YAMAHA
AUTHORIZED
PRODUCT MANUAL
DIGITAL PROGRAMMABLE ALGORITHM SYNTHESIZER
YAMAHA
DIGITAL PROGRAMMABLE ALGORITHM SYNTHESIZER
OWNER’S MANUAL
CONGRATULATIONS!
Your Yamaha DX21 Digital Programmable Algorithm Synthsizer incorporates the
state-of-the-art in digital FM tone generation technology, providing extraordinarily vibrant, rich voices and outstanding playability. The DX21 actually has two FM tone
generators, permitting two voices to be combined and played simultaneously, or in a
split keyboard configuration. The DX21 has a 32-voice programmable RAM memory
from which any voice can be selected at the touch of a button, a built-in 128-voice preset
ROM memory from which voices can be called into the RAM memory as desired, 32
special “performance” memories which can be programmed with combinations of voices
and some function parameters, and a cassette interface that permits unlimited storage
of FM voices. Of course, the DX21 is fully programmable, allowing you to create your
own FM voices or sound effects. Broad MIDI compatibility is also provided so the DX21
can control or be controlled via other MIDI compatible music equipment. To ensure
that you gain maximum benefit from all the performance and flexibility provided by the
DX21, we urge you to read this owner’s manual thoroughly while actually trying out
all of the available functions.
CONTENTS
PRECAUTIONS.. ............................................................
2
CHAPTER I: SETTlNG UP...........................................
1. Audio Outputs ..........................................................
2. Optional Foot Controllers & Foot Switches..............
3. Optional BC1 Breath Controller ...............................
4. Headphones .............................................................
5. MIDI Terminals.........................................................
6. Cassette...................................................................
7. AC Power .................................................................
8. Music Rack Installation ............................................
3
3
3
3
3
3
3
4
4
CHAPTER II: PLAYING THE DX21 .....................................
1. The 32 Voice RAM Memory ...................................
2. The SINGLE Play Mode ..........................................
3. The DUAL Play Mode .............................................
4. The SPLIT Play Mode .............................................
5. The 128 Voice ROM Memory.......................................
5
5
5
6
6
7
CHAPTER III: THE FUNCTION MODE................................
1. Accessing the FUNCTION Mode ..............................
2. Entering Function Data..........................................
3. The Performance Parameters...............................
4. Tuning Functions ....................................................
5. Memory Management Functions...........................
6. MIDI Functions .........................................................
9
9
9
10
14
15
20
1
CHAPTER IV: THE PERFORMANCE MEMORY.................
1. Performance Memory Organization ........................
2. Accessing the Performance Memory ......................
3. Programming the Performance Memory...............
4. The Performance Name.........................................
22
22
22
22
22
CHAPTER V: VOICE PROGRAMMING..............................
1. The Basics of F M Synthesis...................................
2. The EDIT and COMPARE modes .............................
3. The Voice Parameters .............................................
4. STORING Voice Data ..............................................
5. Two Approaches to Creating
Your Own Voices .....................................................
23
23
28
28
36
GENERAL SPECIFICATIONS ......................................
37
MIDI DATA FORMAT .......................................................
1. Transmission Parameters .......................................
2. Transmission Data................................................
3. Reception Parameters .............................................
4. Reception Data .......................................................
5. System Exclusive Data ............................................
38
38
38
41
42
45
DATA CHART ..................................................................
49
36
PRECAUTIONS
1. Location
2. Cleaning
3. Service and
Modifications
4. Relocation
5. Handing
6. Electrical Storms
(Lightning)
7. Electromagnetic
Fields
Choose the installation location for the DX21 with caution. Avoid locations exposed
to direct sunlight or other sources of heat. Also avoid locations subject to vibration,
excessive dust, cold or moisture.
Do not attempt to clean the exterior with chemical solvents, as this may damage
the finish. Clean with a soft, dry cloth.
Do not open the cabinet or attempt to make your own repairs or modifications to
any part of the instrument. Such actions may not only result in electrical shock or
damage, but will also void the product warranty. Refer all servicing to a qualified
Yamaha service center.
When moving the instrument be sure to unplug the AC mains cord as well as all
other connecting cables.
Avoid applying excessive force to switches and slide controls, dropping or rough
handling. The DX21 is ruggedly constructed using reliable solid-state circuitry,
nonetheless it is a fine instrument that should be treated with care.
Digital circuitry such as that used in the DX21 is sensitive to voltage spikes and
surges. Because of this, the DX21 should be turned off and unplugged from the
AC power outlet in the event of an electrical storm.
Digital circuitry is also sensitive to electromagnetic fields such as those produced
by television sets, radio receivers, transmitters, transceivers, etc. The DX21 should
kept at least several feet from such sources in order to prevent possible random
malfunctions.
2
CHAPTER I: SETTING UP
1. Audio Outputs
2. Optional Foot
Controllers & Foot
Switches
3. Optional BC1
Breath Controller
4. Headphones
5. MIDI Terminals
6. Cassette
The DX21 has two audio outputs, one for each tone generator channel. These
are labelled MIX/A and B. When a plug is plugged into the MIX/A output only,
the sound from both the A and B tone generator channels will be mixed and delivered
via the MIX/A output. When plugs are inserted into both MIX/A and B jacks,
however, the channel A signal will be delivered via the MIX/A output and the
channel B signal will be delivered via the B output. This permits sending a stereo
signal to a stereo sound system giving you the full impact of the DX21’s dual and
split play mode capability.
The remaining three phone jacks are for optional foot controllers and footswitches.
The VOLUME jack accepts a Yamaha FC-7 or equivalent foot controller for volume
or “swell” control. The SUSTAIN jack accepts a Yamaha FC-4 or FC-5 footswitch
or equivalent for sustain control–press for sustain, release to damp. And the
PORTAMENTO jack also accepts a Yamaha FC-4 or FC-5 footswitch or equivalent
permitting ON/OFF switching of the programmed portamento effect.
Yamaha’s unique BC1 breath controller is plugged into the mini-jack on the rear
panel.
The PHONES jack accepts any standard pair of stereo headphones. When a plug
is plugged into the B output only, the channel A and channel B signals are delivered
to the headphones in stereo. When the B output is not connected, both A and B
signals are mixed and output from PHONES jack. Headphone volume is controlled
via the VOLUME control on the top panel.
These terminals are used when connecting the DX21 to other MIDI (Musical Instrument Digital Interface) compatible equipment such as digital sequence recorders, modular FM voice generators, drum machines, etc. The MIDI OUT terminal
transmits MIDI data from the DX21 to other MIDI equipment. The MIDI OUT
terminal will normally be connected to the MIDI IN terminal of the receiving
equipment. The MIDI IN terminal accepts MIDI data from external MIDI equipment
such as a digital sequence recorder, music computer or modular FM voice generator.
THe DX21’s MIDI IN terminal will normally be connected to the MIDI OUT terminal
of the transmitting equipment. The MIDI THRU terminal re-transmits the data received at the MIDI IN terminal. Thus, data received via the DX21 MIDI IN terminal
can be simultaneously sent to other MIDI equipment.
The DIN connector end of the supplied cassette cable is plugged into the DX21
CASSETTE connector. The three plugs on the other end of the cable should be
connected to a cassette data recorder (the kind normally used with personal
computers, etc.) as follows:
RED
WHITE
BLACK
cassette deck microphone input.
cassette deck earphone output.
cassette deck remote input (The DX21 does not output cassette
remote control signals, so this jack may be left unconnected).
3
7. AC Power
8. Music Rack
lnstallation
Plug the DX21 AC power cord into an AC wall socket. Be sure that your local line
voltage matches that specified on the DX21 rear panel. The POWER switch is located
next to the AC cord on the rear panel. NOTE: When setting up your system, be
sure to turn the DX21 and any effects units used on BEFORE turning the main
amplifier system on. This will prevent the initial power-on shock surge from possible
damaging your amplifier and speaker system.
The music rack supplied with the DX21 fits into the two sockets on the rear panel
(see diagram).
4
CHAPTER II: PLAYING THE DX21
1. The 32 Voice RAM
Memory
The DX21 has 32 RAM (Random Access Memory) locations which contain voices
which may be immediately selected by pressing the appropriate memory button
while the DX21 is in one of the three normal play modes described below. The two
rows of 16 buttons- row A and row B-are the voice memory selectors. The row
A and B selectors correspond to DX21 tone generator channels A and B.
These same buttons actually serve a number of functions. In the FUNCTION and
EDIT modes--which are covered in later sections of this manual--they serve to select
the function or edit parameter to be programmed.
Since RAM memory can be written to as well as read from, it is possible to load
new voice data into the 32 available RAM memory locations. This can be done
by loading new voices from the DX21’s internal 128-voice ROM (Read Only
Memory) into the RAM memory (see “5. The 128 Voice ROM Memory”, below),
or by loading a new set of 32 voices from data stored on an external cassette tape
(see CHAPTER III: THE FUNCTION MODE, Memory Management Functions).
If you’re just starting out, you’ll want to try out the voices that are pre-programmed
into the RAM memory. You can do this in any of the three play modes discussed
below. Make sure your DX21 is properly connected to a sound system--or you
could simply plug a pair of headphones into the PHONES jack (see CHAPTER
I: SETTING UP), set the VOLUME control to about its center position, and set
the A/B BALANCE control to its center position (you’ll feel a “click” stop at dead
center). Now you can start experimenting with the DX21 voices in the SINGLE
play mode.
2. The SINGLE Play
Mode
Enter the SINGLE play mode by pressing the PLAY SINGLE button to the left of
the LCD (Liquid Crystal Display) panel. When you do this the LCD will look
something like this.
The top line of the LCD reads “PLAY SINGLE”, indicating that you are in the SINGLE
play mode, and the bottom display line will indicate the selected memory location
number (A1 through A16, or B1 through B16) followed by the name of the selected
voice.
In this mode you can play any of the voices currently in the DX21’s RAM memory
individually.
5
Enter the DUAL mode by pressing the PLAY DUAL button.
3. The DUAL Play
Mode
“DU” in the upper left display corner indicates that the DUAL mode has been selected. The upper display line now shows the number and name of any voice selected
from the “A” row of memory selectors, while the lower display line shows the number
and name of any voice selected from the “B” row of voice selectors. The number
in the lower left display corner indicates the value of the DUAL MODE DETUNE
(see CHAPTER III: THE FUNCTION MODE).
In this mode any of the 16 voices from tone generator “A” can be selected and
combined with any voice from tone generator “B”. As you try out various combinations of voices you’ll notice that some voices are louder than others, resulting
in one voice overpowering the other when two are combined. This can be adjusted
by using the A/B BALANCE control. Moving this control more towards the “A”
end of the scale increases the volume of the “A” voice while decreasing the volume
of the “B” voice. Moved in the opposite direction—towards “B”—it has the opposite
effect. Use this control to achieve the best balance between the two selected voices
in the DUAL play mode.
Enter the SPLIT mode by pressing the PLAY SPLIT button.
4. The SPLIT Play
Mode
“SP” in the upper left display corner indicates that the SPLIT mode has been selected.
As in the DUAL mode the number and name of one voice selected from the “A”
row will be shown on the upper display line, and the number and name of a voice
selected from the “B” row will be shown on the lower display line. Unlike the DUAL
mode, however, the two selected voices do not sound together. The “A” voice
will be assigned to all keys to the left of and including the programmed “split” key,
while the “B” voice will be assigned to all keys to the right of the split key. This
type of arrangement lets you assign a bass voice, for example, to the left side of
the keyboard and a piano voice to the right side of the keyboard so you can play
bass and piano lines simultaneously.
The next step is to decide on and program the desired split point. Note that at the
extreme left of the lower display line there is a number followed by a colon. This
is the number of the selected split point. To program the split point, simply press
the S POINT button (notice that this is the same as the PLAY SPLIT button you
used to enter the SPLIT mode) and while holding it down press the key on the
DX21 keyboard at which you want the split point to be programmed (C3=60).
6
SPLIT POINT set at 60
The split point number on the LCD will change to the number of the pressed key,
the “A” voice will be playable on all keys to its left (Including the split key), and
the “B” voice will be playable on all keys to the right of the split key. This process
can be repeated as many times as necessary, but you must press and hold the S
POINT key again every time you re-program the split point.
As in the DUAL mode, the A/B BALANCE control can be used to balance the volume
levels of the “A” and “B” voices.
5. The 128 Voice ROM
Memory
The DX21 comes with 128 different pre-programmed voices in an internal ROM
(Read Only Memory). These voices can be loaded in groups of 8, or singly, into
the DX21’s selectable 32-voice RAM memory. The ROM voices are organized into
16 different categories, each containing 8 voices (see next page).
To load these voices into the DX21's RAM voice memory for selection and use,
see A11: LOAD INTERNAL MEMORY and A12: LOAD SINGLE INTERNAL MEMORY in the following chapter.
7
THE 128 VOICES PROVIDED IN THE INTERNAL ROM
GROUP 1 : PIANO
1
GROUP 5 : BRASS
GROUP 9 : PERCUSSION 2
GROUP 13 : BASS
Electro Tom
1
2
Deep Grand
1
Horns
1
Brite Brass
2
Timpani
3
Breakin
3
Solid Bass
2
Uprt Piano
2
3
Honkey Tonk
3
4
Ivory Ebony
4
Flugel horn
Trombone
4
Xylo snare
4
Mono Bass
5
Phase Grand
5
Hard Brass
5
Synballs
5
Flap Bass
6
Elec Grand
6
Power Brass
6
Hand Drum
6
Elec Bass
7
Piano bells
7
BC1 Horns
7
Clock werks
7
Uprt Bass
8
BC1 Trumpet
8
Heifer Bell
8
Fretless
8
Acous Elec
GROUP 2 : ELECTRIC PIANO
GROUP 10 : LEADSYNTH
GROUP 6 : PLUCKED
Pluck Bass
Synthe Bass
GROUP 14 : SOUND EFFECT 1
1
Old Electra
1
Pluk guitar
1
Heavy synth
1
Racing Car
2
New Electra
2
Brt Guitar
2
Harmo solo
2
Helicopter
3
High Tines
3
Soft Harp
3
Feed Lead
3
Alarm Call
4
Hard Tines
4
Sitar
4
Mono Lead
4
Ghosties
5
Perco Piano
5
Jazz Guit
5
Hollow lead
5
Dopplar FX
6
Wood Piano
6
Old Banjo
6
Lyrisyn
6
Storm Wind
7
Vibra belle
7
Kotokoto
7
Schmooh
7
Space Talk
8
Piano brass
8
Folk Guit
8
Huff Talk
8
Birds
GROUP 3 : ORGAN
GROUP 11 : OTHER KEYBOARD GROUP 15 : SOUND EFFECT 2
GROUP 7 : COMPING
1
Diesel
Harpsi Hi
2
Hole in 1
3
<<Smash>>
1
Easy Synth
1
Harpsi low
1
C l i c k Organ
2
3
Jazz Organ
2
Easy Clav
2
Ham < n > Eggs
3
>> WOW <<
3
4
Drawbars
4
Metal Keys
4
Fuzz Clav
Clear Clav
4
FM SQUARE
5
Club Organ
5
Cheeky
5
Mute Clav
5
FM PULSE
6
<6 Tease>
6
6
Squeeze box
6
FM SAW TOOTH
7
Gentle Pipe
7
Rubber Band
Pick Pluck
7
Celeste
7
LFO NOISE
8
8 Full Ranks
8
S/H Synth
8
Circus time
8
PINK NOISE
GROUP 4 : STRINGS
1
Strings
GROUP 12 : WIND REED
GROUP 8 : PERCUSSION 1
1 Glocken
GROUP 16: SOUND EFFECT 3
1
Claranette
1
Bassoon
2
Synvox
Wind bells
2
Silk Cello
2
Hamarimba
2
3
Orchestra
3
Steel Drums
3
Pan Floot
3
Punk funk
4
Solo Violin
4
Tube Bells
4
Lead Reed
4
Whistling
Mono Sax
5
Voices
Flute wood
6
Zing Plop
<BC1>Sax
7
Valve Plop
8
Mars to ??
5
Box Cello
5
Temple gong
5
6
Rich string
6
Mamarimbs
6
7
5th String
7
Good Vibes
7
8
Pizzicato
8
Bells Bells
8
8
BC1 Hrmnca
CHAPTER III: THE FUNCTION MODE
The FUNCTION mode permits access to four groups of functions: tuning functions, MIDI functions, memory
management functions, and performance functions. In this chapter we’ll describe each of these functions; what
they do and how they are programmed.
1. Accessing the
FUNCTION Mode
The FUNCTION mode is accessed by pressing the FUNCTION button to the left
of the LCD display. Individual parameters to be programmed are then called by
pressing the appropriate voice selector button. Note that when the FUNCTION
mode is active pressing a voice selector button calls the corresponding FUNCTION
parameter—these are the functions printed in brown below each voice selector.
Note that there are two exceptions: the PB (Pitch Bend) MODE SET and KEY SHIFT
KEY SET buttons are NOT included among the voice selectors. These function
selectors are located immediately above the DATA ENTRY -1 and +1 switches.
When the FUNCTION mode is called, the LCD should look something like this.
The top line of the display will. read “FUNCTION CONTROL”, indicating that the
FUNCTION mode is active, and the lower display line will show the name of the
selected function and its current data. In the example above, the MASTER TUNE
function is called (press the Al button) and the data is currently set at 0.
2. Entering Function
Data
Once the desired function has been selected, its value can be altered using either
the linear DATA ENTRY control located to the left of the panel, or the adjacent
-1 /NO/OFF and +1 /YES/ON switches.
Moving the DATA ENTRY control away from you increases the value of the selected
parameter, and moving the control towards you decreases the data value. Pressing
the -1 switch decreases the value of the selected parameter by one (decrements),
and pressing the +1 switch increases the value by one (increments). While the
DATA ENTRY control is valuable for quickly approaching the desired value with
parameters that have a large data range, the +1 and -1 switches permit precise
step-wise location of a specific value. The switches are also easier to use with
parameters that only have two values–e.g. ON (1) or OFF (0). In some cases you
will be required to answer YES or NO to prompts which will appear on the LCD
display. The -1 /NO/OFF and +1 /YES/ON switches are used for this.
9
3. The Performance
Parameters
“Performance parameters” are programmable parameters which pertain mainly to
real-time performance effects such as how the pitch bend and modulation wheels
affect the sound. All performance parameters can be memorized for each voice.
If you wish that the edited data will remain in the memory, you have to store it
using STORE function (see CHAPTER V: VOICE PROGRAMMING, “4. Storing
Voice Data”).
B1: POLY/MONO
This function selects either the POLY or MONO note output mode.
There are two possible settings: POLY and MONO. In the POLY mode the note
output configuration depends on the currently selected play mode. When the play
mode is set to SINGLE up to 8 notes can be played simultaneously. When set to
SPLIT up to 4 notes may be simultaneously played on each side of the split point
(4 + 4). When set to DUAL up to 4 notes may be played simultaneously. In the
MONO mode, which can only be selected when the DX21 is in the SINGLE play
mode, the DX21 acts as a last-note-priority monophonic keyboard. In the SPLIT
mode it is possible to select one POLY voice and one MONO voice. In this case
up to seven notes can be played simultaneously on the POLY voice and one on
the MONO voice.
Once the POLY/MONO function is selected, subsequent presses on the B1 button
alternate between the POLY and MONO modes. The DATA ENTRY switches can
also be used: the -1 switch selects POLY and the +1 switch selects MONO.
B2: PITCH BEND RANGE
This function sets the pitch range of the pitch bend wheel located to the left of
the DX21 panel. The pitch bend wheel automatically centers at normal pitch. It then
may be moved upward (away from the player to raise the pitch, or moved downward
(toward the player) to lower the pitch by the specified amount.
The data range is from 0 to 12. At 0 the pitch bend wheel is off. Each increment
between 1 and 12 represents a semitone. Thus if this function is set to 12, maximum
travel of the pitch bend wheel produces a one-octave pitch variation, both above
and below center position–i.e. plus or minus one octave.
The DATA ENTRY control and -1/+1 switches can be used to enter data. Once
the PITCH BEND RANGE function is called, subsequent presses on the B2 button
increment the data value.
PB MODE: MODE SET
This function selector, located immediately above the DATA ENTRY -1 button,
offers a choice of three pitch bend wheel modes: Low, High and K-on. In the Low
mode, the pitch bend wheel affects only the lowest note played on the keyboard.
In other words, if a chord is played, the pitch bend wheel affects only the pitch
of the lowest note in the chord—this makes it possible to produce some interesting
effects. The High mode is just the opposite—only the highest note played will be
affected by the pitch bend wheel. In the K-on (Key on) mode, all notes played
are affected simultaneously by the pitch bend wheel, while notes sustained using
the sustain footswitch will not be affected.
The DATA ENTRY control and -1/+1 switches can be used to select the desired
mode.
B3: PORTAMENTO MODE
Two different portamento modes are available: Full Time Portamento and Fingered
Portamento. When the POLY/MONO function is set to POLY (button B1), only
the Full Time Portamento mode is accessible. In the MONO mode, you have a choice
between the Full Time and Fingered portamento modes.
(1) “Full Time Porta” (MONO and POLY modes): Aconventional portamento effect
in which portamento occurs whenever a new note is played.
10
(2) “Fingered Porta” (MONO mode only): Portamento only occurs if the previously
played note is held while the next note is played. This mode is useful in recreating the effect of guitar string bending techniques, wood bass slide effects,
etc. If you lift your hand off the DX21 keyboard between notes, there will be
no portamento effect.
Once the PORTAMENTO MODE function is called, subsequent presses on the
B3 button alternate between the two available portamento modes only if the MONO
note output mode is selected. The DATA ENTRY switches can also be used to
select the desired portamento mode.
B4: PORTAMENTO TIME
This function sets the speed of the portamento effect.
The data range is from 0 to 99. At 0 portamento is off. A setting of 99 produces
the longest portamento effect.
Data can be entered using the DATA ENTRY control and -1/+1 switches. Once
the PORTAMENTO TIME function has been called, subsequent presses on the
B4 button increment the data value.
B5: PORTAMENTO FOOT SWITCH
This function turns the rear-panel PORTAMENTO footswitch jack ON or OFF.
When OFF the PORTAMENTO footswitch will NOT function. When ON the
PORTAMENTO footswitch connected to the PORTAMENTO jack can be used to
turn the portamento effect ON or OFF. (accepts Yamaha FC-4 or FC-5 foot
switches)
Once the PORTAMENTO FOOT SW function has been called, subsequent presses
on the B5 button alternate between the ON and OFF settings. The DATA ENTRY
switches can also be used to turn this function ON and OFF.
B6: FOOT CONTROL VOLUME
This function sets the control range of a foot volume controller (Yamaha FC-7,
optional) connected to the rear-panel VOLUME jack.
The data range is from 0 to 99. At 0 the foot controller is OFF and will have no
effect on the sound. At 99 the foot controller can be used to adjust output volume
over the full range from maximum volume to zero volume. Intermediate settings
permit corresponding degrees of “expression” or “swell” control.
The DATA ENTRY control and -1/+1 switches can be used to enter the data for
this function. Once the FOOT CONTROL VOLUME function is called, subsequent
presses on the B6 button increment the data value.
B7: FOOT CONTROL SUSTAIN
This function turns the sustain footswitch (Yamaha FC-4 or FC-5, optional)
connected to the rear-panel SUSTAIN jack ON or OFF. When ON the sustain pedal
can be used to control the sustain effect. When OFF the sustain pedal will not
function.
Once this function is called subsequent presses on the B7 button alternate between
the ON and OFF states. The DATA ENTRY switches can also be used to select the
desired mode.
B8: WHEEL RANGE (MODULATION WHEEL), PITCH
As you move the DX21 modulation wheel away from you an increasing amount
of LFO (Low Frequency Oscillator) modulation is applied to the selected voice.
LFO modulation can be made to modulate the pitch of the voice producing a range
of vibrato type effects. This function is used to set the maximum depth of pitch
modulation which can be applied using the modulation wheel. The actual effect
11
produced depends on the settings of the LFO parameters—these will be discussed
in CHAPTER V: VOICE PROGRAMMING. Note, however, that the appropriate
voice PITCH MODULATION SENSITIVITY parameter must be set to a value higher
than 0 for pitch modulation to be effective. The voice PITCH MODULATION
SENSITIVITY parameter will also be discussed in CHAPTER V. The data range is
from 0 to 99. At 0 pitch modulation is OFF, and rotating the modulation wheel
will cause no pitch modulation to be applied to the voice. A setting of 99 produces
the greatest possible pitch modulation depth.
Data is entered using the DATA ENTRY controls or switches. Once this function
is called, subsequent presses on the B8 button increment the data value.
B9: WHEEL RANGE (MODULATION WHEEL), AMPLITUDE
As you move the DX21 modulation wheel away from you an increasing amount
of LFO modulation is applied to the selected voice. LFO modulation can be made
to modulate the amplitude (level) of specified voice elements (operators), producing
a range of tremolo or timbre modulation (wow-wow) type effects. This function
is used to set the maximum depth of amplitude modulation which can be applied
using the modulation wheel. The actual effect produced depends on the settings
of the LFO parameters—these will be discussed in CHAPTER V: VOICE PROGRAMMING. Note, however, that the appropriate voice AMPLITUDE MODULATION SENSITIVITY parameter must be set to a value higher than 0 for amplitude
modulation to be effective. The voice AMPLITUDE MODULATION SENSITIVITY
parameter will also be discussed in CHAPTER V.
The data range is from 0 to 99. At 0 amplitude modulation is OFF, and rotating
the modulation wheel will cause no amplitude modulation to be applied to the voice.
A setting of 99 produces the greatest possible amplitude modulation depth.
Data is entered using the DATA ENTRY controls or switches. Once this function
is called, subsequent presses on the B9 button increment the data value.
The Yamaha Breath Controller
The optional Yamaha BC1 Breath Controller is a unique way of adding expression
to the sound as you play the DX21 keyboard. The BC1 is held in the mouth just
like the mouthpiece of a wind instrument. Blowing harder or softer into the BC1
mouthpiece then produces a corresponding effect. The Breath Controller can be
used to apply varying amounts of pitch or amplitude LFO modulation, just like the
modulation wheel. In addition it can be set up to directly affect pitch, amplitude
or timbre in response to breath pressure. Set to directly affect amplitude (EG BIAS),
for example, the Breath Controller can be used to apply realistic tonguing effects
to brass and other wind instrument sounds.
The four BREATH parameters listed below determine just how the Breath Controller
will affect the DX21 sound. These parameters may be set individually, or combined
for more complex effects.
B10: BREATH RANGE, PITCH
This function is used to set the maximum depth of LFO pitch modulation which
can be applied using the Breath Controller. The actual effect produced depends
on the settings of the LFO parameters—these will be discussed in CHAPTER V:
VOICE PROGRAMMING. Note, however, that the appropriate voice PITCH
MODULATION SENSITIVITY parameter must be set to a value higher than 0 for
pitch modulation to be effective. The voice PITCH MODULATION SENSITIVITY
parameter will be discussed in CHAPTER V: VOICE PROGRAMMING.
The data range is from 0 to 99. At 0 pitch modulation is OFF, and applying breath
pressure to the Breath Controller will cause no pitch modulation to be applied to
the voice. A setting of 99 produces the greatest possible pitch modulation depth.
12
Data is entered using the DATA ENTRY controls and -1/+1 switches. Once this
function is called, subsequent presses on the B10 button increment the data value.
B11: BREATH RANGE, AMPLITUDE
This function is used to set the maximum depth of LFO amplitude modulation which
can be applied using the Breath Controller. The actual effect produced depends
on the settings of the LFO parameters—these will be discussed in CHAPTER V:
VOICE PROGRAMMING. Note, however, that the appropriate voice AMPLITUDE
MODULATION SENSITIVITY parameter must be set to a value higher than 0 for
amplitude modulation to be effective. The voice AMPLITUDE MODULATION
SENSITIVITY parameter will be discussed in CHAPTER V: VOICE PROGRAMMING.
The data range is from 0 to 99. At 0 amplitude modulation is OFF, and applying
breath pressure to the Breath Controller will cause no amplitude modulation to
be applied to the voice. A setting of 99 produces the greatest possible amplitude
modulation depth.
Data is entered using the DATA ENTRY controls or switches. Once this function
is called, subsequent presses on the B11 button increment the data value.
B12: BREATH RANGE, PITCH BIAS
This function permits breath pressure applied to the BC1 Breath Controller to directly
control the pitch of the voice. In other words the LFO has no effect—only your
breath pressure directly affects the pitch of the voice.
The data range is from 0 to 99. At 50 pitch bias is OFF. A setting of 99 permits
the greatest pitch increase to be produced through the Breath Controller, while a
setting of 0 permits the greatest pitch decrease.
Data is entered using the DATA ENTRY controls and -1/+1 switches. Once this
function is called, subsequent presses on the B12 button increment the data value.
B13: BREATH RANGE, EG BIAS
This function permits breath pressure applied to the BC1 Breath Controller to directly
control the amplitude or timbre of the voice, according to settings of the corresponding voice parameters which will be covered in CHAPTER V. The LFO has
no effect—only your breath pressure directly affects the amplitude or timbre of the
voice.
The data range is from 0 to 99. At 0 EG bias is OFF. A setting of 99 permits the
greatest amplitude or timbre variation to be produced through the Breath Controller.
Data is entered using the DATA ENTRY controls and -1/+1 switches. Once this
function is called, subsequent presses on the B13 button increment the data value.
B14: CHORUS
The DX21 has a built-in chorus effect which can be used to add a pleasant stereo
chorus effect to any voice. Chorus is only effective, however, when both the DX21’s
A and B audio outputs are connected to a sound system in a stereo configuration.
The full impact of the chorus effect will not be heard if the A/MIX output only is
used, or if the A and B outputs are externally mixed to a mono signal.
This function turns the chorus effect ON or OFF. Once the CHORUS function is
called, subsequent presses on the B14 button alternate between the ON and OFF
states. The DATA ENTRY buttons can also be used to select the desired mode.
13
B15: TRANSPOSE
This function permits transposing the pitch of the entire DX21 keyboard up or down
two octaves in semitone steps.
The data range is from C1 to C5 (C3 is middle C, C2 is one octave lower than
middle C, C4 is one octave higher than middle C, etc.). Immediately after this
function is called the desired new key can be selected simply by pressing any note
on the keyboard between C1 and C5 (within a plus or minus two octave range
of middle C). The keyboard’s C3 key then assumes the pitch of the key played,
and all other keys are adjusted automatically. Pressing any key higher than C5
on the keyboard results in a transposition to C5. This method of data entry can
only be used once, immediately after the TRANSPOSE function is called. Further
alterations may be made using the DATA ENTRY control or switches.
KEY SHIFT: KEY SET
During any of the normal DX21 play modes, pressing the KEY SHIFT button instantly
transposes the pitch of the entire DX21 keyoard up or down to a key programmed
using this function. When KEY SHIFT is engaged the letter “K” will appear at the
upper left corner of the LCD display until the KEY SHIFT button is pressed again,
returning the keyboard to normal pitch.
Pressing this button in the FUNCTION mode permits programming the desired
pitch to be shifted to when the KEY SHIFT button is pressed during any of the
play modes.
The transpose range for the KEY SHIFT function is plus or minus two octaves. The
data range is from -24 to +24, with 0 corresponding to standard keyboard pitch.
Each increment corresponds to a shift in pitch of one semitone—a setting of 2 would
therefore raise the pitch of the entire keyboard a whole tone.
Immediately after calling the KEY SET function, data can be entered simply by
pressing a key on the keyboard within a plus/minus two octave range of C3 (middle
C). The pressed key then assumes the pitch of C3 and all other keys are adjusted
accordingly. Pressing the A2 key, for example, produces a setting of -3. Pressing
a key higher than C5 results in a +24 setting, and pressing a key lower than C1
results in a -24 setting. This method of data entry, however, can only be used once
after this function is called. Subsequent changes can be made using the DATA
ENTRY control and -1/+1 switches.
This group includes just two functions: MASTER TUNE and DUAL MODE DETUNE.
4. Tuning Functions
A1: MASTER TUNE
This is the DX21 MASTER TUNE function. All voices are affected simultaneously.
The programmable data range is from -63 to +63. When set to +0 the pitch of
the A3 key is the standard 440 Hz. At the lowest setting of -64 the overall pitch
of the keyboard is 100 cents (1 semitone) lower than standard pitch, and at the
highest setting of +63 the overall pitch of the keyboard is 100 cents higher than
standard pitch.
Use the DATA ENTRY control or -1/+1 switches to enter the data for this parameter.
Once the MASTER TUNE function is called, subsequent presses on the A1 button
increment the data value.
A2: DUAL MODE DETUNE
This function makes it possible to detune the pitches of channels A and B to create
chorus-type effects. DUAL MODE DETUNE only functions when the DUAL play
mode is selected.
The data range is from 0 to 99. At 0 no detune effect is produced. The greatest
(deepest) detune effect—a pitch difference between channels A and B of approx-
14
imately 1/4 semitone—is achieved with a setting of 99. The detune effect is produced
by simultaneously increasing the pitch of channel A while decreasing the pitch
of channel B.
Use the DATA ENTRY control or -1/+1 switches to enter the data for this parameter.
Once the DUAL MODE DETUNE function is called, subsequent presses on the
A2 button increment the data value.
5. Memory
Management
Functions
The memory management functions include functions for loading voices from the
DX21’s internal ROM voice memory, for storing and loading RAM memory voices
to and from cassette tape, initializing the voice memory, recalling voice data from
a special “safety” buffer memory, and turning the DX21 memory write protect
function ON and OFF.
A9: EDIT RECALL
In addition to the voice edit buffer, the DX21 has a special edit recall buffer memory
which maintains the last edited voice data. If, after editing or creating a new voice,
you inadvertently call new data into the voice edit buffer by pressing one of the
voice selector buttons before storing the edited voice data, the voice you had spent
so much time editing will be erased from the edit buffer. If only one error of this
type has been made, the edited data still resides in the edit recall buffer and can
be recalled into the voice edit buffer using this function.
When this function is initially called the LCD will read “Recall Edit ?". Confirm your
intention to recall the data into the voice edit buffer by pressing the DATA ENTRY
YES button. The DX21 will then reconfirm with “Are You Sure ?". Press the DATA
ENTRY YES button again to actually execute the edit recall function. The EDIT
mode will then be automatically entered, and the voice edit buffer will contain the
data called from the edit recall buffer. Pressing another function button, a PLAY
mode button or the EDIT mode button during the above process will abort the edit
recall function.
A10: INIT VOICE
This function sets all voice parameters in the voice edit buffer to their “initialized”
values, permitting voice programming from an effectively neutral set of values—a
“clean slate”.
15
INITIALIZED VOICE DATA LIST
When this function is called the LCD will read “Init. Voice ?“. Confirm your intention
to initialize the voice edit buffer by pressing the DATA ENTRY YES button. The
DX21 will then ask you to reconfirm your intention to initialize the voice with “Are
You Sure ?“. Press the DATA ENTRY YES button again to actually execute the
initialize operation. Once executed, the DX21 will automatically enter the EDIT
mode, ready for voice programming. Pressing another function button, a PLAY
mode button or the EDIT mode button prior to the final step in the above process
will abort the voice initialize function.
A11: LOAD INTERNAL MEMORY
To load a ROM voice group, first make sure that the MEMORY PROTECT function
is OFF (A16). Then enter the FUNCTION mode and press the A11 voice selector
button to call the LOAD INTERNAL MEMORY function. The LCD should now look
like this.
FUNCTION CONTROL
Group to Bank?
Confirm your intention to load a group of 8 ROM voices into RAM memory by
pressing the DATA ENTRY YES button. The top line of the LCD will now ask you
to select the group you wish to load—“Group (1–16) ?“–and the lower display
line will show the currently selected group. The 16 different ROM voice groups
are selected by pressing the A1 through A16 voice memory selectors.
Group (1-16)?
G1: PIANOS
16
Once you have selected the desired ROM memory group, press the DATA ENTRY
YES button to go on to the next step. If you press the NO button at this point the
display will revert to the “Group to Bank?” stage. If you do press the YES button,
the display will prompt you to select the group of 8 memory locations to which
you will load the selected ROM voice group—A1 through A8, A9 through A16,
B1 through B8, or B9 through B16. The desired memory location group is selected
by pressing the A1 through A4 buttons. The selected group will be displayed on
the lower line of the LCD.
Now the selection of the source voices and the destination memory locations is
complete, all you have to do to actually load the selected voices is press the DATA
ENTRY YES button. If you press the NO button at this point the display will revert
to the “Group to Bank?” stage. If you do press YES, the display will announce that
the load job has been accomplished with the “COMPLETED !” display. You can
now press any play mode selector to try out the newly loaded voices.
* * Be sure to turn the MEMORY PROTECT function back ON after the load operation.
A12: LOAD SINGLE INTERNAL MEMORY
This function lets you load one ROM voice into the DX21 voice edit buffer, from
which it can be stored in any of the 32 RAM memory locations using the STORE
function. Make sure the MEMORY PROTECT function (A16) is OFF before attempting to use the LOAD SINGLE INTERNAL MEMORY function.
When this function is first called the LCD will look like this.
Confirm your intention to load a single ROM voice into the voice edit buffer by
pressing the DATA ENTRY YES button. The upper line of the display will now show
the currently selected ROM voice group, and the lower line will show the voice
currently selected from that group.
The various voice groups can be selected by using the A1 through A16 buttons,
and the 8 voices within each group are individually selected using the B1 through
B8 buttons. When the desired voice has been selected in this way, simply hold
down the STORE button and press the memory selector button to which you want
to store the selected voice.
17
You can now try out the newly loaded voice.
** Be sure to turn the MEMORY PROTECT function back ON after the load operation.
A13: CASSETTE SAVE/VERIFY
This function actually incorporates two “sub-functions” —SAVE and VERIFY. After
calling this function, subsequent presses on the B13 button alternate between the
SAVE and VERIFY sub- functions. Normally, however, you will start with the SAVE
function which saves the entire contents of the DX21 RAM voice memory onto
cassette tape. The VERIFY function is then used to check the saved data against
the data still in RAM memory, to ensure that no errors occured in the SAVE process.
Before using this function, make sure that an appropriate data cassette recorder
is properly connected to the DX21 as described in CHAPTER I: SETTING UP.
When this function is initially called the LCD will read “Save to Tape ?”. Confirm
your intention to save the contents of the 32 voice memory locations to cassette
tape by pressing the DATA ENTRY YES button. The display will then read “from
Mem to Tape, all ready ?“. At this point reconfirm that the cassette recorder is set
up properly, make sure a suitable blank tape is loaded into the cassette recorder,
and start the recorder running in the RECORD mode. To begin the actual save
operation press the DATA ENTRY YES button. Pressing another function button,
a PLAY mode button or the EDIT mode button prior to the final step in the above
process will abort the CASSETTE SAVE function. As the DX21 saves each voice
to tape, the upper line of the display will indicate the voice number being saved.
When all 32 voices have been saved the DX21 will automatically go into the tape
verify mode, permitting you to check that the data was recorded properly.
Stop the cassette recorder. The LCD will now read “Verify Tape ?“. To verify, first
rewind the tape to the beginning of the file just saved, then press the DATA ENTRY
YES button in response to the “Verify Tape ?” display which should be showing.
the display will now read “Verify Tape, ready?“. Press the DATA ENTRY YES button
then start the cassette recorder running in the PLAY mode. The DX21 will now
read each voice from the cassette tape, and compare it with the corresponding voice
data in the RAM voice memory. If the cassette and RAM data matches, the display
will read “VERIFY COMPLETED”. Any PLAY mode can then be entered simply
by pressing the corresponding mode selector button.
If an error is encountered this fact will be displayed on the LCD. If this happens,
go back and try the SAVE process—followed by the VERIFY process—one more
time. If the error persists then you may have to carefully adjust the record and/or
playback level of the cassette recorder, or use a higher quality recorder or tape.
Be sure to check that all cassette connections are made properly.
18
A14: CASSETTE LOAD
This function loads a complete set of 32 voices from cassette tape into the DX21
RAM voice memory. Before using this function make sure that all cassette connections have been properly made, as described in CHAPTER I: SETTING UP.
Also make sure that the DX21 MEMORY PROTECT function (A16) is OFF.
When this function is initially called the LCD will read “Load from Tape ?". Confirm
your intention to load a complete set of 32 voices from cassette tape into the RAM
voice memory (remember, this will erase any voices previously in the RAM memory)
by pressing the DATA ENTRY YES button. The LCD will now read “from Tape to
Mem, all ready ?". At this point make sure the appropriate cassette tape is loadeed
into the recorder and is rewound to the beginning of the desired voice file. To execute
the load operation press the DATA ENTRY YES button again and start the cassette
recorder running in the PLAY mode. Pressing another function button, a PLAY
mode button or the EDIT mode button prior to the final step in the above process
will abort the CASSETTE LOAD function. The DX21 will indicate each voice on
the LCD as it is loaded. When the load function is completed the LCD will read
“LOAD COMPLETED”.
Stop the cassette recorder and turn the DX21 MEMORY PROTECT function back
ON. The LOAD function can be interrupted at any time by pressing the DATA ENTRY
NO button. This can, however, result in incomplete data loading, possibly causing
“garbled” voice data to appear in one memory location.
** Be sure to turn the MEMORY PROTECT function back ON after a successful
LOAD operation.
A15: CASSETTE LOAD SINGLE
This function allows you to load a single voice from a previously saved 32-voice
cassette file into the DX21 voice edit buffer, from which it can be stored in any
voice memory using the STORE function. Make sure the MEMORY PROTECT
function (A16) is OFF before using the LOAD SINGLE function.
When this function is initially called the LCD will read "Load Single ?". Press the
DATA ENTRY YES button to go on to the next step. The LCD will now read “Tape
# ? to Buff, (A1-A16,B1-B16)". You must now enter the voice number of the
voice you wish to load from cassette—do this by pressing the corresponding voice
memory seleactor button. The selected voice number will appear in place of, the “#
?” portion of the upper display line. Then, make sure the appropriate cassette is
loaded in the cassette recorder and is rewound to the beginning of the voice file
containing the desired voice, press the DATA ENTRY YES button, and start the
cassette recorder running in the PLAY mode. The DX21 will automatically locate
the selected voice and load it into the voice edit buffer. When finished, the display
will read “LOAD COMPLETED”. The LOAD SINGLE function can be interrupted
by pressing the DATA ENTRY NO button. Doing this may, however, result in garbled
voice data in the DX21 voice edit buffer.
The loaded voice may now be stored into any voice memory location by holding
down the STORE button while pressing the voice selector of the destination voice
memory location. This must be done before another voice selector button is pressed
if you desire to keep the loaded voice in memory, otherwise the loaded voice data
in the voice edit buffer will be erased and replaced by the voice data called by the
pressed voice selector button.
** Be sure to turn the MEMORY PROTECT function (A16) back ON after a successful LOAD operation.
19
A16: MEMORY PROTECT
This function turns the DX21 MEMORY PROTECT function ON or OFF. When
ON, the internal RAM voice memory cannot be altered using the STORE, INTERNAL
MEMORY LOAD, or CASSETTE LOAD functions. The MEMORY PROTECT
function does not affect the voice edit buffer, so the INIT VOICE, EDIT RECALL,
INTERNAL MEMORY LOAD SINGLE, and CASSETTE LOAD SINGLE functions
will operate whether MEMORY protect is ON or OFF.
When this function is called the LCD will read either "Mem Protect:ON" or "Mem
Protect:OFF", according to the current state of the MEMORY PROTECT function
(MEMORY PROTECT is automatically turned ON each time mains power to the
DX21 is turned ON). The DATA ENTRY ON and OFF buttons are used to turn
MEMORY PROTECT ON or OFF.
6. MIDI Functions
This group of functions deals with parameters which control the transmission and
reception of MIDI data via the DX21 MIDI OUT and MIDI IN terminals. This section
will describe how each parameter is programmed, while actual operational details
will depend entirely on the type of MIDI equipment with which the DX21 will be
used.
A3: MIDI ON-OFF
This function turns transmission and reception via the DX21’s MIDI terminals ON
or OFF. When ON the DX21 can transmit or receive MIDI data to or from external
MIDI equipment. When OFF, no MIDI data interchange is possible.
This function is turned ON or OFF using the DATA ENTRY buttons.
A4: RECEIVE CHANNEL
The DX21 is capable of receiving data on any of the 16 available MIDI channels,
or in the OMNI mode which enables reception on all channels simultaneously.
This function is used to set the desired MIDI receive channel or activate the OMNI
mode. The receive channel is normally set to match the transmission channel of
the MIDI equipment from which the DX21 will be receiving MIDI data.
Data is entered using the DATA ENTRY control or buttons, and the data range is
from "Midi Omni: ON” (in this state the OMNI mode is ON) to "Midi Recv Ch =
1" through "Midi Recv Ch = 16".
A5: TRANSMIT CHANNEL
The DX21 can transmit MIDI data on any of the 16 available MIDI channels. This
function is used to set the desired transmission channel. The transmission channels
will normally be set to match the receive channel of the MIDI equipment to which
the DX21 will be transmitting data.
Data is entered using the DATA ENTRY control or buttons, and the data range is
from 1 to 16.
A6: CHANNEL INFORMATION
This is the MIDI channel information “switch”. It turns transmission and reception
of all MIDI BASIC EVENT DATA and OTHER EVENT DATA (except for key event
data, which is always transmitted when the MIDI ON-OFF function is ON) ON
or OFF. This function is turned ON or OFF using the DATA ENTRY ON and OFF
buttons.
20
The data turned ON or OFF by this function consists of the following:
(a) KEY VELOCITY:
Initial touch response data. Although the DX21
does not have key velocity sensitivity (keyboard
touch response), its tone generator is capable of
receiving key velocity information from an external
keyboard which has this capability.
(b) MODULATION WHEEL:
Modulation wheel position.
(c) BREATH CONTROLLER: Breath controller pressure.
(d) PORTAMENTO FOOT SW: Portamento footswitch on/off status.
(e) PROGRAM CHANGE:
Voice memory selection.
(f) DATA ENTRY CONTROL: DATA ENTRY control position. Transmission only.
(g) DATA ENTRY SW:
DATA ENTRY +1 and -1 switch status. Transmission only.
Pitch bend, sustain switch and “all keys off” status data are transmitted whether
this function is ON or OF.
NOTE:
The above MIDI data is generally common to all keyboards and equipment
compatible with the MIDI system. Due to differences in the features provided
by some manufacturers, however, complete compatibility can not be guaranteed.
A7: SYSTEM INFORMATION
This function turns transmission and reception of MIDI SYSTEM EXCLUSIVE
INFORMATION data ON or OFF. The DATA ENTRY ON and OFF buttons are used
to turn this function ON or OFF. MIDI SYSTEM EXCLUSIVE INFORMATION
consists of voice parameters and related data which is transmitted by the BULK
TRANSMIT function (A8) described below. When this function is ON voice parameter changes made in the DX21 EDIT mode are transmitted in real time.
A8: BULK TRANSMIT
This function permits transmission of "32 Bulk Data”. That is, the data pertaining
to the 32 voices contained in the DX21 RAM voice memory. This data could be
transmitted, for example, to the RAM voice memory of another DX21.
When this function is called the LCD will read "Midi Transmit ?". Pressing the DATA
ENTRY YES button then begins actual transmission of the data. During transmission
the LCD will go blank. Transmission takes about two seconds.
NOTE:
System exclusive communication data is only useable with compatible Yamaha
MIDI equipment.
21
CHAPTER IV: THE PERFORMANCE MEMORY
1. Performance
Memory
Organization
2. Accessing the
Performance
Memory
In addition to the DX21’s 32 voice RAM memories and 128-voice ROM memory,
the DX21 also features a special 32 performance memory system. This performance
memory can be programmed with a voice number or pair of voice numbers (one
from each channel), play mode, split mode split key number or dual mode detune
data, key shift data, and pitch bend mode data. Thus, when one performance memory
is called a voice or combination of voices in any play mode is automatically called
along with the most important real-time performance parameters programmed for
that voice or voice combination.
To access the performance memory, hold down the FUNCTION key and press the
EDIT/COMPARE key. The LCD will look something like this:
The 32 performance memories are then selected by pressing the voice selector
buttons (A1 through A16 = 1 through 16, B1 through B16 = 17 through 32).
The first data on the top LCD line is the selected performance memory and its
performance name. This is followed by the programmed play mode: SI = SINGLE
play mode, DU = DUAL play mode, SP = SPLIT play mode. From left to right on
the lower display line we have the number of the voice or voices selected by this
performance memory, the split or detune data (this depends on whether the SPLIT
or DUAL play mode is selected), the key shift data, and the pitch bend mode.
3. Programming the
Performance
Memory
The performance memory is programmed quite simply by setting up the desired
voice or voices in the desired play mode, with the desired performance parameters.
Then, after making sure the MEMORY PROTECT function (A16) is OFF, access
the performance memory by holding down the FUNCTION button and pressing
the EDIT/COMPARE button. Now you can store the currently set up performance
combination in any memory location by holding down the STORE button and
pressing the performance memory selector button to which you wish to save the
performance data. Before you do this, however, you might want to give your
performance combination a name.
** Be sure to turn the MEMORY PROTECT function (A16) back ON after a successful storage operation.
4. The Performance
Name
The performance name can be programmed at any time while in the performance
mode as follows. Hold down the FUNCTION/CHARACTER button. You’ll notice
that a cursor begins flashing over the first letter of the current performance name.
While holding down the CHARACTER button you can enter characters by pressing
the appropriate character buttons (see CHAPTER V: VOICE PROGRAMMING,
B16: VOICE NAME). The available characters are printed in white on the right
of the voice selector buttons and the buttons to the left of the LCD. Pressing the
PLAY SINGLE button inputs a space. The cursor can be moved backwards or
forwards by pressing the DATA ENTRY +1 and -1 buttons. Once a performance
name has been entered the entire performance combination must be stored using
the STORE function, as described in "3. Programming the Performance Memory”,
above.
22
CHAPTER V: VOICE PROGRAMMING
1. The Basics Of FM
Synthesis
Before you begin actually programming or editing your own voices, a basic understanding of how digital FM synthesis works will be necessary. In the following
explanation we will discover how the DX21’s FM voice generator produces complex
voices. This information will help you to understand the process and make it easier
for you to create and edit your own voices.
OPERATORS
The Yamaha DX series FM digital synthesizers use pure sine waves that interact
to create the full harmonic spectrum for any voice. Each digital sine wave oscillator
is combined with its own envelope generator to form an “operator”.
Note that the operator’s oscillator has two inputs: one for the oscillator pitch data,
and one for modulation data.
CARRIERS AND MODULATORS
The DX21 voice generator has 4 operators. When the output of one operator is fed
to the modulation input of a second operator—i.e. the first operator modulates the
second—a whole spectrum of harmonics is created that can form an incredibly
diverse range of complex waveforms (including the more conventional triangle,
sawtooth, and square waveforms). All this from just two operators!
23
Operators do not have to be connected “vertically” in a modulator-carrier relationship, as shown above. The outputs of two operators can also be mixed—just
as the stops in an organ are mixed. In this case the sounds are simply added together
with no modulation effect.
ALGORITHMS
We have seen two different ways that two operators may be combined. The DX21
uses four operators, offering a lot of potential connection possibilities. These different configurations of operator relationships are called “algorithms”, and the DX21
offers 8 algorithm choices. These are all printed right on the DX21 panel. In the
algorithm diagrams on the panel the small boxes numbered 1 through 4 are the
operators.
HOW ALGORITHMS AFFECT THE SOUND
By changing the relative frequencies between operators in a modulator-carrier relationship, you change not only the fundamental pitch of the note, but also the
frequencies present in the harmonic structure. Thus, the timbre of the voice can
be precisely controlled. In addition, since each operator has its own envelope
generator (and a sophisticated one, too!), the harmonic structure of a note can
be programmed to vary over time, just as a plucked string changes its overtones
as the note decays.
Depending on the selected algorithm, operators can be stacked vertically, connected
horizontally, or both. In the vertical arrangement, when the output of one operator
is connected to the input of another the result is modulation. By convention, the
operator at the bottom—or “output” —of a stack of operators is known as a “carrier”.
All operators in the same stack above the carrier are “modulators”. By increasing
the output level of one or more modulators feeding a carrier, the number of harmonics
in the resultant sound is increased (its “bandwidth” is increased), making it more
brilliant.
24
The bottom operator of a
stack is called a CARRIER.
Any operators which are
connected to the carrier’s
inputs are called MODULATORS. YOU CAN’T HEAR A
MODULATOR DIRECTLY,
but you hear its effect by
listening to the carrier’s
altered (modulated) output.
A given operator can be
either a carrier or a modulator: the only difference is
how it is “connected” in a
given algorithm.
Most algorithms have multiple modulators and carriers. In one algorithm a given
operator may be a carrier, while in the next it might function as a modulator—the
only difference being how it is connected. In algorithm number 5, for example, there
are two vertical stacks of two operators, and the outputs of the carriers in these
stacks are connected in parallel (horizontally). Algorithm 5 has an equal number
of modulators and carriers—two modulators and two carriers.
On the other hand, all operators in algorithm 8 function as carriers. Note that no
modulation can occur in this algorithm (except for the feedback loop on operator
4—we’ll discuss that later). But algorithm 8 is ideal for creating rich organ voices—
think of each operator as different organ “stops”, which can be mixed together
as desired.
The algorithm alone, however, does not determine the actual sound of the voice.
The vital characteristics of the voice you create depend mostly on the frequencies
and levels you program into each operator. The 8 algorithms provided in the DX21
were specially selected because they offer the broadest range of voice programming
possibilities.
The results of using different frequency ratios, as well as different algorithms, are
shown graphically in the accompanying illustration. In the left column you see the
waveforms created by 1:1, 2:1 and 3:1 ratios between one modulator and one carrier.
In the right column you see the waveforms which result from the same three ratios,
but when the two operators used are both carriers (connected horizontally, this
is known as additive synthesis).
25
Still more variations can be achieved by changing the relative output levels between
operators; the greater the level of the modulating operator, the more harmonics
are present.
FEEDBACK
Note that every algorithm has one operator with a “feedback loop”—represented
by a line from the output of the operator which feeds back to the input of the same
operator. In effect, a feedback loop means that the operator is modulating itself.
While every algorithm has one feedback loop, feedback is not necessarily used in
every voice. One of the DX21 editing functions permits the feedback level to be
set between 0 (no feedback) and 7 (maximum feedback).
ENVELOPE GENERATORS
Consider what happens when you play a note on, say, an acoustic instrument. The
level of the sound initially goes up to some value, then eventually falls to nothing,
following some sort of pattern that is characteristic of the particular instrument
played. For example, a low note on a pipe organ starts slowly when you press a
key, because it takes a while for the large column of air within the pipe to build
up to maximum oscillation level, and takes a while to die down once the key is
released a note played on a wood block, on the other hand, starts quickly as the
mallet strikes the block, and stops quickly as the block stops resonating. The
characteristic volume pattern of any note played on any instrument is known as
its “volume envelope”. Most acoustic instruments also have a “timbre envelope”,
in which the harmonic content of the note changes (the timbre changes) from the
time the note is initiated to the time it decays.
26
Each of the 4 operators available in DX21 can be programmed with its own envelope.
The envelope applied to a carrier will, generally, contribute to the overall volume
envelope of the note, while an envelope applied to a modulator will contribute to
the timbre envelope of the note.
Here is a copy of the envelope diagram printed above the LCD on the DX21 panel.
This envelope diagram can be used as a guide in visualizing DX21 envelope settings
while you program or edit a voice.
Each envelope generator can be programmed with five different parameters: ATTACK RATE (AR), DECAY 1 RATE (D1R), DECAY 1 LEVEL (D1L), DECAY 2 RATE
(D2R), and RELEASE RATE (RR). The RATE parameters determine how fast the
envelope moves from one level to the next. The term LEVEL is used rather than
“volume” because the envelope of the operator you're working on could affect
volume or timbre, depending on whether it is a carrier or a modulator.
Any note begins at zero level when you press a key, then begins to approach
maximum EG level at a rate determined by the setting of AR (Attack Rate). The
envelope may reach maximum level instantly, or it may take over 30 seconds depending on the setting of AR.
When the envelope reaches maximum level it immediately begins moving towards
the next level in the envelope—D1L (Decay 1 Level)—at a speed determined by the
setting of D1R (Decay 1 Rate).
The change from maximum EG level to D1R can be either a decrease in level or a
sustain at maximum level, depending on the values you choose for D1L.
After reaching D1L, the envelope then begins to decay toward zero level at a speed
determined by the setting of the D2R (Decay 2 Rate) parameter. If D2R is set to
0 (no decay), however, the note will be sustained at D1L for as long as you hold
the key. Now, when you release the key you have been holding the envelope will
immediately begin to decay towards 0 level at a speed determined by RR (Release
Rate). In fact, at whatever point in the envelope you release the key the envelope
will immediately begin moving toward 0 level at the set Release Rate. Thus ends
the note envelope “history”.
THE PITCH EG
In addition to the EG functions just described, the DX21 also has a "PITCH EG".
This pitch envelope generator is programmed in a similar way to the individual
operator EGs, but it affects the overall pitch of the voice rather than level and timbre.
This is useful for creating certain types of voices-whistling, for example, in which
the pitch of the note usually starts out a bit low and then sweeps up to the proper
pitch.
27
2. The EDIT and
COMPARE Modes
To actually program or edit a voice you need to enter the EDIT mode. This is done
by pressing the EDIT/COMPARE switch in the group of switches to the left of the
LCD panel.
When the EDIT mode is activated the upper line of the LCD will indicate the currently
selected algorithm, the operator ON/OFF status (the group of four 1s or 0s) and
the currently selected operator. The lower disply line will indicate the currently
selected voice parameter. The last voice selected in the PLAY mode will be selected
for editing. The individual voice parameters are then selected by pressing the
corresponding voice selector switches—all voice parameters are printed in violet
above the voice selector switches. The selected parameter is then programmed using
the DATA ENTRY control and/or switches. The individual parameters will be described in detail below.
Once the EDIT mode has been called and any parameter change has been made,
a small letter “e” will appear in the top left corner of the LCD, indicating that editing
is in progress. You can play the DX21 keys and listen to how parameter changes
are affecting the voice as you edit. In many cases you will be editing an existing
voice and will want to compare the sound of the edited voice with the original
voice. This is done simply by pressing the EDIT/COMPARE switch again. The small
letter “e” in the top left corner of the LCD will change to a “c” indicating that the
COMPARE mode has been activated, and that the voice you will now hear is the
original voice before editing (the parameters displayed on the LCD will also revert
to those of the original voice). You can then return to the voice being edited by
pressing the EDIT/COMPARE switch again. This can be repeated as many times
as needed during the editing process. The COMPARE mode can only be entered
from the EDIT mode, however, after at least one data change has been made to
the original voice.
The EDIT/COMPARE mode can be exited by entering any PLAY mode or the
FUNCTION mode. Please note, however, that if you exit the EDIT/COMPARE mode
and then select a new voice ANY DATA YOU HAVE EDITED WILL BE ERASED!!!
This is because all editing is performed in a special edit buffer memory which is
the same memory to which a voice is called when its selector button is pressed.
To save edited data you must use the STORE function to save the new data in one
of the DX21’s 32 RAM voice memories. The STORE function will be discussed in
CHAPTER III: MEMORY MANAGEMENT. If you do make a mistake and lose the
edited data, the DX21 has been provided with a special temporary buffer memory
from which the lost data can be recalled (assuming only one error has been made)
using the EDIT RECALL function. The EDIT RECALL function will also be discussed
in CHAPTER III: MEMORY MANAGEMENT.
3. The Voice
Parameters
The following is a brief description of each available voice parameter, how it iS
programmed, and its effect. These parameters are selected by pressing the appropriately labelled (violet labels indicate voice parameters) panel switch while the
DX21 is in the EDIT mode.
PB MODE: OPERATOR SELECT
This switch (located immediately above the -1 DATA ENTRY button) selects the
operator to be worked on. Only one operator can be selected at a time. Only the
parameters for the selected operator will be displayed on the LCD panel.
In the EDIT mode the currently selected operator number is generally displayed
28
at the top right corner of the LCD: for example, “OP 3”. This only applies to parameters which can be individually programmed for each operator, however, when
parameters that affect all operators simultaneously are called (the LFO WAVE,
SPEED and DELAY parameters, for example) the current operator display will disappear from the LCD and individual operators cannot be selected.
A1-A4: OPERATORlAMS ON-OFFIEG COPY, 1—4
Individually turns operators 1 through 4 ON or OFF. In many cases a voice will
not require all operators in an algorithm. Operators that are not needed should
be turned OFF while editing. Also, during the voice creation process, it is a good
idea to start with all operators OFF and then turn them ON one at a time as you
program and add them to the algorithm. The four digits immediately following the
algorithm number on the upper display line represent the four operators, 1 through
4, in order from left to right. When an operator is ON a “1” appears in the corresponding position, and when an operator is OFF a “0” appears in the corresponding
position. Each press on the A1 through A4 buttons alternately turns the corresponding operator ON and OFF.
When the AMPLITUDE MODULATION SENSITIVITY parameter is selected (A14),
these switches are used to determine to which operators the sensitivity setting
will apply.
When the EG COPY function is in use these switches are used to select the operator
to which the data from the currently selected operator will be copied. The parameters
which can be copied are those of ENVELOPE GENERATOR and KEYBOARD
SCALING.
A5: ALGORITHM
Permits selection of any of the 8 available algorithms. The desired algorithm number
is selected using either the DATA ENTRY control or switches.
A6: FEEDBACK
Feedback can be applied to one operator in each algorithm. Pressing this switch
permits setting the amount (level) of feedback which will be applied.
The feedback level range is from 0 to 7. At 0 feedback is OFF, and at 7 maximum
feedback is applied.
Data is entered via the DATA ENTRY control or switches.
A7: LFO WAVE
Permits selection of the low frequency oscillator waveform. The available waveforms
are SAW UP (a rising sawtooth waveform), SQUARE, TRIANGLE, and S/HOLD
(sample and hold).
These waveforms are selected using the DATA ENTRY control or switches.
29
A8: LFO SPEED
Permits setting the speed of the low frequency oscillator. The data range is from
0 to 99. 0 corresponds to the slowest LFO speed (0.06 Hz), and 99 corresponds
to the fastest LFO speed (50 Hz).
A9: LFO DELAY
Permits setting a delay of between 0 and approximately 3 seconds before the LFO
modulation effect begins after a key is played. This is particularly useful for simulating
brass instruments, human voice, etc., in which the a vibrato effect is gradually applied
after the note has been initiated.
The data range is from 0 to 99. At 0 there is no delay, and a setting of 99 produces
a delay of approximately 15 seconds. With longer delay settings the modulation
effect begins gradually for a remarkably natural sound.
A10: LFO PMD
This parameter sets the depth of pitch variation produced by LFO modulation for
all operators simultaneously. This function is independent from pitch modulation
produced by the modulation wheel and breath controller, and is always ON once
set.
The data range is from 0 to 99. At 0 pitch modulation is OFF, and a setting of 99
produces the greatest pitch variation, dependent on the PITCH MODULATION
SENSITIVITY setting, described below (A13). When the PITCH MODULATION
SENSITIVITY parameter is set to maximum (7), the maximum pitch variation (PM
DEPTH = 99) will be plus and minus 1 octave.
Even if this parameter is set to 0, pitch modulation can still be applied via the
modulation wheel or breath controller.
A11: LFO AMD
This parameter sets the depth of amplitude variation (tremolo or wow) produced
by LFO modulation for all operators simultaneously. This function is independent
from amplitude modulation produced by the modulation wheel or breath controller,
and is always ON once set.
The data range is from 0 to 99. At 0 amplitude modulation is OFF, and a setting
of 99 produces the greatest variation, dependent on the AMPLITUDE MODULATION SENSITIVITY setting (A14). When the AMPLITUDE MODULATION
SENSITIVITY parameter is set to maximum (3), the maximum amplitude variation
(AM DEPTH = 99) will be plus and minus 42 dB peak-to-peak.
Even if this parameter is set to 0, amplitude modulation can still be applied via the
modulation wheel and breath controller.
30
A12: LFO SYNC
The beginning of the LFO cycle is normally synchronized with key-on timing. This
parameter permits turning this synchronization ON or OFF. All operators are affected
simultaneously.
When this parameter is ON, the LFO cycle always begins from the peak of a positive
half-cycle (90 degrees phase angle) when a key is played. This produces a clear,
consistent attack on all notes.
When LFO KEY SYNC is OFF, the LFO cycle starts from a random point when a
key is played. This is the ideal setting when the LFO is being used to create natural-sounding chorus or phasing type effects.
NOTE:
When in DUAL or SPLIT mode, the “B” voice is affected by the LFO data
of the “A” voice.”
A13: PITCH MODULATION SENSITIVITY
This parameter sets the sensitivity of all operators to pitch modulation applied either
via the LFO PMD parameter, above, or via the modulation wheel of breath controller.
The data range is from 0 to 7. At 0 no pitch modulation can be applied, and at 7
the maximum pitch modulation can be achieved. When LFO PMD, above, is set
to 99 a setting of 7 produces a plus and minus one-octave pitch variation.
A14: AMPLITUDE MODULATION SENSITIVITY
This sets the operator’s sensitivity to LFO and EG BIAS effects applied via the LFO
PMD or AMD functions, or via the modulation wheel or breath controller.
The application of LFO modulation to a carrier results in tremolo, and applied to
a modulator the result is a periodic variation in timbre which is similar to wow effects.
When EG BIAS is applied to a carrier via the breath controller the result is volume
(expression) control, and applied to a modulator the result is brilliance control.
The data range is from 0 to 3. At 0 amplitude modulation sensitivity is OFF and
no LFO or EG BIAS effects can be applied to the selected operators. A setting of
3 produces maximum sensitivity and therefore maximum effect depth.
The operators to which modulation sensitivity is to be applied are selected using
buttons A1 through A4. The four digits—1 or 0—at the top right corner of the LCD
correspond to operators 1 through 4. When an operator is turned ON—i.e. able to
receive amplitude modulation—the corresponding digit will be a “1”. When OFF
the corresponding digit will be a “0”. The operators are turned ON and OFF alternately each time the corresponding OPERATOR/AMS ON-OFF/EG COPY button
is pressed.
A15: EC BIAS
This sets the operator’s sensitivity to EG BIAS effects applied via the breath controller. EG bias changes the overall output level from the operator. The harder you
blow into the breath controller, the higher the maximum envelope level. When EG
31
BIAS is applied to a carrier via the breath controller the result is volume (expression)
control, and applied to a modulator the result is brilliance control.
The data range is from 0 to 7. At 0 EG BIAS sensitivity is OFF and no EG BIAS
effects can be applied to the selected operators. A setting of 7 produces maximum
sensitivity and therefore maximum effect depth.
A16: KEY VELOCITY
While the DX21 has no key velocity sensitivity of its own, its voice generators will
accept key velocity data from an external MIDI controller keyboard which does
have this feature. This function determines the sensitivity of each operator to
keyboard velocity sensitivity data from an external keyboard connected to the DX21
MIDI IN terminal (key velocity sensitivity = the harder you play a key, the louder
the note. Timbre variations are produced when keyboard sensitivity is applied to
a modulator).
The data range is from 0 to 7. At 0 key velocity sensitivity for the selected operator
is OFF. A setting of 7 produces the highest sensitivity, and therefore the greatest
effect.
B1: FREQUENCY RATIO
These parameters determine the actual frequency of each operator. For operators
which function as carriers this determines the actual pitch of the sound produced.
For operators functioning as modulators this determines the harmonic spectrum
of the sound produced.
Each operator can be set to any of 64 different frequency ratios as follows:
DX21 OPERATOR FREQUENCY RATIOS
0.50
1.57
3.46
5.65
7.85
9.89
12.00
14.00
15.70
19.03
22.49
0.71
1.73
4.00
6.00
8.00
10.00
12.11
14.10
16.96
19.78
23.55
0.78
2.00
4.24
6.28
8.48
10.38
12.56
14.13
17.27
20.41
24.22
0.87
2.82
4.71
6.92
8.65
10.99
12.72
15.00
17.30
20.76
25.95
1.00
3.00
5.00
7.00
9.00
11.00
13.00
15.55
18.37
21.20
1.41
3.14
5.19
7.07
9.42
11.30
13.84
15.57
18.84
21.98
These frequency ratios have been carefully chosen as the most useful for voice
programming. A ratio of 1.00 sets the selected operator to standard pitch—a pitch
of 440 Hz will be produced when the A3 (A above middle C) key is pressed. A
ratio of 0.50 produces a pitch one octave lower, and a ratio of 2.00 produces a
pitch one octave higher than standard pitch, and so on. The fractional ratios—1.73,
for example—produce extremely complex waveforms when combined with operators
set to other ratios, permitting the creation of an unlimited variety of sound effects
32
including extremely realistic bells, explosions, etc. More even ratios are more useful
for creating musical instrument sounds. It is possible to combine a modulator set
to a fractional ratio at a low operator level with even-ratio operators to add bite
to a string sound and many other effects.
The standard DX21 keyboard pitch is 8’, therefore in terms of footage: 0.50 = 16’,
1.00 = 8’. and 2.00 = 4’.
B2: DETUNE
This parameter permits slight detuning of the selected operator in relation to the
others, making it possible to create richer, fuller voice effects. If detune is applied
to carriers, the result is a thick, multi-instrument effect. Applied to modulators the
result is a slight periodic variation in timbre— sometimes similar to a phase shift
effect.
The data range is from -7 to +7, for a maximum detuning range of plus or minus
2 cents (1/600th of an octave). At +0 no detune effect is produced.
B3-B7: ENVELOPE GENERATOR, AR, D1R, D1L, D2R, RR
These switches select the specific envelope generator parameters to be worked
on: ATTACK RATE, DELAY 1 RATE, DELAY 1 LEVEL, DELAY 2 RATE and RELEASE
RATE.
The data range for AR, D1R, and D2R is 0 to 31, with 31 being the fastest rate
(instantaneous) and 0 the slowest. RR and D1L have a data range of from 0 to
15, with 15 being maximum level and 0 being zero level for D1L
The following EG curve shows the relationship between the RATE and LEVEL
parameters.
33
The following EG curves show the parameters for some common instruments.
When creating voices on the DX21 you should apply these EG curves to the carrier
operator.
B8: OPERATOR OUTPUT LEVEL
Permits setting the output level of the selected operator. The data range is from 0
to 99. At 0 the operator is OFF, and a setting of 99 produces maximum output level
from the selected operator.
Varying the output level of an operator which is functioning as a carrier results in
a change in the overall level of the sound contributed to the voice by that operator,
while varying the output level of an operator which is functioning as a modulator
results in a change in the harmonic spectrum produced by the carrier, thereby
changing the timbre of the sound.
Data is entered using the DATA ENTRY control or switches.
B9: RATE SCALING
This parameter makes it possible to gradually shorten the overall envelope length
(increase EG rate) as higher notes on the keyboard are played. This is particularly
useful for simulating the sound of stringed instruments such as piano or guitar, in
which the envelope of the higher notes is noticeably shorter than the lower notes.
The data range is from 0 to 3. At 0 RATE SCALING is minimum and at 3 the greatest
variation in envelope length is produced.
34
B10: LEVEL SCALING
This parameter produces a gradual decrease in note output level towards the right
(high) end of the keyboard. Many acoustic instruments exhibit a gradual falloff in
level towards the higher notes, and this function can be used to create the most
natural overall keyboard response for any particular voice.
The data range is from 0 to 99. At 0 no level scaling is applied. At 99 the deepest
keyboard level scaling effect is produced, and highest note on the keyboard may
be almost inaudible. Set this parameter for the most natural overall balance for the
particular voice.
B11-16: PITCH ENVELOPE GENERATOR, PR1, PL1, PR2, PL2, PR3, PL3
These switches select the specific pitch envelope generator parameters to be worked
on: PITCH RATE 1, PITCH LEVEL 1, PITCH RATE 2, PITCH LEVEL 2, PITCH RATE
3 and PITCH LEVEL 3. Note that the parameters are slightly different from those
of the volume envelope generator, and that there are six parameters instead of five.
The data range for all pitch envelope generator RATE and LEVEL parameters is 0
to 99. For the RATE parameters 99 is the fastest rate (instantaneous) and 0 the
slowest. The LEVEL parameters actually correspond to pitch. 50 is standard pitch,
99 is the highest pitch and 0 is the lowest pitch reachable with the pitch EG. The
following chart shows the relationship between the pitch EG rate and level parameters.
PL3 (Pitch Level 3) is the level (pitch) at which the note begins and finishes. PR1
(Pitch Rate 1) is the speed at which the pitch moves from PL3 to PL1 when a key
is pressed. PR2 is the speed at which the EG curve moves from PL1 to PL2. The
EG begins moving towards PL2 immediately after PL1 is reached. PL2 is the
“sustain” pitch. The pitch determined by PL2 will be maintained as long as a key
is held. Once the key is released the EG will begin moving back towards PL3 at a
rate determined by the setting of PR3. PR3 can be thought of as the “release” rate.
If all level parameters are set at 50 (standard pitch), the pitch envelope generator
will have no effect.
35
B16 (FUNCTION MODE): VOICE NAME
This function is accessed in the FUNCTION mode—press the FUNCTION button
and then the B16 (VOICE NAME) button.
The VOICE NAME function can be used to give voices in the voice memory an
identifying name. Voice names can be up to 10 characters in length.
When this function is called the flashing cursor will appear at the beginning of the
voice name entry space on the lower display line, ready for you to enter the first
character. The available characters are printed in white on the right side of the voice
selector buttons and the group of six buttons to the left of the LCD. To enter a
character in the FUNCTION: VOICE NAME mode, hold down the CHARACTER
button (this is the same as the FUNCTION button) and press the corresponding
character button. As each character is entered the flashing cursor moves one space
to the right, ready for the next character. A character entered over an existing
character will replace the original character. The PLAY SINGLE button can be used
to enter a space. To correct errors, the cursor may be moved to the left and right
using the - 1 /NO/OFF and +1 /YES/ON switches, respectively.
4. STORING Voice
Data
5. Two Approaches to
Creating Your Own
Voices
If you have edited any of the above voice parameters and wish to keep the edited
data, you must use the STORE function BEFORE PRESSING ANY VOICE SELECTOR AFTER EXITING THE EDIT OR FUNCTION MODES, and store the new
data in one of the 32 RAM voice memories. Of course, this means it is a good idea
to have a free memory location ready before you begin editing. If you are editing
voices that were originally called from the internal ROM memory, don’t worry, you
can recall the ROM memory voices any time you need to. If the RAM is full of original
voices, however, make sure any voices you want to keep have been saved to cassette
tape from which they can later be recalled.
To STORE a newly edited voice, first exit the EDIT mode by pressing the PLAY
SINGLE mode selector. Then, hold down the STORE button (this is the same as
the EG COPY button) and press the voice selector to which you want to save the
edited voice data. The MEMORY PROTECT function must be OFF before attempting
to use the STORE function.
There are two basic approaches which can be taken when programming voices
on the DX21. First, you can run through all the existing voices, choose one that
is close to the sound you wish to create, and then edit that voice to create your
own. Second, you can “initialize” the the voice edit buffer (see CHAPTER III:
MEMORY MANAGEMENT, 1. The Initialize Voice Function) setting all parameters
to their initial values and begin programming your voice from scratch.
The first method, editing a preset voice, is generally a much more efficient approach,
provided the sound you want is close to something that already exists. If, however,
you are looking for a unique voice that is totally unlike any of the available presets,
it is probably best to initialize and start from scratch.
If you choose to program a voice from scratch, you’ll need to have a clear memory
location (or one containing a voice that you either don’t want or have already backed
up on cassette tape), so that when you’ve completed programming the voice, you
can save it. Since all editing is done in the separate voice edit buffer, nothing is
erased while you are actually programming the voice. But when you save the new
voice whatever was in that memory location will be erased and replaced by the
new data.
36
GENERAL SPECIFICATIONS
Keyboard
61 keys, C1 ~ C6
Sound Source
FM Tone Generator: 4 operators x 2 (A,B), 8 Alogorithms
Simultaneous Note Output
Polyphonic: 8 notes (SINGLE), 4 notes (DUAL), 4 + 4 notes (SPLIT)
Monophonic: 1 note (SINGLE, DUAL), 1 + 1 note (SPLIT)
Internal Memory
128 ROM voice memory
32 RAM voice memory
32 RAM performance memory
Controls
PITCH BEND WHEEL, MODULATION WHEEL, VOLUME, BALANCE, DATA
ENTRY
Display
LCD (16 characters x 2 lines)
Connecting Terminals
OUTPUT MIX/A, OUTPUT B, PHONES, CASSETTE
Control Terminals
FOOTSWITCH (SUSTAIN, PORTAMENTO), FOOT CONTROLLER (VOLUME),
MIDI (IN, OUT, THRU)
Dimensions/Weight
909 (W) x 82 (H) x 270 (D) mm (35.8” x 3.2” x 10.6”)
8 kg (17.6 Ibs)
Power Requirements
U.S. & Canadian Models: 120 V, 50/60 Hz
General Models: 220 ~ 240 V, 50 Hz
Power Consumption
7 W
Standard Access
Music Stand, Cassette Cable
Optional Access
FC7 Foot Controller, FC4 or FC5 Footswitch, BC1 Breath Controller, LC-21S or
SC-21 Carrying Case, LG-21 Stand, MIDI-03 or MIDI-15 MIDI Cable
37
MIDI DATA FORMAT
1. Transmission
Conditions
2. Transmission Data
All MIDI data is transmitted when the MIDI ON/OFF function is ON. The MIDI
transmission channel is determined by the setting of the MIDI TRNS CH function.
2-1. CHANNEL INFORMATION
2-1-1. Channel Voice Message
(1) Key Off
Status
Note no.
Velocity
1000nnnn
0kkkkkkk
01000000
38
n=channel no.
k=36(C1) ~ 96(C6)
(2) Key On
Status
Note no.
Velocity
1000nnnn
0kkkkkkk
01000000
n=channel no.
k=36(C1) ~ 96(C6)
1011nnnn
0ccccccc
0vvvvvvv
n=channel no.
(3) Control Change
Status
Control no.
Control code
a) Transmitted whether MIDI CH INFO is ON or OFF
Control no.
C=64: Sustain
Control code.
V=0:OFF, 127:ON
SW.
b) Transmitted when MIDI CH INFO is ON
Control No.
C=1 : modulation wheel
C=2: breath control
C=6: data entry slider
C=7: foot volume
C=65: portamento SW.
Control code
V=0~127
V=0~127
V=0~127
V=0~127
V=0:OFF, 127:ON
(4) Program Change
1100nnnn
0ppppppp
Status
Program no.
n=channel no.
p=0~31
This data is transmitted when a voice selector is pressed during the play mode,
when MIDI CH INFO is ON and MIDI SYS INFO is OFF. Also transmitted when
a performance selector is pressed in the performance mode.
(5) Pitch Blend
1110nnnn
0uuuuuuu
0vvvvvvv
Status
Code (LSB)
Code (MSB)
n=channel no.
The transmitted data is as follows:
LSB
00000000
00000000
01111110
MSB
00000000
01000000
01111111
Lowest value
Center value
Highest value
2-2. SYSTEM INFORMATION
2-2-1. System Real-time Message
Active sensing
Status
11111110
Transmitted once approximately every 200 milliseconds
2-2-2. System Exclusive Message
Transmitted only when MIDI SYS INFO is ON
39
(1) Parameter Change
Status
ID no.
Substatus/ch. no.
Parameter group no.
Perameter no.
Data
EOX
11110000
01000011
0001nnnn
00010010
0ppppppp
0ddddddd
11110111
n=channel no.
This data is transmitted when voice or function parameters are changed in the EDIT
or FUNCTION modes. The voice parameters transmitted are those given in voice
parameter table 5-2, and the function parameters transmitted are shown in function
parameter table 5-3.
(2) 1 Voice Bulk Data
Status
ID no.
Substatus/ch. no.
Format no.
Byte count
Byte count
Data
11110000
01000011
0000nnnn
00000011
00000000
01011101
0ddddddd
Checksum
EOX
0ddddddd
0eeeeeee
11110111
n=channel no.
93 bytes
The data for one voice is transmitted when a voice selector is pressed in the PLAY
SINGLE mode. Data in the voice edit buffer is transmitted when a format no. f=3
dump request is received. The transmitted data is showm in voice parameter table
5-2. The check sum is the lowest 7 bits of the two’s complement sum of all data
bytes (the same applies below).
(3) 32 Voice Bluk Data
11110000
01000011
0000nnnn
00000100
00100000
00000000
0ddddddd
Status
ID no.
Substatus/ch. no.
Format no.
Byte count
Byte count
Data
n=channel no.
4096 bytes
0ddddddd
0eeeeeee
11110111
Checksum
EOX
The data of all 32 voices in RAM memory will be transmitted if the YES (or BULK)
key is pressed in response to the “MIDI Transmit?” display which appears when
the BULK key is pressed in the FUNCTION mode. The data for all 32 voices will
also be transmitted when a format no. f=4 dump request is received. The transmitted
data is shown in voice data table 5-1. 55 bytes of 0’s are added to the 73 bytes
in this table, so 128 bytes are transmitted for each voice. 4096 bytes are therefore
transmitted for all 32 voices.
40
3. Reception
Conditions
MIDI IN
41
4. Reception Data
All MIDI data is received when the MIDI ON/OFF function is ON. When a specific
MIDI receive channel has been selected using the MIDI RECV CH function, and
the OMNI mode is OFF, MIDI data will be received only on the specified channel.
MIDI data will be received on all channels when the OMNI mode is ON.
4-1. CHANNEL INFORMATION
4-1-1. Channel Voice Message
(1) Key off
Status
Note no.
Velocity
1000nnnn
0kkkkkkk
0vvvvvvv
n=channel no.
k=0(C-2) ~ 127 (G8)
v is ignored
1001nnnn
0kkkkkkk
0vvvvvvvv
n=channel no.
k=0(C-2) ~ 127(G8)
v=0: key off
v=1 ~127: key on
(2) Key On/Off
Status
Note no.
Velocity
The key on note level will vary according to the received velocity value. The range
of this instrument is C-1 to B6. If a higher or lower key number is received, it will
be output within the range limits. For example, received C7 through B7 data will
be output as notes in the C6 through B6 range.
(3) Control Change
1011nnnn
0ccccccc
0vvvvvvv
Status
Control no.
Control code
n=channel no.
a) Received whether MIDI CH INFO is ON or OFF
Control no.
C=64: Sustain SW.
Control no.
C=1: modulation wheel
C=2: breath control
C=5: portamento time
C=6: data entry slider
C=7: foot volume
C=65: portamento SW.
Control code
V=0:OFF, 127:ON
Control code
V=0~127
V=0~127
V=0~127
V=0~127
V=0~127
V=0:OFF, 127:ON
(4) Program Change
Status
Program no.
1100nnnn
0ppppppp
n=channel no.
Received only when MIDI CH INFO is ON. If received during the PLAY or PERFORMANCE mode the voice or performance number will be changed accordingly.
The lower five bits of the program no. contain the voice or performance number.
(5) Pitch Bend
1110nnnn
0uuuuuuu
0vvvvvvv
Status
Code (LSB)
Code (MSB)
42
n=channel no.
Function s only on MSB data:
MSB
00000000
01000000
01111111
4-1-2.
Lowest value
Center value
Highest value
Channel Mode Message
Status
1011nnnn
0ccccccc
0vvvvvvv
n=channel no.
a) Recived whether MIDI CH INFO is ON or OFF
C=123
V=0
All notes OFF
b) Received only when MIDI CH INFO is ON
C=126
C=127
V=1
V=0
MONO mode ON
POLY mode ON
4-2. SYSTEM INFORMATION
4-2-1. System Real-Time Message
Active sensing
Status
11111110
Sensing begins when this code is received once. If status and data are not received
within 300 milliseconds the MIDI receive buffer will be cleared and the currently
output note will be turned OFF.
4-2-2. System Exclusive Message
( 1 ) parameter Change (swithch mode)
Status
ID no.
Substatus/ch. no.
Parameter group no.
Switch no.
Data
EOX
11110000
01000011
0001nnnn
00001000
n=channel no.
0mmmmrnmm
0ddddddd
11110111
d=0:OFF, 1~127:ON
All panel swithces are controlled. The switch numbers are arranged as shown in
the illustration below. Receiced only when MIDI SYS INFO is ON.
(2) Parameter Change
The format is the same as the transmitted parameter change data. Received only
when MIDI SYS INFO is ON. Permits changing voice and function parameters
while the EDIT mode is active. It is also possible to change modes: PLAY, EDIT,
etc. The parameter no. and data received are shown in voice parameter table 5-2
and function parameter table 5-3.
43
(3) 1 Voice Bulk Data
Received only when MIDI SYS INFO is ON. The format is the same as for the
transmitted 1 voice bulk data. The 93 voice data bytes are read into the voice edit
buffer, replacing the current voice data. The 93 received data bytes are shown in
voice parameter table 5-2.
(4) 32 Voice Bulk Data
Received only when MIDI SYS INFO is ON. The format is the same as for the
transmitted 32 voice bulk data. This data can be received only when the MEMOTY
PROTECT function is OFF. The received voice data is stored in the 32 RAM voice
memory locations. The “MIDI RECEIVED!!” display appears to confirm complete
reception of 32 voice bulk data.
(5) Dump Request
Status
ID no.
Substatus/ch. no.
Format no.
EOX
11110000
01000011
0010nnnn
0fffffff
11110111
n=channel no.
f=3,4
Received only when MIDI SYS INFO is ON. When received the bulk data corresponding to the received format code will be dumped via MIDI OUT.
f=3:
f=4:
1 voice bulk data
32 voice bulk data
44
5. System Exclusive
Data
5-1. VOICE DATA (VMEM format)
Parameter no.
P
0
1
2
3
4
5
6
7
8
9
10
19
20
29
30
39
40
41
42
43
44
45
Parameter
ATTACK RATE
DECAY 1 RATE
DECAY 2 RATE
RELEASE RATE
DECAY 1 LEVEL
OP4
KEYBOARD SCALING LEVEL
AMPLITUDE MODULATION ENABLE/EG BIAS
SENSITIVITY/KEY VELOCITY
OUTPUT LEVEL
OSCILLATOR FREQUENCY
KEYBOARD SCALING RATE/DETUNE 1
SAME AS FOR OP4
OP2
SAME AS FOR OP4
OP3
SAME AS FOR OP4
OP 1
49
50
51
52
53
54
55
56
57
LFO SYNC/FEEDBACK LEVEL/ALGORITHM
LFO SPEED
LFO DELAY
PITCH MODULATION DEPTH
AMPLITUDE MODULATION DEPTH
PITCH MODULATION SENSITIVITY/AMPLITUDE
MODULATION SENSITIVITY/LFO WAVE
TRANSPOSE
PITCH BEND RANGE
CHORUS SWITCH/PLAY MODE/SUSTAIN FOOT
SWITCH/PORTAMENTO FOOT SWITH/PORTAMENTO
MODE
PORTAMENTO TIME
FOOT VOLUME
MODULATION WHEEL PITCH MODULATION RANGE
MODULATION WHEEL AMPLITUDE MODULATION RANGE
BREATH CONTROL PITCH MODULATION RANGE
BREATH CONTROL AMPLITUDE MODULATION RANGE
BREATH CONTROL PITCH BIAS RANGE
BREATH CONTROL EG BIAS RANGE
VOICE NAME 1
66
67
68
69
70
71
72
VOICE NAME 10
PITCH EG RATE 1
PITCH EG RATE 2
PITCH EG RATE 3
PITCH EG LEVEL 1
PITCH EG LEVEL 2
PITCH EG LEVEL 3
46
47
48
45
5-2. VOICE PARAMETERS (VCED format)
Parameter no.
P
0
1
2
3
4
5
6
7
8
9
10
11
12
Parameter
ATTACK RATE
DECAY 1 RATE
DECAY 2 RATE
RELEASE RATE
DECAY 1 LEVEL
KEYBOARD SCALING LEVEL
KEYBOARD SCALING RATE
EG BIAS SENSITIVITY
AMPLITUDE MODULATION ENABLE
OP4
KEY VELOCITY
OUTPUT LEVEL
OSCILLATOR FREQUENCY
DETUNE 1
LCD Display
Data
EG AR
EG D1R
EG D2R
EG RR
EG D1L
LEVEL SCALE
RATE SCALE
E BIAS SENS.
A MOD SENS.
KEY VELOCITY
OUTPUT LEVEL
FREQUENCY
DETUNE
0 ~31
0 ~31
0 ~31
0 ~15
0 ~15
0~99
0 ~ 3
0 ~ 7
0 , 1
0 ~ 7
0~99
0~63
0~7
13
SAME AS FOR OP4
OP2
SAME AS FOR OP4
OP3
SAME AS FOR OF4
OP1
25
26
38
39
51
52
ALGORITHM
ALGORITHM SELECT
0 ~ 7
53
54
55
56
57
FEEDBACK LEVEL
LFO SPEED
LFO DELAY
PITCH MODULATION DEPTH
AMPLITUDE MODULATION DEPTH
FEEDBACK
LFO SPEED
LFO DELAY
LFO PMD
LFO AMD
0 ~ 7
0~99
0~99
0~99
0~99
58
59
60
61
LFO SYNC
LFO WAVE
PETCH MODULATION SENSITIVITY
AMPLITUDE MODULATION SENSITIVITY
0 , 1
0 ~ 3
0~7
62
63
64
TRANSPOSE
PLAY MODE POLY/MONO
PITCH BEND RANGE
LFO SYNC
LFO WAVE
P MOD SENS.
A MOD SENS.
Middle C
Poly Mode
P Bend Range
65
66
67
68
PORTAMENTO MODE
PORTAMENTO TIME
FOOT VOLUME
SUSTAIN FOOT SWITCH
Full Time Porta
Porta Time
Foot Volume
Foot Sustain
69
70
71
72
73
74
75
76
77
PORTAMENTO FOOT SWITCH
CHORUS SWITCH
MODULATION WHEEL PITCH MODULATION RANGE
MODULATION WHEEL AMPLITUDE MODULATION RANGE
BREATH CONTROL PITCH MODULATION RANGE
BREATH CONTROL AMPLITUDE MODULATION RANGE
BREATH CONTROL PITCH BIAS RANGE
BREATH CONTROL EG BIAS RANGE
VOICE NAME
Foot Porta
Chorus
MW Pitch
MW Amplitude
BC Pitch
BC Amplitude
BC Pitch Bias
86
87
PITCH EG RATE 1
88
89
90
91
92
PITCH
PITCH
PITCH
PITCH
PITCH
PEG RATE 1
PEG RATE 2
PEG RATE 3
LEVEL 1
LEVEL 2
LEVEL 3
EG
EG
EG
EG
EG
RATE 2
RATE 3
LEVEL 1
LEVEL 2
LEVEL 3
46
BC EG Bias
0 ~ 7
0 ~ 48
0, 1
0 ~ 12
0 , 1
0~99
0~99
0 , 1
0 , 1
0 , 1
0~99
0~99
0~99
0 ~99
0 ~99
0~99
1 ASCII
10~99
0 ~99
0 ~99
0 ~99
0 ~99
0 ~99
0~99
Note
5-3. FUNCTION PARAMETERS
Parameter no.
P
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
Parameter
LCD Display
OPERATOR ENABLE/DISABLE
OPETATOR SELECT
EDIT MODE 1=ON
FUNCTION MODE 1=ON
STORE MODE 1=ON
PLAY DUAL MODE 1=ON
PLAY SPLIT MODE 1=ON
PLAY SINGLE MODE 1=ON
DUAL MODE DETUNE
SPLIT POINT
MASTER TUNE $40=CENTER
MIDI SWITCH 1=ON
MIDI CH INFO
OMNI 0=OFF 1=ON
MIDI TRANS CH
MIDI RECV CH 0=OMNl ON
MIDI SYS INFO
32 VOICE BULK DUMP
EDIT RECALL
INIT VOICE
FUNCTION CONTROL
MEMORY STORE
PLAY SINGLE
Dual Detune
Master Tune
Midi Switch
Midi is OFF!
Midi Omni
Midi Trns Ch
Midi Recv Ch
Midi Sy Info
Midi Transmit?
Recall Edit?
Init. Voice?
Mem Protect
Key Shift
Bend Mode
MEMORY PROTECT 1=ON
KEY SHIFT 24=CENTER
PITCH BEND MODE 1=ON
KEY SHIFT
COMPARE
PITCH BEND MODE
47
Data
0 ~15
0~3
0, 1
0,1
0, 1
0,1
0, 1
0, 1
0 ~ 99
0 ~ 127
0 ~ 127
0, 1
0, 1
0, 1
0 ~ 15
0 ~ 16
0,1
0,1
0, 1
0, 1
0,1
0 ~ 48
0,1
0, 1
0,1
0 ~2
Note
[ Digital Programmable Algorithm Synthesiser 3
Date : 3/9, 1985
Model DX21
MIDI Implementation Chart Version : 1 . 2
Transmitted
Recognized
Remarks
Function ...
Basic
Channel
Default
Changed
1 - 16
1-16
1-16
1-16
memorized
3
x
1, 2, 3, 4
POLY, MONO(M=1)
x
memorized
Mode
Default
Messages
Altered
**************
36 - 96
Note
Dumber : True voice **************
Velocity Note ON
Note OFF
After
Touch
x
x
Key' s
Ch's
Pitch Bender
1
2
9nH,v=64
8nH,v=64
0 - 127
0 - 127
o
x
x
x
x
o
o
*1
*1
o
o
o
*1
*1
x
6
7
o
o
x
o
o
64
65
o
*1
o
96
97
o
o
*1
*1
x
Prog
Change : True #
o 0-31
**************
o
System Exclusive
o
o
System : Song Pos
: Song Sel
Common : Tune
x
x
x
x
x
x
System
:Clock
Real Time :Commands
x
x
x
Aux :Local ON/OFF
:A11 Notes OFF
Mes- :Active Sense
sages:Reset
x
x
x
o
o
x
x
5
Control
v=1-127
x
7 bit resolution
0-12 semi
*1 Modulation wheel
*1 Breath control
*1 Portamento time
Data entry knob
*1 Foot volume
o
Change
o
x
Sustain foot sw
*1 Portamento f sw
o
*2
Data entry +1
Data entry -1
x
0-127
0-31
*2 Voice parameters
o (123,126,127)
Notes: All MIDI communications are enabled if MIDI switch is on.
*1 = transmit/receive if CH information switch is on.
*2 = transmit/receive if system information switch is on.
Mode 1 : OMNI O N , POLY
Mode 3 : OMNI OFF, POLY
Mode 2 : OMNI ON, MONO
Mode 4 : OMNI OFF, MONO
o : Yes
x : No
49
50
FCC CERTIFICATION (USA)
While the following statements are provided to comply
with FCC Regulations in the United States, the corrective
measures listed below are applicable worldwide.
This series of Yamaha combo equiptment uses frequencies that appear in the radio frequency range and
if installed in the immediate proximity of some types of
audio or video devices (within three meters), interference
may occur.
This series of Yamaha combo equipment has been
type tested and found the comply with the specifications
set for a class B computing device in accordance with
those specifications listed in subpart J of part 15 of the
FCC rules. Those rules are designed to provide a reasonable measure of protection against such interference.
However, this does not guarantee that interference will
not occur. If your combo equipment should be suspected
of causing interference with other electronic devices,
verification can be made by turning your combo equiptment off and on. If the interference continues when
your equipment is off, the equipment is not the source
of interference. If your equipment does appear to be
the source of the interference, you should try to correct
the situation by using one or more of the following
measures:
Relocate either the equipment or the electronic device
that is being affected by the interference.
Utilize power for the combo equipment and the
device being affected that are on different branch
(circuit breaker of fuse) circuits, of install AC line filters.
In the case or radio or TV interference, relocate the
antenna or, if the antenna lead-in is 300 ohm ribbon
lead, change the lead-in to co-axial type cable.
If these corrective measures do not produce satisfactory results, please contact your authorized Yamaha
combo equipment dealer for suggestions and/or corrective
measures. If you can not locate a authorized Yamaha
combo equipment dealer in your general area contact
the Combo Service Departments, Yamaha International,
6600 Orangethorpe Ave., Buena Park, CA 90620, USA.
If for any reason, you should need additional information relating to radio or TV interference, you may
find a booklet prepared by the Federal Communications
Commission helpful; “How to Identify and Resolve
Radio-W Interference Problems”. This booklet is
available from the U.S. Government Printing Office,
Washington D.C.
20402-Stock No. 004-000-00345-4.
YAMAHA
NIPPON GAKKI CO., LTD. HAMAMATSU, JAPAN
OMD-127-1 85 05 2.0 CR Printed in Japan
®
YAMAHA
DX21 OM
Yamaha Corporation of America
6607 Orangethorpe Avenue, P.O. Box 6600, Buena Park, CA 80622-6600
DX21 OM
9/10/98 29645