Performances

Beside the Voice button is Performance. Hit this one. You are now in Performance mode.
Performance mode can be used to make a multi (up to 16 voices, on different MIDI channels) - but it can also be used to make layers of voices, keyboard splits, or master keyboard-type setups that will allow you to control other synthesisers more efficiently. There are 128 Performance locations, all of them user re-writable.

The EX Architecture

Basic Structure
The EX has two basic play modes, Voice and Performance. Voice is used for a single voice, which can contain up to 4 elements. Performance is used for designing a multi setup to use several voices simultaneously, or else to make voice layers, splits, etc. and provides Master Keyboard functions that allow you to control many synthesisers from the EX front panel.

DSP Limitations

Sampling and AWM elements are not limited. Using just this technology you can access the full polyphony of your machine (64 on EX7, 128 on EX5) without any restriction. However, the insert effects, the VL algorithm, AN algorithm, and FDSP algorithms are all sharing the same area of DSP. When an element of this type is selected, the DSP area is sent the formula to perform that specific function. Since these formulae are of various sizes, it is hard to predict how many AN elements, or insert effects you may be able to use in a certain situation. Your EX will tell you when the DSP area is full, though, with a very irritating message that youll see over and over again when trying to make multi setups on your EX:
DSP Resource Full means you cant turn on any more DSP algorithms without first going back and turning another off. It is up to you to manage the DSP and to decide how best to use it within any song setup. As you might imagine, the VL algorithm is very large, but a chorus effect is comparatively small.
Voice Types and Structure
There are several different TYPES of voices. Put your machine in Voice mode, and hit EDIT/COM/Param. (By the way, we will continue to use this method of describing a page, in which the sequence of keys pressed are separated by a forward slash.) This is one of the Voice Common edit pages.
The first parameter on this page is your Voice Type. The first choice of voice type is AWM. This is the common wave playback type of voice, and lets you include up to 4 elements, which can each include either ROM waveforms or your sampled RAM/FlashROM waveforms. On the EX5 and EX5R you will find the second Voice Type VL+AWM. This means that element #1 is now a VL algorithm (the other 3 elements are still AWM). Unfortunately the EX7 cant do this one, because the DSP area is half the size of the EX5. Next you will find the Voice Type FDSP. This Formulated DSP voice is a collection of processes to which you must send an AWM element. So in this case, your elements are still AWM, but each can be sent (or not) through the chosen FDSP algorithm. If youre still in the Voice Common page, youll notice that when you selected FDSP a new F-key appeared called FDSP (on F-key #5). Push this one, and you can decide what type of FDSP to use, and adjust its parameters. To the left of this page, you can see the ElmSw list of the 4 possible elements. Each of these can be turned ON or OFF to be either sent, or not sent, to the FDSP algorithm youve chosen. Well get further into FDSP in a little bit. The following Voice Type is AN+AWM in which element #1 becomes an AN algorithm. The AN algorithm is the analogue model, very similar to that used in the AN1x synthesiser. In the EX5 and EX5R you can choose either AN (Poly) and AN (Layer). AN (Poly) is two notes of polyphony in one element, AN (Layer) is two elements that are each monophonic. The EX7 has only monphonic AN. You can also choose Voice Type AN+FDSP (EX5 and EX5R only), in which the AN algorithm is sent through the FDSP algorithm. Lastly, you can choose type Drum in which each key can access a different element, or group of elements. You can have 128 Elements in a Drum Voice. Each of these voice types will be described in more detail in the Voice section.

Probably by now, your Open Saw is starting to get a little less than useful. You might want to consider adding some real-time control to the voice, so that you can adjust the filter cutoffs, resonance, LFO speeds, etc. Lets exit the LFO page, and drop back down into CTRL (Control). There are two sections to the Control page, the first is called Pitch. In pitch you can set the pitch bend range, and the upper and lower ranges can be set seperately. The lower range can be set to a maximum of 4 octaves (48 semitones). This page also lets you set the portamento time, and mode. The two modes are fingered and full time. If you set it to fingered portamento only happens when notes overlap. In full time portamento is always active. Now, lets check the second of the CTRL pages, called SET. SET is basically a controller matrix. You have 16 sets, which is a little like having 16 patch cords. You can assign 16 controller+destination combinations. Okay? You can turn each of these sets on or off for each element (to the left of the screen), although at the moment were working with a single element voice. So lets turn set 1 ON for Element 1, and the turn Knob 1 on, and then down at the bottom set the Destination (DST) to 077: AWM DCF Reso. The DEPTH of this control can be set to the right, lets set it to maximum for now.
Of course, when youre in EDIT mode, the knobs are functioning as data entry for the edits you make, so if you want to check the results of this assignment without leaving the editor, hit the hardware switch called KNOB MODE to test the results of Knob 1. Hit it again to turn it off. Now lets say we want to use the Mod Wheel to control the amount LFO1 modulates the Filter Cutoff. Go up to the top of the screen and increment set1 to set2. Turn it ON for the element. Then turn on MW1. Set the Destination to 053: AWM LFO1 FMD (filter modulation depth), and set the depth again to maximum, like this:
Got it? Okay, lets have a look at the effects. As we mentioned earlier, there are two types of effects: insert effects and send effects. The insert effects can be configured in three ways; parallel, or serial (with either INS1 or INS2 first in line). You can set this on the front page of the EFCT section. This page also lets you set up the send and return levels for each of the send effects (REV and CHO) :

Of course, were not hearing this yet because the effects are currently off. Hit the INS 1 key to enter the first insert effect. Set the effect type to 16: Amp Simulator. Again, you choose which element is going into which insert effect on the left, where each element can be OFF, 1, or 2. Like this:
You can cursor through the parameters within the effect. Now try INS2. There are 79 different effects in here. Try 64 the Talking Mod, go back out to the controller set, and assign set3 so that something modulates Dest:020: EF2 Vowel. mad stuff. If your voice sounds great, store it somewhere using the STORE button.
Lets move on to another voice type.
Getting more gain If you need to get some extra gain onto an element, use the SCF (static filter). The last three types are just gain boost. Also check Common Volume, Oscillator Mix Level, Amplitude Parameter Level, and the Filter DCF Gain.

FDSP Voice

Voice Type FDSP
4 Elements Effects and FDSP Setup FDSP Algorithm

Element 1 to 4 AWM type

FDSP is a kind of process. As you see in the diagram above, it needs an input. In an FDSP voice, one or more of the elements must be routed through it in order to hear any effect of the FDSP. Generally, we route AWM elements through the FDSP. In the EX5 and EX5R and AN element can also be routed through the FDSP. There are ten FDSP algorithms. Lets have a quick look at what FDSP is, and what each of the algorithms do. FDSP is very similar to an effect processor, in that it takes an audio source and processes it. The main difference between FDSP and a traditional Effect is that the FDSP is polyphonic. Each note going through it is processed individually, thus allowing pickup simulators and so on to be modelled more accurately, and keyscaling and velocity to be used within the algorithms. Due to the limitations of the DSP used in the EX, an FDSP voice has a polyphony of 8. Look at this diagram:
Here is a brief explanation of the ten FDSP Algorithms in your EX: 1. EP Pickup This is a model of the electro-magnetic pickup used in electric pianos. 2. EG Pickup This is a model of the electro-magnetic pickup used in electric guitars. 3. Water 4. PWM 5. Flange 6. Phaser 7. Self FM 8. Tornado This is an original effect in which sample&hold filter modulation is passed through a resonating string algorithm. Sounds kind of watery. This realizes a Pulse Width Modulation sound on any input wave. A Flanging effect with envelope control for each note individually. A Phaser effect with envelope control for each note individually. The input is frequency modulated by itself. This is another kind of FM synthesis in which the signal is modulating a frequency of 0 frequency (DC). By varying the modulation intensity and feedback, FM type sounds or analog sync sounds can be created. Ring Modulation. This modulates the input signals amplitude with an oscillator and creates mad sidebands. (in fact there are two oscillators: main and sub) This is an overdriving, time-variant low-boost filter. Good for distortions and for fattening sounds. The attack transients can be clipped resulting in emphasized or compressor-like sounds.

Some parameters are controlled upside down There are several parameters that behave upside down when assigning controllers to them. Most parameters are offset by the controller value (filter cutoff, EG times, etc.), and the controller adds or subtracts from the default value. However, some strange exceptions to this rule exist, and Feedback is one of them. The Feedback default is the MAXIMUM that the knob can take it at full depth. FM Depth also works this way around. Its weird, but true. And we dont know why.

VL Voice

( EX5 and EX5R Only )

Voice Type VL

Element 1 VL Element Pointing at VL model Key/Velocity Range Voice Element 2 - 4 AWM type
Voice I1:004: Moby is a demonstration of the VL algorithm used in the EX5 and EX5R. You dont have access to the actual editing of the physical model within the EX, so instead there is a supply of basic models (256 of them, found in the OSC page). You can then offset parameters using the EX editing system. This isnt necessarily a bad thing. Firstly, the VL model is extraordinarily complex, and editing it through a little screen would be a nightmare. Secondly, the interesting side of VL is not the way it sounds, but the way it moves. You have full access to the controller map, and this is what makes VL interesting. Lets check out Moby. The concept behind Moby is that its a mutated brass type of instrument, much bigger than anything you could blow with your skinny little lips. Thus, when you blow into it, you can barely control the embouchure enough to keep it in one mode at a time. Try this, move Mod Wheel 2 all the way back to zero, and play very lightly in the middle of the keyboard. Just hold the note down, and try not to trigger the aftertouch. The sounding pitch should be quite high. Its a ridiculously high harmonic or overtone. Now slowly bring Mod Wheel 2 up, and youll hear the mode shift down through the overtone series until it finally reaches the fundamental. If you open the controller page, youll see a similar layout to that in the other voice types, but again, notice that the destinations are VL specific.
VL Embouchure is probably the coolest of all the VL destinations. This approximates lip tension in a brass or reed instrument. If you dont care about that, then just remember that its the parameter that can force the VL model into different harmonic modes. In Moby, its assigned to both the Ribbon, Aftertouch, AND Mod Wheel 2. Aftertouch is also adding to the VL Pressure destination, which is the amount of energy being forced into the VL model. If you want to get realistic, this is best done with a breath controller, but Moby is pretty far from realistic. The VL Pressure and VL Embouchure are not fixed discrete paths, as you find in most machines, but influence each other in a non-linear, organic kind of way. For example, if the pressure is higher, the embouchure will tend towards more stable modes. VL is a mysterious creature. Other interesting VL destinations to explore are Tonguing, Scream, Growl, Throat, Damping, and Absorption.

The Insert Effect can be turned on or off for each part, but of course, due to the limited DSP resource, you cant turn on very many at once. The Insert Effect will be that which was stored in the Voice itself.
The next page, LYR (Layer) include the Layer Switch (more on that in moment), and other Layer type parameters, key range, velocity range, note shift, and several switches for the MIDI transmit and Arpeggio settings.
The SOUND page is a collection of offsets to the Voice synthesis parameters themselves, such as filter and envelope settings.
The CTRL (Control) page lets you adjust the pitch bend range, and turn the controllers (knobs, etc.) on or off for each Voice. You cannot re-assign the controllers from within the multi. The PRE (Preset) page lets you setup whether or not the controllers and knobs are transmitted over MIDI. And also their default value which is sent when the Performance is selected. This can be very useful for Master Keyboard or Live Applications in which the EX is used to control other synthesisers. The COM (Common) page lets you set up parameters for the entire performance, such as the arpeggiator and controller numbers and effects settings.
DSP Resource Full ! - DAMMIT!
Yes, we know this message will drive you crazy. But try to be cool and lets sort out whats going on and how to cope with this. First of all, remember all the things that use up our DSP resource: VL Algorithms (EX5 and EX5R only) AN Algorithms FDSP Algorithms Insert Effects Any time any of these are in use, some amount of DSP is being eaten up. In the EX5 you may have 4 insert effects, or else 1 insert effect plus 1 synthesis algorithm. If you decide to turn on the insert effect on the Tube Crunch organ (channel 2), this will be fine. But if you now want to turn on the insert effect on anything else, youll get the dreaded message. However, if you decide you really dont need that AN voice after all, change that voice to a non-DSP voice, and then you can turn on 4 different insert effects. In the EX7 you may have 1 insert effect OR 1 synthesis algorithm. On the Tube Crunch organ (channel 2 of our multi) you cannot turn on the insert effect at all until you select a nonDSP voice on channel 1. In all the EX synths, if you try to turn ON an insert effect and the DSP resource is full, youll get the message and the insert effect will stay OFF. If you try to select a voice that requires DSP and the DSP resource is full, the following non-DSP voice will be selected instead. Theres only so much power, and you must decide how to best distribute it.

This one is easy enough. Make sure your floppy disk is DOS format (format it inside the EX if you want to be sure), copy a WAV file onto it. (Theres one on our floppy if you want to try it.) If youre a Mac user you must remember to name the file in an DOS manner. That is, 8 (or less) characters, a dot, and WAV. For example: SAMPLE.WAV Or in the case of an AIFF file, it must be named SAMPLE.AIF On the EX, go to Disk mode, File Load, and choose WAVE. The WAV file on your disk will appear as a WAVE to the EX. Load it. Your done. With an Akai floppy disk, you will see the Akai Program. So make sure that any Akai samples you want are included in the program.
Loading a sample via SMDI
You must have a SCSI board in your EX for this transfer. If you have sample data inside your PC or Mac, you can send this data over SCSI directly to the EX RAM area. SMDI is a standard protocol for sending samples via SCSI like this. Check the SCSI ID of the EX (in Utility/Other), since your computers sample editor will want to know this. Open your sample in the editor, and set it up to send to an SMDI device on the correct SCSI ID. Be careful that you dont duplicate any SCSI IDs with other devices on your SCSI bus. Basically, you just SEND from the computer. If all is correct the EX will show an SMDI Data message while its receiving the data. Be careful not to overwrite existing sample data. SMDI requires a sample position to be designated. If you tell it to send to sample position #1, any existing data in the EX at sample postion #1 will be overwritten.

Sample Editing

So now your sample is inside the EX. (If not, load the LARGE.WAV) from our floppy disk. Enter the SAMPLE Mode, and choose the sample number with our sample in it (very likely, its number 1). So hit EDIT and well enter the Sample Editor. There are three main pages in this section, PARAM, LOOP, and NAME. NAME is quite obvious. Choose PARAM first.
Here you can set some defaults for the playback of this sample. Many of these things can be changed from the Voice using this sample, so its not necessary to make any permanent decisions here. You can choose a default original key (Center), and a default playback method. Lets turn the Loop ON for a moment. (FwdLp = Forward Loop) The bottom three numbers are the sample start, loop start, loop length, and sample end points. But lets hit the LOOP page while we look at these.

The first column here represents the timing of the pattern. The second column represents the KEY CODE. Now the key code is not the note itself, but is a representation of the position of the notes. For example, the key codes were assigned by the arpeggiator to our original sequence like this:
The lowest note is 1, the next lowest is 2, and so on.
When you hold down keys on the keyboard, it replaces these CODES with the lowest note you play, the second lowest you play, etc. Got it? Now, you could theoretically leave the sequence as is, and use it. But, if we tweak it out further, it can be much more useful, or more musical. The main problem with it at the moment, is that if I put my fist down on the keyboard, the pattern that results doesnt have any of the internal relationships or vibe of the original. Why? Because the arpeggiator called the third note CODE 6, whereas my interpretation is that its actually the first note up an octave. So lets fix it. In the ARP EDIT window, lets go in and change the third event to CODE 2, but in the O column (octave column) lets put +1. Lets also do this for the tenth (second last) event.
Lets do the same for the 8th and 11th events. Lets call these CODE 5 (second note) but with an octave setting of -1. And while were at it, well call the fourth event CODE 4, +1. Basically, Im just removing all excessive key codes, and replacing them with octaves. Now my entire pattern looks like this:
There are only four key codes remaining, 2,3,4, and 5. So now, I only have to hold down four notes on the keyboard to supply enough information to get the arpeggiator to crank out something resembling my original idea. And to tidy it up further, Im going to rename all CODE 2 to CODE 1, rename the CODE 3 to CODE 2, rename CODE 4 to CODE 3, and rename CODE 5 to CODE 4. This just makes it easier to read, and understand what is going on when you hold down 4 notes on the keyboard. So finally, my pattern looks like this:
Of course, you dont have to begin with a MIDI file. If you know what you want you could enter it directly in this editor (using the INS function). You could import any data from a track or pattern on the internal sequencer. Or you could record the basic pattern in real time in the Arp Editor itself. Exit this page to go back to the top of the Arpeggiator section. You can name your arpeggio pattern in the NAME page. And before we leave this, lets have a look at the Play Effects (PFX) page. First of all, turn OFF the FxThru switch on the main ARP page:

These default to FxThru ON (which means bypassing the play effects). Now enter the PFX page itself. Play Effects are offsets to the MIDI data, much like the inspector parameters in Cubase. You can offset velocity, note length, etc. and you can impose quantize grids (called Groove) on the data, without actually changing the data itself. Hold down some keys, and start scrolling through the Groove Type list.
You should be able to hear how each groove is altering the feel of your arpeggio. The last groove is called 101:user, and you can design your own groove in this space. You can also edit the existing grooves if you want. Below the groove stuff are some basic offsets, so you can alter the Velocity, Gate time, etc. of your pattern as well.
Arpeggios disappear when you power down. Back em up.
Yes, this is a drag. Sorry, but it uses the same RAM area that the patterns and songs are in. You can save the Arpeggios to disk. And you can load them again, individually if necessary, to build up your own library of patterns. In the Utility/Voice Mode page you can set the Arpeggiator to HOLD or not, and you can choose a MIDI channel to transmit the Arpeggiator to another synthesiser.

Patterns and Keymaps

Making a Pattern
Patterns are actually used in various ways in the EX, so its useful to learn how to make your own. Basically, you can think of the Pattern as a loop, much as if youd loop a section in a sequencer. It can have 8 tracks, each of which can be on any MIDI channel and can have different loop lengths. A Pattern cannot be longer than 16 bars, though. To demonstrate a Pattern that uses different Voices at the same time, lets choose a Performance to work within. Choose the one called Pattern, Performance #004. This Performance has a Drum Kit on channel 1, then an Rhodes sound on 2, and upright bass on 3. Hit the Pattern button.
Set a tempo you want to try working at, and then hit the PFX f-key. This takes you to the Performance Effects page, which we looked at briefly in the Arpeggiator chapter.
Again, with the patterns, you can offset many performance features of each track. You can introduce groove quantisations, transpositions, etc. But for now, all were interested in is the Length parameter, since this sets the length of the track.
Notice that when changing the Length parameter, there are two options. These are shown by the f-key selection below, which can be ALL or TR. If its set to ALL, then ALL the tracks in that pattern will be set to the Length you choose. If its set to TR then only the selected Track will be affected. Same goes for the other parameters in the PFX page. Im going to set the entire pattern to 1 bar for the time being. You can always go in and lengthen another track afterwards. So I set the Length to 2. Why 2? Well, because actually youre setting the Loop position, not the Length. So I want it to Loop at Bar 2, Beat 1. You can actually Loop at a beat within a bar if you like. (The Beat is displayed after the Bar in the Length parameter). This setting sets all tracks to Loop after 1 bar:

So now, I go back to the PTN page, and I see that the MaxLng display says 1. This displays the length of the longest track in the pattern. Using the Sequencer transport controls, hit Record then Play. The metronome should be ticking out its beat, and you can play something into the pattern. Its just looping the one bar around. Play a simple Kick drum or something into the pattern. Hit Stop, and then hit Play to make sure youve got it. Notice that when you hit the Record button, there are a choice of Record modes, which are equivalent to similar choices you have in your software sequencer. Here are your options.
This can be on or off. If Multi is on, then all tracks are recording at once. This can be useful if you have several MIDI channels coming into the EX from another sequencer, or from live MIDI instruments. When its OFF, you can choose which track to record on the bottom left of the screen. Replace mode. Whatever you record will replace the entire contents of that track. Overdub mode. Whatever you record will be added to the existing data within that track. Step entry. Sometimes this is the quickest way to get what you want, just choose the value of each note and hit it. In it goes. Notice two things in the step entry: 1- the note values can be quickly chosen with the ten-key pad 2 - if you set the velocity to EXT it uses the velocity at which you enter it

RPLC OVER STEP

So well record a kick drum, or something simple into Track 1. Make sure MULTI is OFF, and put yourself in RPLC mode. Hit Record, Play. Play a couple of beats. Stop, play it back, make sure youve got it. Second track. Choose Track 2 with the Program buttons on the right end of the front panel. Its a Rhodes piano sound. Hang on, we want another drum track, but this Performance has a Rhodes on channel 2. No problem, we change the channel of track 2 to 1. Hit TCH page:
Change the Transmit Channel of track 2 to 1 (as Ive done in the screen above). Notice you can also choose which MIDI output to transmit on, and you can disable the internal tone generator (TG) if you want to control only other instruments with the pattern. Now go back to the main PTN page, and Record another drum part into track 2. Some kind of rhythm happening? Alright, lets add a bassline on track 3. I want to make my bassline two bars long, so now I choose Track 3, and hit the PFX page. I choose TR, so that what I do only happens to the selected track, and change the Length parameter to 3.1 (2 bars). Like this:
Go back to to the PTN mode and record it in real time. Damn! I played it perfect, except for one note. No problem, I just hit EDIT and fix that note.

In EDIT you can see the start times of the events in the left column, written as bars, beats, and ticks (at 480 ticks per quarter note). The next column is the note, then the duration, and finally the velocity. You may see controllers and so on in here also, particularly CAT, which is Yamahas way of saying Aftertouch. (Its actually Channel AfterTouch, as opposed to Polyphonic AfterTouch). If you see too much CAT in your Patterns, you could always disable the Aftertouch in Utility/SEQ.) More on this EDIT of MIDI data in the following Sequencer section. For now, lets add some more tracks, and name the pattern something to differentiate it from all the other New Tracks in there. But lets see how we can use this pattern.
Patterns disappear when you power down. If you make a good one, save it to disk.

Making a Keymap

Lets imagine youve got a whole load of patterns, a bunch of samples, etc. and you want to jam them out and try various combinations of them. Wouldnt it be cool if you could have each key on the keyboard trigger a different one of these? Well thats what the Keymap mode is for. There is only one KeyMap at a time in the EX. It can be turned ON from either Voice or Performance mode, and this Performance or Voice remains active UNDERNEATH the KeyMap. In the case of our Pattern, we need the Performance called Pattern to be active for it to make sense. So lets stay with Pattern and turn on KeyMap (hardware button called KEYMAP). The first thing you might see after turning it on is a screen like this:
If you dont, hit it again twice. The KEYMAP button is a toggle switch. Choose C1 either by scrolling or using the SEL key while hitting the MIDI note. Where it currently says off you can choose either a Pattern (ptn) or a Sample (smpl). Choose a pattern first of all. To the right, youll see a choice of all the Patterns in memory. In between youll also see a column that describes either a Track number or all. This is so you can choose just one track from a pattern. Set it to all for the time being:
Choose the Pattern that we made in the previous section, and hit C1. The Pattern starts to play. Hit it again to stop it. Cursor over to the far right where it says toggle. There are three modes of triggering for each Key in the KeyMap. Toggle means that you hit the key once to start and again to stop. Oneshot means that it will play the entire pattern once, without any looping. Gating means that it triggers much like a normal MIDI note, ie. it plays while you hold it down, and stops when you release it. This is my preferred mode, but its up to you whats appropriate for your pattern or jamming style.

Again, like Pattern, you have a choice of recording MULTI or not, as well as a choice of recording modes. There is one new mode here, not found in Pattern, called PNCH. Punch recording lets you define an in and an out time for recording. This is useful if you just want to record for a short passage within an otherwise perfect take. The Src parameter you see highlighted above can be switched between normal and arpeggio. If its in arpeggio then the output of the arpeggiator is captured as MIDI data, thus freeing the arpeggiator for use elsewhere.
So lets hit record and play any old stuff into a MIDI track. Done? Does it play back? Okay, hit EDIT.
This is the display of note data, hopefully the notes youve just played. There are two modes to this editor, INS (insert) and CHNG (change). In Change mode, as above, you can alter any of the fields with the data wheel or the ten-key pad. The first field is the time of the event, in bars/beats/ticks (at 480 ticks per quarter note). The next field is the pitch of the note. The third field is the length of the note, again in beats/ticks. (The note-like display to the right of this tries to display the closest musical equivalent of the duration, by the way.) Finally, at the last column, we have the velocity. If you hit INS, you can insert a note or any other MIDI event. Like this:
Lets have a look at the other types of tracks. The Tempo track lets you define tempos at any position in the song, so that you can change speed when necessary. Its usually best to do this with INS, since you cant record this real time anyway. If youve imported a MIDI file with tempo changes in it, you can edit them with CHNG. The Play Effect track (Fx) is a strange one. Here you can record real-time. Try it out. You can adjust the Groove template, change velocity offsets, etc. through the Song. Record some of this and then hit EDIT.
Mad, innit? Now, no-ones suggesting that this is the way you would like to write music, and its probably not your preferred way to edit your MIDI data. However, this sequencer can be useful for several things. For example: Import a MIDI file to see what the Play Effects can do to your grooves There are many Groove templates in here, and you can apply them in varying amounts to the timing, the velocity, etc. Use the Sequencer to play sequences live. Okay, theres only one Song, but it can hold up to 30,000 MIDI events. So you could have an entire set in there, just chained together as one song. Use the Sequencer to capture improvised ideas that occur to you when the computer is off. Use the Sequencer to capture the Arpeggiator output, copy it to a Pattern, and free up the arpeggiator for something else The only way to get your MIDI files into Patterns or Arpeggios is VIA the Sequencer Impress your Mum with a MIDI file of Griegs Piano Concerto while you pretend to play

So you see, its not a complete waste of space.
Sequences disappear when you power down. If you make a good one, save it to disk.
Data Management and Utilities
How do I Save/Load one Voice?
You cannot Save one voice. You can Load one voice from either a Voice file or a Synth ALL file. (see next section) So, if you want to build your own library of Voices, you load individual voices from disks with Synth ALL or Voice files on them. Lets say you want to load a single voice from the floppy that came with this guide. Go to DISK/File Load/Synth ALL
Where it says All Data, you can change this to decide exactly what you want to load. You can also choose the destination if you are loading a single object. Like this:
This is what we call heirarchical loading. With this technique, you can fill the user memory of your EX with your preferred voices. For more detail on the various saved files, check the next section.

What is Saved With What?

There are many different file types for use with the EX. Its important to know what is saved, and how you can load it again. Here are the Save types, along with a description of what can be loaded from them.

ALL Data

This saves everything; samples, sequences, patterns, etc. If you save this you are sure to be able to get it all back again. The only drawback with Save All is that its non-hierarchical when loading. In other words, when you have an ALL file on disk, you can only load ALL of it again.

Synth ALL

Synth ALL saves all the voices and performances. Not waves, not patterns, not arpeggiators, not sequences, not keymap. Synth ALL DOES save your system setup, however. Synth ALL is hierarchical, so you can load one single voice or performance from a Synth ALL file on disk.
This saves only the Voices, not the performances or system. Voice is hierarchical when loading.
Wave saves all the samples and waves created that use those samples. Wave is hierarchical when loading. You can load a single wave from a Wave file on disk.
Standard MIDI File. This saves the Song to disk as a MIDI file type 0. Play effects are not included in a MIDI file, obviously. Note: You can play MIDI file type 0 directly from disk if you like.

Where do we go from here?
If youve followed through most of this book, you should have an idea of the huge potential of this instrument. Not only is the EX providing the raw elements of many synthesis techniques, its also giving you some compositional tools to help generate ideas. Once you understand the basic architecture of the machine, its often useful to forget about the designed purpose of the various sections, and to strike out on your own. For example, the Electric Piano Pickup FDSP algorithm might sound great with a vocal sample going through it. Using the resampling feature, you can sample the results of one DSP, and then apply another to the result. You could import your drum grooves, turn them into arpeggio patterns and manipulate them with the Play Effects, transpose it realtime, etc. Put your breakbeats through the Seismic FDSP and turn them to mush. You could import each section of your track as a seperate pattern, and jam the tune live from a KeyMap. Or make a sixteen part keyboard split as a Performance, set them all to the same MIDI channel, turn the arpeggiator on and stand back. A lot of interesting sounds are creating by merely trying stuff out, experimenting without a fixed purpose. Mistakes can be very cool. Miles Davis said There are no wrong notes. Brian Eno said Make the machinery fail. In other words, go for it, break the rules, and make it squeal like a pig.
And now, my son, you must eat this book.