This page is read from MIDI Instrument Definitions, a Python library used in the author's other projects, in which every number is traced to a document its maker published. Its source is on GitHub.
Akai Professional MPC Live MIDI implementation
The MIDI implementation of the Akai Professional MPC Live: what it answers to, and the 2 controls it offers, on firmware 3.9.1.
At a glance
- Firmware
- This page describes firmware 3.9.1.
- MIDI channel
- Any one of 1 to 16.
- Voices
- Not recorded.
- Note range
- Not recorded.
- Velocity
- Sent and received.
- Aftertouch
- Channel pressure and polyphonic, each answered. It sends aftertouch too.
- Pitch bend
- How far it bends is not recorded, and the range can be changed.
- Program change
- Received and sent.
- MIDI clock
- Sent and received.
- Start and stop
- Not recorded.
- Control change
- Assigned by the player, apart from the 2 controls below.
- NRPN
- Not recorded.
- System exclusive
- Not recorded.
Controls
In the order the definition gives them, by group. Under each control is the name a program calls it by, once pymidiinstrumentdefs.load("akai/mpc_live") has loaded them.
| Control | CC | Values |
|---|---|---|
Mod Wheelmodulation_wheel |
1 | 0-127 Received, never sent |
Expressionexpression |
11 | 0-127 Received, never sent |
Where these facts come from
Source documents
- MPC Standalone OS User Guide, v3.9, dated 15 June 2026, its user manual, read on 4 October 2026 (the file read).
- MPC 3.9.1 Release Notes, 3.9.1, its release notes, read on 4 October 2026 (the file read).
- MPC 3.9 Release Notes, 3.9, its release notes, read on 4 October 2026 (the file read).
How they were read
MPC Standalone OS User Guide, 530 pages, the edition Akai names v3.9. Cited throughout, and the pages that carry the weight are these: the introduction on p. 7; the project on p. 45; the track's outgoing program change on p. 54; every MIDI preference on p. 64; the pitch bend preference on p. 65; the keyboard screen on p. 102; the pad note map on pp. 81 and 128; the MIDI track's controller sliders on p. 255; the CV chapter on p. 256; the track's automation on p. 310; the Q-Link modes on p. 314; the track settings on p. 267; the controller filter on p. 268; MIDI Learn on p. 320; MIDI Control Mode on pp. 333 to 337; the hardware chapter's own scope on pp. 344 and 358; this machine's panels on pp. 352 and 353; the glossary on p. 386; MIDI Machine Control on p. 456; the specifications on pp. 468, 472 and 479; and the unlisted addenda on pp. 487 and 528. MPC 3.9 Release Notes and MPC 3.9.1 Release Notes, for which firmware the guide describes and which machines each release is for.
ONE DOCUMENT, TEN MACHINES, AND IT SAYS SO ITSELF. "This user guide explains how to create music using the MPC 3 operating system on first and second generation standalone MPC hardware, including: MPC X, MPC X Special Edition, MPC Live, MPC Live II, MPC One, MPC One+, MPC One G2, MPC Key 61, MPC Key 37, and MPC Key 37 G2" (p. 7), and from there on they are, in the guide's own words, "MPC hardware" (p. 7) unless otherwise noted. Two others are named and excluded: "MPC Live III and MPC XL, which have a separate User Guide" (p. 7).
AND IT STATES ITS OWN COVERAGE THREE TIMES, ONCE DIFFERENTLY. The sentence above names ten; so does the hardware chapter, "This chapter explains the features and functions of each MPC v3.9-supported model: MPC X / MPC X Special Edition, MPC Live, MPC Live II, MPC One / MPC One+, MPC One G2, MPC Key 61, MPC Key 37, and MPC Key 37 G2" (p. 344). The one in between names eight, dropping the MPC X and the MPC X Special Edition - and it is the one that defines the term "MPC hardware" (p. 7) that the whole book then uses. Read literally, every unqualified sentence in 530 pages excludes a machine that has its own chapter, its own appendix and its own specification. That is plainly not meant, and it is what the page says. It does not touch this instrument, which is in all three lists, and it is recorded because the next person to ask this document who it is about will meet it.
AND IT INHERITS BETWEEN SOME MACHINES AND NOT OTHERS. "Mentions of the features and functions of MPC X hardware throughout this guide also apply to the MPC X Special Edition unless otherwise noted" (p. 344), and the MPC One does the same for the MPC One+ (p. 358). There is no such rule for this machine and the MPC Live II, though the guide writes "MPC Live/Live II" (p. 64) freely. So a sentence that names the MPC Live II alone is not this machine's, and nothing may be borrowed from one to the other.
SO THE SCOPING WAS ENUMERATED RATHER THAN READ, and the guide marks it two ways. The first is a parenthetical after the thing it qualifies - 44 distinct ones, 132 times - and every one of the ones that exclude this machine was read, not a sample of them. Not one changes what the instrument does over MIDI: they are an Ethernet adapter, a USB audio channel count, a USB-C data role, a VU meter source, an on-screen display strip, and for the rest, which button opens a given screen on which machine. Even the two that sit beside the words MIDI Control Mode are of that last kind: the mode is reached from the menu on every machine and only the shortcut differs (p. 333).
THE SECOND MECHANISM IS A SECTION'S OPENING SENTENCE, AND IT IS THE ONE THAT MATTERS. "The keyboard control screen allows you to edit the functions of the keybed on MPC Key 61, Key 37, and Key 37 G2" (p. 102) - and everything under that heading is scoped by it, including a MIDI channel setting and an aftertouch setting, neither of which is marked in any way. A reading that enumerated the parentheticals alone would have taken both for this machine's.
AND THE MODEL NAMES ARE PREFIXES OF EACH OTHER, WHICH MAKES A NAIVE READING WRONG IN ONE DIRECTION. This instrument's name is the whole of the next one's: MPC Live, MPC Live II. So is MPC One against MPC One+ and MPC One G2, MPC Key 37 against MPC Key 37 G2, and MPC X against MPC X SE. Any test that looks for a name inside a note matches the longer machine's notes too, silently, and always by claiming more than the maker said. The first version of the script that read this document did exactly that and reported two notes excluding this machine as including it.
A FOURTH PLACE IS AN OMISSION FROM THIS MACHINE'S OWN SPECIFICATION, WHICH IS BINDING AND IS NEVER SAID IN PROSE. The connections list on p. 472 has no CV or gate jack, no MIDI thru-port and no footswitch input, where the MPC X's lists eight CV/Gate outputs and two footswitch inputs (p. 468) and the MPC Key 61's lists a thru-port (p. 479). So the guide's unqualified CV Tracks chapter (p. 256), its footswitch settings and its thru settings cannot be about this machine, and no sentence anywhere says so. An absence from a table is as binding as a parenthetical and cannot be searched for.
AND THE MACHINE ITSELF OFFERS CONTROLS IT DOES NOT HAVE, WHICH THE GUIDE WARNS ABOUT. "The Control field menu will show many more hardware controls than are actually available on your MPC hardware. This is because the list includes all possible controls from all current MPC models" (p. 334). So the shared operating system reaches the screen as well as the manual, and a reading taken off this instrument's own display would be wrong in the same direction as a reading taken off an unqualified sentence. It is the third place the ten machines have to be told apart, and the only one a document cannot be checked against.
EVERY ASSIGNMENT BUT TWO IS THE OWNER'S. "You can use the MIDI Learn function to assign external MIDI controllers to various parameters in your specific MPC project" (p. 320) - mixer parameters, pad parameters, track parameters and effect parameters - and "These assignments will be saved with your MPC project" (p. 320). The same is true in the other direction: in MIDI Control Mode each button has a "CC Number: This determines what MIDI Control Change number the button will send to the software" (p. 335), and each Q-Link knob has the same field for itself (p. 336). So apart from the two below, which number reaches which parameter is not a property of the instrument.
THE TWO ARE 1 AND 11, AND THEY ARE CARRIED, RECEIVED ONLY. The arpeggiator's modifiers "enable Expression messages (MIDI CC #11) from external MIDI controllers to modify the arpeggiator" (p. 85), and the same for "Modulation messages (MIDI CC #1) from external MIDI controllers" (p. 85); a keygroup's LFO has a knob "to adjust how much control mod wheel data (MIDI CC1) has over the LFO." (p. 239). Each effect is one the owner switches on or turns up, and each number is fixed, the MIDI specification's own modulation and expression. control_change IS learned: the word says there are numbers that are the owner's and that no file can list, and it stands beside the numbers a maker fixes, which here are these two. Until 2026-10-13 the field was left unset, the word having been read as one that could not say both.
AND THE GUIDE SAYS A FIXED SET EXISTS WITHOUT PRINTING IT. "MIDI: In this mode, the Q-Links are fixed to a selection of MIDI performance controls, allowing you to control modulation, pitch bend, sustain and other common MIDI messages as you play" (p. 314). Fixed, named by function, and not one number given. The same thing happens twice more: a track's MIDI automation offers "Standard MIDI control change assignments" (p. 310), which defers to the specification rather than enumerating, and the pad note map offers a "Classic MPC" preset whose numbers are never printed (p. 128). Three times this document asserts that a fixed map exists and declines to publish it, which is a sharper absence than silence.
THE ONE LIST OF CONTROLLER NUMBERS IN THE DOCUMENT IS NOT A CONTROL MAP. "MIDI CC Filter: Check these boxes to allow the listed MIDI Control Change messages to pass through the track" (p. 268), and the thirteen it lists are 1, 2, 3, 4, 5, 7, 10, 11, 64, 65, 66, 67 and 68 - modulation, breath, undefined, foot, portamento, volume, pan, expression, sustain, portamento on/off, sostenuto, soft pedal and legato. Those are the MIDI specification's own names for those numbers, and the list says which message types a track will forward, not what this instrument does when it receives one. No control is written below on the strength of it.
AND THAT LIST NUMBERS FOUR THINGS THAT ARE NOT CONTROL CHANGES AT ALL. Beside the thirteen it prints CC128 Pitchbend, CC129 Channel Pressure, CC130 Program Change and CC131 Aftertouch - numbers above the 127 a control change can carry, used as labels for other message types. A reader who took the list for a control map would write four controls this instrument cannot have.
NO NOTE MAP IS PUBLISHED EITHER, AND THE PADS' NOTES ARE ASSIGNED. "The Edit Pad Note Map function lets you assign specific MIDI notes to your MPC hardware pads" (p. 128), with three presets: "Chromatic C1 (an ascending chromatic scale, beginning with C1), Chromatic C-2 (an ascending chromatic scale, beginning with C- 2), or Classic MPC (the default MIDI note map of classic MPCs)" (p. 128). The classic map's numbers are not given, no preset is named as the default, and nothing says which is loaded on a new track. So voices holds nothing.
THE SAME THREE PRESETS ARE NAMED DIFFERENTLY TWENTY-SEVEN PAGES EARLIER. "You can also choose from three Custom presets: Chromatic C1, Chromatic C2, or Classic MPC" (p. 81) - C2 where p. 128 has C-2. One document, one list of three things, two spellings of the second of them, and nothing here chooses between them.
WHAT THE INSTRUMENT DOES WITH THE CLOCK IS A FOUR-WAY SETTING AND THIS FORMAT HOLDS TWO OF IT. "Receive: This determines whether your MPC hardware receives MIDI Clock information (MIDI Clock), MIDI Time Code information (MIDI Time Code (MTC)), communication from Ableton Link (Ableton Link), or none of these (Off)" (p. 64), and "Send: This determines whether your MPC hardware sends MIDI Clock information (MIDI Clock), MIDI Time Code information (MIDI Time Code (MTC)), or neither (Off)" (p. 64). So clock travels both ways and either direction can be switched off; MIDI Time Code and Ableton Link are the other choices and no field here records them.
IT ANSWERS TO MIDI MACHINE CONTROL AND SENDS IT, which no field here records either. "Your MPC hardware can send and receive MIDI Machine Control (MMC) messages, a standard protocol for transport controls" (p. 456), and the appendix tabulates which commands go each way. The pair of switches is in the preferences: "Receive MMC: When enabled, your MPC hardware will be able to receive MIDI Machine Control (MMC) information" (p. 64), and Send MMC is the same for the other direction (p. 64).
PROGRAM CHANGE GOES BOTH WAYS AND REACHES TWO DIFFERENT THINGS. "Program Change: This determines what an incoming MIDI program change message will change: a Sequence or Track" (p. 64), and a track sends one: "To select a MIDI program change value for the track to send, check the Program Change box to enable program changes" (p. 54), with a bank change beside it.
AND THE TWO FIGURES THE DOCUMENT GIVES FOR A PROGRAM CHANGE DISAGREE, AND NEITHER IS THE PROTOCOL'S. What a track sends is "enter a value from 1-127" (p. 54); the glossary says "Programs 1-128 can be changed via program change messages" (p. 386). A program change carries 0 to 127. Nothing here is recorded from either.
NOT RECORDED: how many presets a program change reaches. The guide never says how many sequences a project holds, and what an incoming program change selects is a setting with two positions, so one number would be wrong for one of them. "A single project can hold up to 128 tracks" (p. 45) is the only count of either, and it is not offered as a program change range. The glossary's "Programs 1-128 can be changed via program change messages" (p. 386) is the MIDI specification being explained and not this instrument being described.
AND THE GLOSSARY IS WORTH NAMING, BECAUSE IT READS LIKE A SPECIFICATION AND IS NOT ONE. Printed pp. 384 to 387 explain MIDI itself - what a control change is, that a controller number "can range from 0 to 127" (p. 384), what note on and note off do. Nothing in it is a statement about this machine, and three of the terms a control map would be written in appear nowhere else in 530 pages.
THE PADS SEND AFTERTOUCH, SO aftertouch_transmits IS TRUE, AND WHICH KIND THEY SEND IS A THREE-WAY SETTING NO FIELD HOLDS. "Aftertouch: This determines how the pad's aftertouch (pressure applied to the pad after the initial press) behaves" (p. 334), and the three positions are given: "Off: The pad will not send any aftertouch messages" (p. 334), "Channel: If you press multiple pads that have this setting, the aftertouch messages they send will be identical" (p. 334), and "Poly: If you press multiple pads, the aftertouch message each pad sends will be independent from the others" (p. 334). The specification calls them "(16) velocity- and pressure-sensitive pads, RGB-backlit" (p. 472).
AND IT ANSWERS TO BOTH KINDS AS THE GUIDE DIVIDES THEM, SO aftertouch IS both. What the pads send is the owner's choice, and this field records what an instrument answers to, which the guide states as well. The arpeggiator's modifiers "enable Aftertouch messages from the pads or external MIDI controllers to modify the arpeggiator" (p. 85) and "enable Channel Pressure messages from external MIDI controllers to modify the arpeggiator" (p. 85), and a keygroup has both: "Aftertouch To Cut 1/2 determines how much a connected MIDI keyboard’s aftertouch data affects the filter cutoff." (p. 242), and "Pressure To Filter determines how much a connected MIDI device’s channel pressure data affects the filter envelope." (p. 242). The guide never says polyphonic or key pressure, and elsewhere calls channel pressure aftertouch too (p. 334), but here and in its event list (p. 184) it sets Aftertouch beside Channel Pressure as a kind of its own, and the instrument takes each, so both. The keyboard screen's own aftertouch setting is three other machines' (p. 102), and nothing here rests on it.
VELOCITY TRAVELS BOTH WAYS. The pads send it: "Velocity: This determines whether the pad will be velocity-sensitive (On) or not (Off). When set to Off, pressing the pad will send a note at full-level (127) always" (p. 334), and this machine's own chapter can fix the level they trigger at - full at 127, half at 64, or a level the owner sets (p. 352). And an incoming note's velocity is used: "Use the Velocity Scaling field to apply scaling to the incoming velocity." (p. 193), and a track's "Velocity Range: This determines the Velocity Min and Velocity Max that passes MIDI." (p. 268). So note_on is both.
TWO LIMITS ON THE ABSENCE ABOVE, BOTH WORTH STATING. Every MIDI screen in this guide is an image, so a default controller number rendered in a field would be invisible to any search of the text - and p. 255, where a MIDI track shows "an overview of all available MIDI CCs with a slider for each" (p. 255), is exactly where such a default would be shown. Nothing here claims more than that the words are not in the text.
AND FORTY-THREE PAGES ARE MISSING FROM THE TABLE OF CONTENTS. Its last entry is p. 486, and pp. 487 to 529 are an unlisted Addenda section for a paid upgrade - "The MPC Pro Pack can be purchased as an upgrade for all Gen 1 MPC standalone hardware running MPC v3.6 or later" (p. 487), which includes this machine. They were searched with everything else. The only MIDI table in them belongs to a plugin and routes incoming mod wheel, aftertouch, expression and a footswitch to a destination the owner picks (p. 528) - a plugin's parameter, sold separately, and one of its four rows is a footswitch this machine has no socket for. It is not the only place an incoming controller is given an effect: the arpeggiator and a keygroup's LFO, above, are two of several, and the only ones that print a controller number.
NOT RECORDED: system exclusive, NRPN and RPN. None of the three words appears anywhere in the guide or in either set of release notes - silence rather than a checked absence.
TWO MIDI INPUTS AND TWO MIDI OUTPUTS, which no instrument in this corpus has had before. "(2) 5-pin MIDI inputs" (p. 472) and "(2) 5-pin MIDI outputs" (p. 472), and the panel calls them in the plural: "MIDI In: Use standard 5-pin MIDI cables to connect these inputs to the MIDI outputs of optional external MIDI devices" (p. 353). Nothing here records a port, so a channel number in this definition does not say which socket it arrives on.