HINT You can rapid/y move through the values by holding down one of the VALUE buttons. You can move even more rapidly by holding down one button and then pressing and holding down the other. For example, to rapidly advance (increase) the value, hold down the VALUE button and simultaneously press and hold down the VALUE button.

Changing the Voice Bank

In the current Sound Module mode (XG), several banks of Voices are available. Each bank can contain up to 128 different Voices. Operation Use the SELECT buttons to move the arrow cursor to the left side of the instrument icon, as shown below.

Use the VALUE

buttons to change the Bank number.
Finally, use the SELECT buttons again to move the arrow cursor back to the right side of the instrument icon for Voice selection.
Selecting Voices From Your MIDI Keyboard
You can also select Voices remotely from the connected MIDI keyboard. Though the actual operation may differ depending on the keyboard used, the general procedure is the same. (Refer to the owners manual of your instrument for specific instructions.) Operation Make sure that your keyboard is set up to send Program Change messages. Use the panel controls to select a program on your keyboard. Generally, if everything has been set up properly, the Voice number and name on the MU50 will change, and will be the same number as the program number you selected on your keyboard.
Changing Some of the Settings Part Controls
You can make changes to each individual Part by using the Part controls. These are always displayed in the Play mode, giving you at-a-glance confirmation of the various basic settings of the MU50. Lets take a look at the Play display again:
Each of these settings can be made independently for each of the Parts. For example, each Part could have a different Volume setting, or a different Pan setting. Try going through the brief sections below and making some changes in the Part controls yourself.
Selecting another Part and changing its MIDI channel
Here well select Part 2 and change its MIDI channel to 1, so that it matches the MIDI channel of Part 1, This allows you to play the Voices of both Part 1 and Part 2 over MIDI channel 1.
Operation Use the PART buttons to select Part 2.

Use the SELECT below.

buttons to call up the Rcv CH display

Use the VALUE nel to 01.

buttons to change the MIDI Receive Chan-
Finally, use the SELECT buttons to move the arrow cursor back to the instrument icon (so that the Voice name is displayed), and play the MIDI keyboard again.
If both Parts 1 and 2 have been set to MIDI channel 1, both of their level meters should move together as you play. And, if the two Parts are set to different Voices, you should hear two different Voices sound at the same time. (To change the Voice for a Part, refer back to Selecting Voices above.) 21

This determines the Expression setting for the selected Part's Voice
Settings: Rnd (Random), L63 C R63 This determines the stereo position of the selected Parts Voice. A setting of Rnd (Random) randomly assigns the Voice to a pan position. This is useful when you want to have different Voices sound from different random parts of the stereo image.

Reverb Send (RevSend)

Range:
This determines the level of the selected Parts Voice that is sent to the Reverb effect. A value of 0 results in a completely dry Voice sound. NOTE Keep in mind that the Reverb effect must be properly enabled and set for this parameter to work as intended. (See page 86.)

Chorus Send (ChoSend)

Range: This determines the level of the selected Parts Voice that is sent to the Chorus effect. A value of 0 results in a completely dry Voice sound (no Chorus effect). NOTE Keep in mind that the Chorus effect must be properly enabled and set for this parameter to work as intended. (See page 88.)

Variation Send (VarSend)

Settings: off, on (when Variation Connection is set to INS); (when Variation Connection is set to SYS) This determines whether the selected Parts Voice is sent to the Variation effect or not. A setting of off results in no Variation effect being applied to the Voice.

Note Shift (NoteSft)

Range: -24 +24 semitones This determines the key transposition setting for the Parts Voice.

All Part Control

The All Part controls include: Device Number, Master Volume, Reverb Return, Chorus Return, Variation Return and Transpose. Keep in mind that these controls affect all Parts equally, and either add to or subtract from their individual values. For example, if Note Shift on one Part is set to -12, and Transpose (in All Part) is set to +12, that Parts pitch value will actually be 0 or normal.
Selecting All Part Control
To select All Part control, press both PART buttons simultaneously. (or press the EXIT button). (All appears in the PART section of the display.)

Editing in ALL Part

Operation Select the desired control for all Parts by using the SELECT buttons. Change the value of the selected control by using the VALUE buttons.

EG Parameters

1) Short Attack, Decay, Release times:
2) Long Attack, Decay, Release times:
Even though the key is held for the same length of time in both examples, the sound of the second example slowly reaches full volume and decays over a longer time. It also sustains longer after the key is released. Keep in mind that the EG parameters affect each other, and are affected by how long a note is held. For example, if Decay is set to a low value and the note is held for a long time, you may not be able to hear changes made to the Release parameter. EG Attack Time
Range: -64 +63 Default: 0 This determines the Attack Time of the EG, or how long it takes for the sound to reach full volume when a note is played. For the Filter, this determines how long it takes for the sound to be affected by the maximum Filter values. EG Decay Time
Range: -64 +63 Default: 0 This determines the Decay Time of the EG, or how rapidly the sound dies out as a note is held. For the Filter, this determines how long it takes for the Filter effect to die out. EG Release Time
Range: -64 +63 Default: 0 This determines the Release Time of the EG, or how long the sound sustains after a note is released. For the Filter, this determines how long the Filter effect continues after a note is released.

Pitch EG Parameters

The Pitch EG parameters determine how the pitch of a Parts Voice changes over time. This lets you produce subtle or pronounced pitch changes as a note is played. In the example Pitch EG settings below, a played note is gradually bent up to its normal pitch, and kept there as the note is held. When the note is released, the pitch rapidly climbs up.

Pitch PEGReleLvl

Normal

PEGInitLvl

NOTE The Pitch EG parameters day have little or no effect, depending on the particular Voice used and the settings made to the main EG parameters.
Pitch EG Initial Level (PEGInitLvl)
Range: -64 +63 Default: 0 This determines the initial pitch of the Parts Voice, when the note is first played. A setting of 0 corresponds to normal pitch.
Pitch EG Attack Time (PEGAtakTime)
Graphically shows Pitch EG Attack Time setting for each Part.
Selected Part and MIDI channel
Range: -64 +63 Default: 0
This determines the Attack Time of the Pitch EG, or how long it takes for the pitch to return to normal (from the pitch value set in Initial Level above).
Pitch EG Release Level (PEGReleLvl)
Graphically shows Pitch EG Release Level setting for each Part.
This determines the final pitch of the Parts Voice, or the pitch that is reached after the note is released. A setting of 0 corresponds to normal pitch. NOTE

The Pitch EG Release Level and Time parameters may have no effect if the Voice itself does not sustain after the note is released. (Short percussive Voices may fall into this category.) Also make sure that the main EG Release Time is set to an appropriate value so that the sound sustains.
Pitch EG Release Time (PEGReleTime)
Graphically shows Pitch EG Release Time setting for each Part.
Range: -64 +63 Default: 0 This determines the Release Time of the Pitch EG, or how long it takes for the pitch to change to the pitch value set in Release Level above.

Vibrato

Vibrato produces a quavering, vibrating sound in the Parts Voice, by regularly modulating the pitch. You can control the speed and depth of the Vibrato, as well as the time it takes before the Vibrato effect is applied. Vibrato Rate
Graphically shows Vibrato Rate setting for each Part.
Range: -64 +63 Default: 0 This determines the speed of the Vibrato effect. Higher values result in a faster Vibrato sound. Vibrato Depth
Graphically shows Vibrato Depth setting for each Part.
Range: -64 +63 Default: 0 This determines the depth of the Vibrato effect. Higher values result in a stronger, more pronounced Vibrato sound. Vibrato Delay
Graphically shows Vibrato Delay setting for each Part.
Range: -64 +63 Default: 0 This determines the delay in the onset of the Vibrato effect. Delay is effective especially on stringed instrument Voices. For example, violin players often use delayed vibrato, especially while playing long notes. The Delay parameter is useful in recreating this effect, producing a more natural, lifelike sound. Higher values result in a longer Delay time.

Others

The Others section of parameters contains miscellaneous controls, including those related to tuning, Part Mode, velocity, portamento, note range, etc. Detune
Graphically shows Detune setting for each Part.
Range: -12.8 +12.7 Default: 0.0 This determines the fine tuning of the Parts Voice.
HINT Detune could be used to slightly detune a Voice compared to the tuning of the rest of the Voices for a richer sound. It could also be used to detune two different Voices being played in unison. For example if two different Parts are set to the same MIDI channel (see page 36) and same Voice, a naturally thick chorusing effect can be obtained by slightly detuning each Voice in opposite directions here.

Part Mode

Height of bars indicates selected Part Mode setting for each Part. (A single bar indicates normal setting.) Settings: normal, drum, drumS1, drumS2 Default: normal (Parts 1 9, 11 16) drumS1 (Part 10) (When Sound Module mode is set to C/M or DOC, Part 10 is set to drumS1.) This determines the mode for the Part. A setting of normal allows selection of the normal instrument Voices. (See the SOUND LIST & MIDI DATA booklet.) The drum setting allows selection of the drum kits. (See the SOUND LIST & MIDI DATA booklet.) The drumS1, drumS2 settings are locations for storing specially programmed drum setups. These setups can be edited by using the Drum Setup controls in the Multi Edit mode. (See page 61.) The drum and drumS1, drumS2 settings are not available in the Performance mode. The Part Mode settings differ depending on the Sound Module mode selected, as described below. For XG mode: All settings described above are available. When normal is selected, any of the basic or the extended set of Voices can be used for the Part.

In the Performance mode, the MU50 performs as a four-Part tone generator, with all Parts controlled over a single MIDI channel. The Performance mode is so named because its ideally suited to live performance situations. It allows you to play four different Voices at the same time from your MIDI keyboard either in a layer, or in sophisticated keyboard and velocity splits. It also gives you comprehensive control over each of the four Parts for maximum performance flexibility. A total of 256 Performances are available: 32 x 4 Preset and 128 Internal.
NOTE Drum Parts are not available in the Performance mode.

Performance Play Mode

To set the Performance mode and call up Performance Play:

Use the VALUE ANCE).

button to select PERFORM (PERFORM-
Selecting a Performance and its Individual Parts
To call up a Preset or Internal Performance and select individual Parts of that Performance: From the Performance Play mode (see section above), press the PART buttons simultaneously, so that the All Part display is shown.
lndicates All Part display.
The display is automatically set to Ail Part when the Performance Play mode is selected.

Use the SELECT

use the VALUE (Pre) or Internal (Int). Preset bank:
buttons to select the Bank parameter, then
buttons to select the desired bank, Preset

Internal bank:

Use the SELECT buttons to select the Program Number parameter, then use the VALUE buttons to select the desired Performance number.
From the All Part display, press the PART buttons simultaneously, so that the Single Part display is shown.
Select the desired Part by using the PART
Select the desired Voice bank for the Part by using the SELECT buttons to select the Bank parameter, and using the VALUE buttons to change the bank number. Select the desired Voice for the Part by using the SELECT buttons to select the Program Number parameter, and using the VALUE buttons to change the Voice number. To return to the All Part display from the Single Part display, either press the PART buttons simultaneously, or press the EXIT button.

Performance Part Control

Single Part

Settings:

000, 001, 003, 006, 008, 012, 014, 016 020, 024, 025, 027, 028, 032 043, 045, 064 072, 096 101, SFX
This determines the bank number of the selected Parts Voice. Each bank contains 128 Voices. (Refer to the SOUND LIST & MIDI DATA booklet.
Settings: Rnd (Random), L63 C R63
This determines the stereo position of the selected Parts Voice. A setting of Rnd (Random) randomly assigns the Voice to a pan position. This is useful when you want to have different Voices sound from different random parts of the stereo image.

Reverb Send (Rev Send)

This determines the level of the selected Parts Voice that is sent to the Reverb effect. A value of 0 results in a completely dry Voice sound.
NOTE Keep in mind that the Reverb effect must be properly enabled and set for this parameter to work as intended. (See page 86.)

Chorus Send (Cho Send)

Range: 0 127
This determines the level of the selected Parts Voice that is sent to the Chorus effect. A value of 0 results in a completely dry Voice sound (no Chorus effect).
NOTE Keep in mind that the Chorus effect must be properly enabled and set for this parameter to work as intended. (See page 88.)
Variation Send (Var Send)
Settings: off, on (when Variation Connection is set to INS); (when Variation Connection is set to SYS)
This determines whether the selected Parts Voice is sent to the Variation effect or not. A setting of off or 0 results in no Variation effect being applied to the Voice.

-24 +24 semitones 71

This determines the key transposition setting for the Parts Voice.

All Part

System MIDI Channel (Sys CH)

Settings: 1 16, all

This determines the MIDI receive channel for all Parts of the Performance. When this is set to all, the Performance responds over any incoming MIDI channel.

Performance Bank

Settings: Pre (Preset), Int (Internal)
This determines the bank of Performance programs: Preset or Internal. Preset Performances are those that have been created and loaded at the factory; the Internal bank is reserved for user-created Performances.

Performance Number

Range: This determines the Performance number.
Performance Volume (Pfm Vol)
This determines the overall Volume of the Performance.
Performance Pan (Pfm Pan)

Range: L63 C R63

This determines the overall Pan position of the Performance.

Reverb Return (RevRtn)

Range: This determines the amount of Reverb return for the Performance in the overall mix.

Chorus Return (ChoRtn)

The Part menu contains the Filter, EG, Vibrato and Others parameters for the Performance.
The Filter section of Performance Edit parameters are the same as the corresponding parameters in the Multi Edit mode. (See page 43.)
The EG section of Performance Edit parameters are the same as the corresponding parameters in the Multi Edit mode. (See page 45.)
The Vibrato section of Performance Edit parameters are the same as the corresponding parameters in the Multi Edit mode. (See page 52.)
The Others section of Performance Edit parameters contains miscellaneous controls, including those related to tuning, velocity, portamento, note range, etc. With a few additional parameters, these are the same as the parameters in the Multi Edit mode.
NOTE The Part Edit parameter (available in Multi Edit) is not available in Performance Edit mode. No Drum Parts can be selected for a Performance.

Detune

Same as the corresponding parameter in the Multi Edit mode. (See page 52.)

Mono/Poly Mode

Same as the corresponding parameter in the Multi Edit mode. (See page 54.)
Velocity Sensitivity Depth (VelSensDpt)
Same as the corresponding parameter in the Multi Edit mode. (See page 55.)
Same as the corresponding parameter in the Multi Edit mode. (See page 56.)
Same as the corresponding parameter in the Multi Edit mode. (See page 57.)
Note Limit High (NoteLimitH)
Portamento Switch (PortametSw)
Same as the corresponding parameter in the Multi Edit mode. (See page 58.)
Portamento Time (PortametTm)
Same as the corresponding parameter in the Multi Edit mode. (See page 59.)
Velocity Limit High (VelLimitHi)

Dry Level

Same as the corresponding parameter in the Multi Edit mode. (See page 60.) This parameter is only available when the Variation Connection parameter is set to SYS. (See page 92.)
Modulation Wheel LFO Pitch Modulation Depth (MWLFO PMod)
Same as the corresponding parameter in the Multi Edit mode. (See page 60.)
Modulation Wheel LFO Filter Modulation Depth (MWLFO FMod)
This determines how widely the filter is modulated by the LFO (low frequency oscillator). This is generally controlled from a modulation wheel on a MIDI keyboard and, depending on the character of the sound, produces a swoosh filter sweep effect. The higher the value, the deeper the filter modulation, and hence, the more pronounced the filter sweep effect.

Basic Concepts

The XG format maintains the universality and compatibility of the MIDI and GM standards while significantly increasing the range of expressiveness. It is designed to ensure data continuity, and to provide equipment manufacturers with considerable flexibility in designing machines that satisfy its requirements. Specifically, the XG format does the following.
Enables production of extremely expressive sound data Significantly expands available voice types and variations Supports future compatibility of sound data among musical instruments, computers, and

other devices

Ensures that data will remain fully usable well into the future Supports standardized handling of new types of effects-inclusive data (such
data) The XG format is founded on the following three principles:

as karaoke

Compatibility Scalability Expandability Compatibility Any XG machine, regardless of model or manufacturer, will provide faithful reproduction of any XG sound data. Because the XG format maintains downward compatibility with the GM format, XG machines will also provide correct reproduction of GM sound data. Scalability Although the XG format provides detailed and extensive specification of voice sets and voice changes, it does not require XG machines to support the full range of functions. Designers are free to develop a wide range of products to meet various cost and performance objectives. Each XG machine will replay XG data in accordance with the machines level of sophistication. If a model does not support a variation voice, it will automatically play the corresponding basic voice instead. If a model includes a graphic equalizer, it can take full advantage of graphic equalizer functions so as to control frequency characteristics to best suit the musical genre being played from lively rock to soothing classical. Expandability The XG format remains open to enhancements and extensions that will allow it to remain in step with future product developments.
Additions to the GM format
The XG offers the following extensions to the GM format.
Voices: The GM format supports 128 voices. The XG format provides for Bank Select messages that significantly expand the number of voices supported. 1) Voice Extension by Bank-Select LSB Variations of basic GM voices are stored in banks. To select a variation, specify the desired bank by sending the appropriate Bank-Select LSB value. Each bank is associated with a specific type of variation, so that voices are easy to locate.
2) Bank-Select MSB adds an SFX bank The Bank-Select LSB method is not useful for extension of distinctive SFX voices that have no meaningful variation. For this reason the XG format supports a full SFX bank of extension effects, which you can select by sending a Bank-Select MSB value of 40H. Bank-Select MSB 7H, in contrast, can be used to set any channel to rhythm-part play.
Voice Modification: The XG format allows creation of extremely expressive control data

Adjusts the reverb send level. Effects Send Level 3 (chorus): 5DH
Cntrl# 93 Parameter Effect: 3 Depth Data Range 0 to 127 Default: 00H
Adjusts the chorus send level. Effects Send Level 4 (variation): 5EH
Cntrl# 94 Parameter Effect-4 Depth Data Range 0 to 127 Default: 00H
Adjusts the variation effect send level. Effective only if Variation Connection = System. Data Increment/Decrement (RPN): 60H/61H
Cntrl# Parameter RPN Increment RPN Decrement Data Range 0 to to 127
The data byte has no meaning. This message adds or subtracts 1 to/from the Pitchbend Sensitivity, Fine Tune, and Coarse Tune MSBs. Note that the Increment (Decrement) message will not change a setting that has already reached its maximum (minimum) value. (Incrementation or decrementation of the Fine Tune value never carries over to the Coarse Tune value.) NRPN (Non-registered parameter number) LSB/MSB: 62H/63H
Cntrl# Parameter NRPN LSB NRPN MSB Data Range 0 TO TO 127
If multipart parameter Rcv NRPN is OFF, the part ignores this message. First send the NPRN MSB and LSB to select the control parameter, then set the value by Data Entry. Once you have selected an NRPN on a given channel, the channel will apply subsequent Data Entry to the selected parameter. After making the necessary settings you should set RPN to Null to reduce the risk of operational errors.
The following NRPN values are supported. NRPN MSB 01H 01H 01H 01H 01H 01H 01H 01H 14H 15H 16H 17H 18H 19H 1AH 1CH 1DH 1EH 1FH LSB 08H 09H 0AH 20H 21H 63H 64H 66H rrH rrH rrH rrH rrH rrH rrH rrH rrH rrH rrH Data-entry MSB mmH mmH mmH mmH mmH mmH mmH mmH mmH mmH mmH mmH mmH mmH mmH mmH mmH mmH mmH Parameter Vibrato Rate Vibrato Depth Vibrato Delay Filter Cutoff Frequency Filter Resonance EG Attack Time EG Decay Time EG Release Drum Filter Cutoff Frequency Drum Filter Resonance Drum EG Attack Rate Drum EG Decay Rate Drum Pitch Fine Drum Pitch Coarse Drum Level Drum Pan Drum Reverb Send Level Drum Chorus Send Level Drum Variation Send Level Data Range mm:00H-40H-7FH(-64-0-+63) mm:00H-40H-7FH(-64-0-+63) mm:00H-40H-7FH(-64-0-+63) mm:00H-40H-7FH(-64-0-+63) mm:00H-40H-7FH(-64-0-+63) mm:00H-40H-7FH(-64-0-+63) mm:00H-40H-7FH(-64-0-+63) mm:00H-40H-7FH(-64-0-+63) mm:00H-40H-7FH(-64-0-+63) rr:drum instrument note number mm:00H-40H-7FH(-64-0-+63) rr:drum instrument note number mm:00H-40H-7FH(-64-0-+63) rr:drum instrument note number mm:00H-40H-7FH(-64-0-+63) rr:drum instrument note number mm:00H-40H-7FH(-64-0-+63) rr:drum instrument note number mm:00H-40H-7FH(-64-0-+63) rr:drum instrument note number mm:00H-7FH(0-Max) rr:drum instrument note number mm:00H-40H-7FH(Random,L-Ccntcr-R) rr:drum instrument note number mm:00H-7FH(0-Max) rr:drum instrument note number mm:00H-7FH(0-Max) n:drum instrument note number mm:00H-7FH(0-Max) rr:drum instrument note number
Note that LSB values 14H through IFH (drum parameters) arc effective only for rhythm parts. RPN (Registered parameter number) LSB/MSB: 64H/65H Cntrl# Parameter RPN LSB RPN MSB Data Range 0 to to 127 Default: 7F 7FH

If multipart parameter Rcv RPN is OFF, the part ignores this message. The following parameters are supported. NRPN LSB 00H Data-entry MSB MSB 00H mmH Parameter Pitchbend Sensitivity Data Range mm: 00H-40H-7FH (-64-0-+63) Default: 02H

LSB value is ignored.

Minimum operational range is 00H00H-0CH00H ( octave).
01H 02H 7FH 00H 00H 7FH mmH mmH Fine Tune Coarse Tune Null mm: 00H-40H-7FH (-64-0-+63) Default: 40 00H mm: 00H-40H-7FH (-64-0-+63) Default: 40 00H

5. Channel Mode Messages

All Sound Off: 78H
Cntrl# Parameter 120 Data Range 0
Switches off sound from all parts. Does not reset the settings established by Channel Messages. Reset All Controllers: 79H
Cntrl# Parameter 121 Data Range 0
Resets the following values: Pitchbend, Modulation, Expression, Sustain, Registered Parameter Number, parameter settings. Resets portamento control. Specifically, clears condition in which source (Portamento Control Message) has been received but target (new Key On) has not. Leaves the following values unchanged: Program Change, Bank Select MSB/LSB, Volume, Pan, Dry Send Level, Effects Send Levels 1 ~ 5, Pitchbend Sensitivity, Fine Tuning, Coarse Tuning. All Notes Off: 7BH
Cntrl# Parameter 123 Data Range 0
Switches off all currently on notes in all parts. Any notes being held by sustain or sostenuto continue to sound until sustain/sostenuto goes off. OMNI Off: 7CH
Cntrl# Parameter 124 Data Range 0
Same processing as for All Notes Off.

OMNI On: 7DH

Cntrl# Parameter 125 Data Range 0
Same processing as for All Notes Off. (OMNI ON operation not supported.) MONO: 7EH
Cntrl# Parameter 126 Mono Data Range 0 to 16
Generates All Sound Off operation. If the value of the third byte (mono number) is 0~16, the channel changes to Mode 4 (m=1). POLY: 7FH
Cntrl# Parameter 127 Data Range 0
Generates All Sound Off operation, and sets the channel to Mode 3.

6. Channel Aftertouch

Status: DnH The message is not effective under initial settings to enable use of this message, send system Exclusive to change the channel aftertouch (CAT) parameters. If multi part parameter Rcv CHANNEL AFTERTOUCH is Off, the part ignores this message.

7. Polyphonic Aftertouch

Status: AnH The message is not effective under initial settings. To enable use of this message, send system Exclusive to change the polyphonic aftertouch parameters. If multi part parameter Rcv POLYPHONIC AFTERTOUCH is Off, the part ignores this message. It is not necessary for the effect to apply to all note numbers (0 ~ 127).

8. Exclusive Messages

XG System On
F0H,43H.10H,4CH,00H,00H,7EH,00H,F7H 0001nnnn F1n 4C 7E 00 F7 Exclusive status YAMAHA ID Device Number Model ID Address High Address Mid Address Low Data End of Exclusive

The message switches SOUND MODULE MODE to XG and initializes the XG settings. The message requires approximately 50ms to execute, so sufficient time should be allowed before the next message is sent. GM System On
F0H,7EH,7FH,09H,01H,F7H F0 7E 7F F7 Exclusive status Universal Non-realtime ID Device ID Sub ID1 Sub ID2 End of Exclusive
Resets all data values to their defaults, with the exception of the MIDI Master Tuning value. MIDI Master Volume
F0H, 7FH,7FH, 04H,01H, 11H, mmH, F7H Ommmmmmm 11110111 F0 7F 7F 11 mm F7 Exclusive status Universal Realtime ID Device ID Sub ID1 Sub ID2 Master Volume LSB Master Volume MSB End of Exclusive
Changes volume of all channels. (Universal System Exclusive) MIDI Master Tuning
F0H,43H,1nH,27H,30H,00H,00H,mmH,llH,ccH,F7H OOOlnnnn Ommmmmmm 01111111 0ccccccc 11110111 F1n mm 11 cc F7 Exclusive status YAMAHA ID Device Number Model ID Sub ID2
Master Tune MSB Master Tune LSB Dont Care End of Exclusive
Changes tuning of all channels.

Parameter Change

OOOlnnnn 01001100 Oaaaaaaa Oaaaaaaa Oaaaaaaa Oddddddd Oddddddd 11110111 F1n 4C aa aa dd dd F7 Exclusive status YAMAHA ID Device Number Model ID Address High Address Mid Address Low Data Data End of Exclusive
Includes 2 or 4 bytes of data, depending on parameter size. The following eight types of parameter change are provided. 1) System Data parameter change 2) Multi Effect Data parameter change 3) Multi EQ Data parameter change 4) Multi Part Data parameter change 5) Drums Setup Data parameter change 6) System Information 7) Display Data parameter change 8) AD Part Data parameter change *6) System Information is sent in response to dump requests. Received parameter changes are ignored. System Exclusive messages are not accepted if Rcv System Exclusive is OFF. Bulk Dump
OOOOnnnn 01001100 Obbbbbbb Obbbbbbb Oaaaaaaa Oaaaaaaa Oaaaaaaa Oddddddd Oddddddd 0ccccccc 11110111 F0n 4C bb bb aa aa dd dd cc F7 Exclusive status YAMAHA ID Device Number Model ID Byte Count MSB Byte Count LSB Address High Address Mid Address Low Data Data Checksum End of Exclusive
For information about Address and Byte Count fields, refer to Table 2 on the SOUND LIST & MIDI DATA booklet. Checksum value is set such that the sum of Address, Byte Count, and Checksum has value zero in its seven least significant bits. No more than 512 bytes should be sent in a single transmission. If the Dump Request asks for more than 512 bytes, data should be sent in packets of 512 bytes or less, with at least 120msec between transmission of consecutive packets.

Parameter Request

OO11nnnn 01001100 Oaaaaaaa Oaaaaaaa Oaaaaaaa 11110111 F3n 4C aa aa aa F7 Exclusive status YAMAHA ID Device Number Model ID Address High Address Mid Address Low End of Exclusive

Notes are cut off or omitted.
Reverb, Chorus and/or Variation effects cannot be heard.
The input sound (keyboard, guitar, etc.) cannot be beard.

Error Messages

Battery Low!
The battery voltage (for internal memory backup) may be too low. Bring the unit to your local Yamaha dealer or any other authorized Yamaha service personnel.

Illegal Data!

A data error resulted during reception of MIDI messages. Try transmitting the data again, or turn the MU50 off and back on again.

MIDI Buffer Full!

Too much MIDI data is being received by the MU50 at one time. Reduce the amount of data being sent to the MU50.

HOST is OffLine!

This message appears when the host computer is not turned on, the connecting cable is not properly connected, or the sequencing software is not active.

SysEx Adrs ERROR!

The data of the received System Exclusive message is incorrect. Check the address of the message and try transmitting again.

SysEx Data ERROR!

The data of the received System Exclusive message is incorrect. Check the data of the message (as to whether it requires an MSB or LSB header) and try transmitting again.

SysEx Size ERROR!

The data of the received System Exclusive message is incorrect. Check the size of the message and try transmitting again.

Check Sum ERROR!

The checksum of the received System Exclusive message is incorrect. Check the checksum of the message and try transmitting again.
This Parameter isnt Excl Data
The selected parameter has no System Exclusive value and cannot be displayed with the Show Exclusive function.

No Parameter

The selected parameter for use with the Show Exclusive function does not exist as a valid parameter.

Rev CH is OFF!

The selected parameter for use with the Show Exclusive function cannot be converted to a MIDI message value since the Receive Channel for the Part is off. Set the Receive Channel to an appropriate value. 124

Specifications

Tone Generation Method
AWM2 (Advanced Wave Memory 2)
Maximum Simultaneous Polyphony

32-note

Sound Module Modes
XG (Extended General MIDI), TG300B, C/M, DOC, and Performance

Multi-timbral Capacity

16-Part (on 16 MIDI channels; with element reserve priority for later notes and dynamic Voice allocation)
Internal Voice/Program Structure
Normal Programs Total Voices. 737 XG mode. 480 TG300B mode. 579 C/M mode. 128 (Parts 1 9), 64 (Parts 11 16) DOC mode. 61 Drum Programs Total Programs.. 22 XG mode.. 11 TG300B mode.. 10 C/M mode.. 1 DOC mode. 1 Performance Programs Up to four Voices plus all effect settings can be memorized to a Performance. Preset Programs. 32 x 4 User Programs. 128

Effects

Five sections of multi-effects: Reverb (12 Types), Chorus (11 Types), Variation (42 Types)

Demo Song

1 (not editable, stored in ROM)

Display

Custom back-lit LCD

Controls

INPUT level control; POWER/VOL control; Mode select buttons: PLAY, UTIL (UTILITY), MODE, EDIT, EFFECT; other buttons: MUTE/SOLO, , Select VALUE ENTER, EXIT, PART

Jacks and Terminals

Front panel: INPUT jack (stereo l/4"), PHONES jack (stereo mini-pin) Rear panel: OUTPUT R, L/MONO jacks (Right, Left/Mono); DC IN jack; TO HOST terminal; HOST SELECT switch; MIDI IN, MIDI OUT, and MIDI THRU terminals

Computer/MIDI Interface

Direct connection to host computer port (RS-232C, RS-422) with optional cables (CCJ-PCl, CCJ-PCl, CCJ-MAC); MIDI terminals allow connection to MIDI sequencer or MIDI controller
Data Transfer (Baud) Rate
MIDI 31,250 bps (bits per second) Mac 31,250 bps PC-1 31,250 bps PC-2 38,400 bps

Power Supply

Yamaha PA-1207 AC Adaptor (included)

Dimensions (W x D x H)

220 x 210 x 44 mm (8-5/8" x 8-l/4" x 1-3/4")

Weight

1.3 kg (2 lbs., 14 oz.)

Included Accessories

Owners Manual, Yamaha PA-1207 AC Adaptor
* Specifications subject to change without notice.

Glossary

Assignable Controller 1 Certain functions on the LFO Abbreviation for low frequency oscillator,
MU50 (such as the Filter, Volume or Variation effect) can be changed in real time by controllers on a connected MIDI instrument. Assignable Controller 1 lets you determine which controller (for example: modulation wheel, breath controller, foot controller, etc.) is used for that purpose.

Format Music Data Production Recommendations
Revision history April 20, 1995 ver.1.03 release 1995 Yamaha Corporation of America, CBX Group, P.O. Box 6600, Buena Park, CA 90622
Introduction This document provides recommendations for creating music data compatible with instruments utilizing the Yamaha XG format. In order to create music data that takes the fullest advantage of the XG format, youll need to have a complete understanding of the XG Specifications. We therefore urge you to refer both to this document and to the XG Specifications as you are creating your data. Note that the XG Specifications includes a number of new features which are not defined in the GM System Level 1 Recommendations (MIDI 1.0 Recommended Practice RP-003) of the JMSC (Japanese MIDI Standards Committee, now AMEIthe Association of Musical Electronics Industry). These new features are marked *Non-GM1* in this document. Models currently compatible with the XG Specifications include: Yamaha MU50 (Mid-range reference module, utilizing XG Level 1) Yamaha DB50XG (Sound card daughterboard, utilizing XG Level 1) Yamaha MU80 (High-end reference module, utilizing XG Level 2)

CONTENTS

1. XG implementation levels 2. Data format 3. Channels 4. Simultaneous polyphony 5. Timing distribution 6. Voice selection 7. Effects 8. System Exclusive parameter changes 9. Fade in/out 10. Miscellaneous
1. XG implementation levels There are currently two levels of XG implementation: Level 1, as utilized by the Yamaha MU50 and DB50XG; and Level 2, as utilized by the Yamaha MU80. There are also plans for a future XG Level 3, with details to be announced. While most aspects of XG Levels 1 and 2 are the same, there are a number of important differences between the two, as follows: XG Level 1 XG Level 2 (MU50 / DB50XG) (MU80) --------------------------------------------------------------------------------------------------------------------Number of drums setups Two Four Number of onboard effects Three (Reverb, Chorus, Variation) 42 Five (Reverb, Chorus, Variation, Distortion, Graphic EQ) 49

Number of SFX voices*

Unless otherwise noted, all references in this document apply to both XG Level 1 and Level 2.
* For a complete listing of all Level 1 and Level 2 SFX voices, see section 6-3 in this document.
2. Data format 2-1. Format We recommend the use of Standard MIDI File format 0 (where the file contains a single multichannel track), as opposed to format 1 (where the file contains multiple tracks). 2-2. Tempo We recommend using tempo settings in the range of 32 - 250 bpm. 2-3. Resolution Any resolution (such as 96 ppq or 480 ppq) may be used. For the MIDI events listed below, be sure to allow appropriate spacing between each event (as described below) in order to ensure correct processing by the sequencer: Continuous Control Change data such as Pitchbend, Expression, Volume, and Modulation - If these messages occur in dense succession, tempo can sometimes become unstable, so use the Thin Out (or equivalent) function on your sequencer to reduce the data stream as much as possible. Wherever possible, leave intervals of at least 1/96 (=5/480) between these events. Between Bank Select MSB/LSB and Program Change - If these three events are placed at the same time clock, the sequencer may produce unexpected results. Always leave intervals of at least 1/96 (=5/480) between these events, and always send the Bank Select MSB first, the Bank Select LSB second, and the Program Change last. Between events such as RPN/NRPN Parameter Select, Data Entry MSB/LSB, and Reset Always leave intervals of at least 1/96 (=5/480) between these events. After a System Exclusive message - Always leave an interval of at least 1/96 (=5/480). We recommend the use of System Exclusive messages only within the setup data (see section 2-4 below), and not within the song data itself.

2-4. Setup Data Reserve one or more measures for tone generator setup data at the beginning of the song, and do not place any music data in this setup measure or measures. Use the minimum number of setup measures that are required, as dictated by the time signature and the amount of MIDI events necessary. Always observe the required spacing between each event as given in section 1-3, above. Taking into account the possibility that XG music data may be played back on a GM tone generator, be sure to include a GM System On message in the setup data. The GM System On message must be transmitted before the XG System On message. Allow empty intervals of at least 200 msec after the GM System On, and at least 50 msec after XG System On. Note that the reception of an XG System On message by an XG-compatible instrument causes all parameters to initialize to their factory defaults (see the instruments owners manual for details). Therefore, you need include in the setup data only those messages required to set parameters to values other than their factory defaults. Allow an empty interval of at least 200 msec after the setup data before the song data begins. We recommend the general order of events in the setup measure to be as follows: 1. SMF header chunk 2. SMF track chunk 3. SMF Meta-Events such as tempo, time signature, text events, copyright notice, sequence name, lyric, etc. 4. GM System On message 5. XG System On message 6. All required System Exclusive parameter changes (Effects selection, followed by Effects customization, followed by voice parameter customization) 7a. Bank Select messages (Control Change #0, followed by Control Change #32) for channel 1 7b. Program Change message for channel 1 7c. All required Control Change messages for channel 1 (except Control Change #0 and #32) 7d. All required RPN messages for channel 1 7e. All required NRPN messages for channel 1 8. Repeat 7a - 7e for channel 2, then channels 3. 16 You can utilize the setup data templates provided on the enclosed floppy disk. Be careful to observe the spacing requirements between MIDI events as given in section 2-3 above.
3. Channels 3-1. Number of Parts A maximum of 16 parts (16-way multitimbral) should be used, since this is the minimum requirement for Level 1 XG instruments (Level 2 XG instruments may have up to 32-way multitimbral capability). 3-2. Channel Assignment for Each Part Depending on the character of the song, assign channels 1 - 5 and channel 10 as follows: Minus-one music: Part MIDI Ch. Note ----------------------------------------------------------------------------------------------------------------------Rhythm 10 dedicated rhythm part Melody Sub-melody/ solo/ accompaniment Bass Pad Riff Band music: Part MIDI Ch. Note ----------------------------------------------------------------------------------------------------------------------Rhythm 10 dedicated rhythm part Melody Sub-melody/ solo/ accompaniment Bass Pad Riff 5 lead part solo part or important backing part bass part backing part with mostly sustained notes rhythmic backing part 5 main part to be muted during minus-one play secondary part to be muted during minus-one play bass part backing part with mostly sustained notes rhythmic backing part

Piano practice piece: Part MIDI Ch. ----------------------------------------------------------------------------------------------------------------------Rhythm 10 dedicated rhythm part Melody Sub-melody/ solo/ accompaniment Bass Pad Riff 5 right hand (1st) part left hand (2nd) part bass part backing part with mostly sustained notes rhythmic backing part
Remaining musical parts should be allocated to channels 6 - 9 and 11 - 16, with parts of greater musical importance assigned to lower-numbered channels. 3-3. Rhythm Channel Settings As specified in 3-2 above (and as per General MIDI), channel 10 is always dedicated to the rhythm part. *Non-GM1* If you wish to specify a secondary rhythm part (in addition to channel 10), use either the Bank Select MSB and LSB* or the SysEx parameter change Part Mode** to designate another channel as a rhythm sound. SysEx parameter changes can be used to create drum setup settings for a maximum of two parts, including channel 10. By default, ch.10 is DRUMS 1. (In SysEx parameter change multi-part Part Mode F0H, 43H, 10H, 4CH, 08H, nnH, 07H, ddH, F7H, Part Number nnH=09H, and Data ddH=02H.) For one drum part other than ch.10, use DRUMS 2. (F0H, 43H, 10H, 4CH, 08H, 08H, nnH, 07H, 03H, F7H.) DRUMS 3 and DRUMS 4 are provided by Level 2 XG instruments only, and should not be used. If you do not need to use drum setup to modify the contents of the kit, you can use the SysEx parameter change F0H, 43H, 10H, 4CH, 08H, nnH, 07H, 01H, F7H to set the Part Mode to DRUM. This allows you to specify three or more parts as rhythm sounds. * See the XG Specifications, #4. Control Change / Bank Select MSB/LSB ** See the XG Specifications, #7. System Exclusive Messages / Parameter Change, and Table 1-8
4. Simultaneous polyphony Always keep the song within 16 - 32 notes of simultaneous polyphony. Since some voices use two notes of the total polyphony (because they contain two elements), the maximum available number of simultaneous notes will depend on the type and combination of voices used. Avoid overlapping note on/offs of the same channel and note number. For example, in the following case, it cannot be definitely stated how note-on processing will take place: note on | note off | | ================= =================== | | note on | note off

ch.1, note#60 ch.1, note#60
5. Timing distribution Do not place messages such as Pitch Bend Change and Control Change, etc., in the same time clock. Since some sequencers rearrange the order of MIDI events that occur in a single time clock, the results may be unpredictable. Also, dense successions of events (even if they do not occur in the same time clock) can cause the tempo to be disrupted, or the hardware to respond in unexpected ways. *Non-GM1* Regarding Bank Select, the XG Specification says that As given in the MIDI specification, MSB + LSB + Program Change must be sent as a unit. Here, as a unit should not be understood to mean at the same time, but rather that other MIDI events may not occur between the MSB, LSB and program change. Hypothetically speaking, if these three messages were sent at the same time and the sequencer read the Program Change first and the MSB and LSB later, the voice would be selected from the previously-specified bank. For more information, refer to XG Specifications #4. Control Changes / Bank Select MSB/LSB. Note that the reception of Bank Select MSB/LSB does not immediately change the tone generator voice. The channel stores the received Bank Select MSB/LSB setting, but does not apply it until receipt of the next Program Change.
6. Voice selection *Non-GM1* 6-1. Voice Selection Procedure The XG voice selection procedure is as follows. Melody voice Bank Select MSB = 00H, use Bank Select LSB to specify the bank, then use Program Change to select the voice Bank Select MSB = 40H, Bank Select LSB = 00H (fixed), then use Program Change to select the voice Bank Select MSB = 7EH, Bank Select LSB = 00H (fixed), then use Program Change to select the kit, finally, use Note Number to select the voice Bank Select MSB = 7FH, Bank Select LSB = 00H (fixed), then use Program Change to select the kit, finally, use Note Number to select the voice

SFX voice

SFX kit

Rhythm Kit

For Kit and Voice mapping, refer to the XG Voice List and XG Drum List that are attached to the XG Specifications. In the SFX kit, there are some notes within the valid range of notes that produce no sound (that is, keys to which no voice is assigned). 6-2. Melody Voice Substitution Tone generators that do not have the specified bank when a new melody voice is selected will automatically substitute the sound of the previously selected bank. (By default, this will be the GM basic sound set in Bank 0.) As an example: If you want Bank Select MSB = 00H, LSB = 2DH, Program Change #29 Jazz Man but a substitution of MSB = 00H, LSB = 2BH, Program Change #29 Funk Gtr 2 is possible, first transmit MSB = 00H, LSB = 2BH, Program Change #29 and then transmit MSB = 00H, LSB = 2DH, Program Change #29 This slight detour will minimize the chances of an undesired replacement by the GM basic sound set.

6-3. SFX Voice Substitution Be aware that Level 1 XG instruments (such as the MU50 and DB50XG) do not contain all the Level 2 SFX voices (as provided by the MU80). Level 1 tone generators that do not have the specified SFX voice when a new SFX voice is selected will play silence, so avoid using any SFX voices that are not common to both Level 1 and Level 2 XG instruments. The table below compares the Level 1 (MU50/DB50XG) SFX voice list with the Level 2 (MU80) voice list:
Level 1 (MU50/DB50XG) SFX Voice List Pch# Bank 0 # of Elements ---------------------------------------------------1 CuttngNz CttngNzStr Slap Fl.KClik 33 Rain Thunder Wind Stream Bubble Feed Level 2 (MU80) SFX Voice List Pch# Bank 0 # of Elements ---------------------------------------------------1 CuttngNz CttngNz3 DstCutNz Str Slap B.Slide P.Scrape Fl.KClik 33 Rain Thunder Wind Stream Bubble Feed 44 45
Level 1 (MU50/DB50XG) SFX Voice List Pch# Bank 0 # of Elements ---------------------------------------------------Dog Horse Bird Ghost Maou Tel.Dial DoorSqek Door Slam Scratch Scratch 70 WindChm TelphonCarEngin Car Stop Car Pass CarCrash Siren Train Jetplane Starship Burst Coaster SbMarine 95
Level 2 (MU80) SFX Voice List Pch# Bank 0 # of Elements ---------------------------------------------------Dog Horse Bird 52 Kitty Growl Haunted Ghost Maou Tel.Dial DoorSqek Door Slam Scratch Scratch 70 WindChm TelphonCarEngin Car Stop Car Pass CarCrash Siren Train Jetplane Starship Burst Coaster SbMarine 95
Level 1 (MU50/DB50XG) SFX Voice List Pch# Bank 0 # of Elements ---------------------------------------------------Laughing Scream Punch Heart FootStep 113 MchinGun LaserGun Xplosion FireWork 128
Level 2 (MU80) SFX Voice List Pch# Bank 0 # of Elements ---------------------------------------------------Laughing Scream Punch Heart FootStep Applaus113 MchinGun LaserGun Xplosion FireWork 128
7. Effects *Non-GM1* The following three effect blocks are available for use by both Level 1 and Level 2 XG instruments: Reverb Chorus Variation The types available in each block are listed in the XG Basic (Required) Effect List included with the XG Specifications. There is a risk of losing song data compatibility if you use optional types, optional parameters, or optional effects. Avoid using, for example, White Room or Pitch Change. If your data requires a distortion effect and is intended for playback on a Level 2 XG instrument such as the Yamaha MU80, use a distortion program within the Variation block rather than a distortion program contained within the Distortion (Insertion Effect 1) block. If you use SysEx parameter change effect parameters, do not modify parameters 11 and higher. (Use only parameters 1 - 10.) Details of the parameters are given in the XG Effect Parameter List in the XG Specifications. In the Variation block, Variation Connection is initially set to INS (that is, in F0H, 43H, 10H, 4CH, 02H, 01H, 5AH, ddH, F7H, the default setting is ddH=00H.) In this state, a Variation effect can be used for only one part, so specify Variation Part as needed. (In F0H, 43H, 10H, 4CH, 02H, 01H, 5BH, ddH, F7H, specify ddH. By default, ddH=7FH i.e., OFF) When Variation Connection = INS, Control Change #94 Effect 4 Depth (Variation send) is ignored. By setting Variation Connection to SYS, you can apply Variation effects to two or more parts. (In F0H, 43H, 10H, 4CH, 02H, 01H, 5AH, ddH, F7H, set ddH=01H.) In this case, Control Change #94 (Variation Send) will specify the send levels of each part. Always set Variation Connection to SYS before setting values for Control Change #94.

8. System Exclusive parameter changes *Non-GM1* Avoid using System Exclusive parameter changes that are designated as optional in the XG Specifications. As exceptions, the following three may be used even though they are listed as optional: Multi EQ Data parameter change* Display Data parameter change A/D Part Data parameter change Because Multi EQ parameters are used only by Level 2 XG instruments (and are ignored by Level 1 XG instruments, which always provide a flat EQ response), EQ may be used only as long as a flat response would not cause the character of the song to be lost. Note that later-received SysEx messages always take priority, so when transmitting messages that reset the entire block of parameters (such as Effect Type), give consideration to the appropriate order and spacing. For example: Multi Effect Data parameter change - Send parameters after specifying the effect type. Multi EQ Data parameter change* - Send parameters after specifying the EQ type. Drums Setup parameter change - Use a Program Change message to select a Rhythm Kit, then (if necessary) use Multi Part Data parameter change Part Mode to select Drums Setup. Finally, send parameters.
9. Fade in/out *Non-GM1* To fade a song in or out, use Control Change #11 (Expression) or the universal system exclusive message MIDI Master Volume (F0H, 7FH, 7FH, 04H, 01H, llH, mmH, F7H). Note that MIDI Master Volume changes the relative values of all parts simultaneously. Also, when Variation Connection is SYS, parts which use a Variation effect may not fade correctly. In this case, use the effect parameter change Variation Return (F0H, 43H, 10H, 4CH, 02H, 01H, 56H, ddH, F7H) in conjunction with the MIDI Master Volume message.
10. Miscellaneous 10-1. When a Program Change message is received, the values of Control Change messages such as Volume or RPN/NRPN are not reset. Be aware of this when you use Program Change messages within a song. 10-2. Use Control Change #7 (volume) to adjust the volume balance between parts. If, when inputting song data, you use a pedal-type volume controller for performance expression, the data will be recorded as Control Change #7 data, but you should replace this with Control Change #11 (expression) data. 10-3. For performance expression within a song (such as crescendos and diminuendos), use Control Change #11 (expression). 10-4. *Non-GM1* Control Change #10 (panpot) may be used while a note is sounding. 10-5. For the following four messages, use only values of either 0 or 127 (i.e., on/off switch values)do not use half-pedal data values: Control Change #64 (Sustain) Control Change #65 (Portamento) Control Change #66 (Sostenuto) Control Change #67 (Soft Pedal) 10-6. *Non-GM1* Some parameters (such as Filter Cutoff Frequency, Filter Resonance, EG Attack Rate, and EG Release Rate) can be altered with Control Change messages as well as with NRPNs and/or System Exclusive parameter messages. We recommend that such parameters always be altered with Control Changes rather than with NRPNs or System Exclusive messages. 10-7. Do not use Control Change #84 (Portamento Control). 10-8. Do not use Control Change #120 (All Sound Off) or #123 (All Note Off). 10-9. Do not use AnH (Polyphonic Aftertouch) or DnH (Channel Aftertouch). When inputting data from a keyboard or other controller in realtime, keep Aftertouch out of the music data with the use of a MIDI input filter.