channel allocation. Assigning channels to separate parts should take into account the polyphony you are going to need for each part. Basslines and melodies usually only need a couple of simultaneous notes, whilst pads could use six or more notes at once — even when you are only playing three-note chords, there are often times when notes from one chord are decaying over the start of the next. Once the parts have been allocated to channels, you should keep a careful note of how they are assigned — the important setup is no longer how the expanders are wired physically, but how they are assigned to MIDI channels.
Here is a sample MIDI channel allocation for two expanders:
|Name||MIDI channel||Polyphony||Expander||Prog. No.||Part|
|Bass||2||2||Boss Dr Synth||37||1|
|Strings||6||6||Boss Dr. Synth||49||4|
The names tell you the sound that should be set for that particular part — you could use a MIDI Program Change message on the appropriate MIDI channel. For GM-compatible expanders, this Program Change number will always select the same sound, regardless of who made the expander. In this case, both the TG100 and Boss Dr. Synth are GM-compatible, and so we automatically know that the Flute is selected by Program 74. The drum part uses another aspect of GM: the definition of drum sounds on specific notes on the keyboard. Once these sounds are set up, we can almost forget them, since the only thing we now need to remember is the MIDI channel for that part. The polyphony column simply refers to how many simultaneous notes the part will require; this is not so important now that almost every multi-timbral expander on the market simply assigns available voices to parts as they are required, but once upon a time it was necessary to pre-define how many notes would be available to each part.
Thinking just about the parts, the tracks on the sequencer could be set up like this:
This is the obvious and simplest way of doing it. Using the idea of splitting tracks up to make things easier when editing, we could use a different set of tracks in the sequencer:
The two flute tracks contain two attempts at the same melody line. This allows you to cut and paste the best bits of each into another track later. The main bass track contains the standard bassline, whilst the bass walk track has a fast arpeggio down to the leading note for the next bass note (see Figure 4). Record on a separate track any fast bits, fiddly phrases, difficult to play counterpoints, or just plain experiments, so you don't mess up the basic bassline. The blocked piano chords are on one track, with some linking arpeggiated chords on a separate track, which allows you to get the phrasing exactly right later instead of in one go! The strings are split into two sections so that they can be octave transposed as required. The drum sounds are each allocated their oen tracks, just as you might do in a multitrack recording studio. This example uses 12 tracks, but only five MIDI channels, which shows that the 16 MIDI channels are actually more useful than they initially appear.
These two ways of using the tracks of a sequencer show that the tracks can be independent of both the MIDI channels and the expander parts. As long as you know the MIDI channel of the sound you want then you can use tracks to suit your musical purposes instead of to suit MIDI's numbering. It is also possible to misuse the MIDI channels for creative purposes — so, for example, you could copy the piano arp track to track 13 and assign it to the string sound on channel 6, thus 'doubling' up the piano notes but not the block chords! Doubling or layering two (or more) sounds is easy with a sequencer — you just copy a track, or even just alter the channelising of the 'soft thru' for instant real-time doubling. Mixtures of string sounds from two differing instrument types (FM and PCM samples, for example) gain a richness and depth, whilst the contrast between a sharp attack sound and a slower sound can produce complex evolving sounds. Sounds with different velocity sensitivity can give very expressive timbre changes as you play.
For playback, the possibilities offered by selective doubling (the strings playing some of the piano, for example) are enormous. By copying just the initial chord of a set of fast chords to another track, you can then emphasise that start by doubling with another sound. The sounds you use for doubling can be different for different sections of the song — so you could use a track with a brass sound for doubling the strings during the verses, but use another track with a synth string sound for the choruses. Once recorded, you can copy or cut out groups of notes and use them on another track. One of my favourite tricks is to split the alternate notes of arpeggios into two tracks, so that you can then play them with the same, or different sounds. Combined with differing pan settings, the results can be very impressive, especially since you can still double with another sound using the original 'unsplit' track. Do not be afraid to use tracks creatively or just make things easier — most sequencers provide plenty of tracks to work with. And don't forget to give your tracks descriptive names which will help you if you ever come back to the song...
That should have given you some ideas to try out. Don't forget to go back and re-read the manuals after you've had the software or expander for a while; you will find that they make a lot more sense when you are familiar with the products. Above all, keep things simple and understandable. Once you have named the channels and set up the sounds, you can forget all the MIDI stuff and deal just with the named Parts. If you can't play something, slow the sequencer down in tempo, or add it in step time. Use the technology as a tool for making your own music. Have fun!
Feature by Martin Russ
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!