Subsystem Search

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.

Intellijel Cascadia MIDI implementation

The MIDI implementation of the Intellijel Cascadia: what it answers to, and the 4 controls it offers, on firmware 1.4.4.0.

At a glance

Firmware
This page describes firmware 1.4.4.0.
MIDI channel
Any one of 1 to 16.
MPE
Received.
Voices
1, or 2 as set on the instrument.
Note range
Not recorded.
Velocity
Received. Release velocity is ignored.
Aftertouch
Channel pressure, one value for every key.
Pitch bend
How far it bends is not recorded, and the range can be changed.
Program change
Not received or sent.
MIDI clock
Sent and received.
Start and stop
Neither sent nor received.
Control change
Assigned by the player, apart from the 4 controls below.
NRPN
None.
System exclusive
None.

Controls

In the order the definition gives them, by group. Under each control is the name a program calls it by, once pymidiinstrumentdefs.load("intellijel/cascadia") has loaded them.

MIDI / CV

Control CC Values
Portamento Control
portamento		 65 0 off, 127 on
Received, never sent
Portamento Time MSB
portamento_time		 5 0-127
Received, never sent
MIDI CC Output
midi_cc_output		 2 0-127
Received, never sent; The factory's number, which the player can change
MIDI MOD Output
midi_mod_output		 1 0-127
Received, never sent; The factory's number, which the player can change

Where these facts come from

Source documents

How they were read

EIGHT JACKS, AND MIDI IS WHAT COMES OUT OF THEM. "This section of Cascadia takes MIDI data received at the back panel's MIDI IN [14.G] jack (or USB MIDI [14.J] port) and extracts 8 user-assignable CV outputs for patching into Cascadia" (manual p. 17). So a controller sent here does not set a parameter: it moves a voltage, and what that voltage reaches is decided with a patch cable. The eight are pitch, velocity, gate, trigger, a clock, an LFO, and the two the controllers below drive.

FOUR CONTROLLER NUMBERS ARE PUBLISHED FOR THIS INSTRUMENT AND NO MORE. All 122 sheets were swept for every form of a controller number, and five appear: 1, 2, 5, 65 and 74. Two of the five are fixed, two are factory choices the owner can replace, and the fifth is a recommendation rather than an assignment, so it is not below.

THE TWO FIXED ONES ARE A PORTAMENTO, AND THEIR VALUES ARE EXACT RATHER THAN BANDED. "CC 65 (Portamento Control) on Cascadia's MIDI Channel" (manual p. 18), with "Value 127 = Turns Cascadia's portamento ON" (manual p. 18) and "Value 0 = Turns Cascadia's portamento OFF" (manual p. 18) - two numbers and not two halves of the range, which is why that control carries choices and not values: nothing says what 1 to 126 do. Beside it, "Sets Cascadia's portamento time from 0 seconds (CC 5 = 0) to about 1.1 seconds (CC 5 = 127), slewing the pitch" (manual p. 18). Neither is learnable and neither can be reassigned; they are the only two controllers that reach into the instrument rather than out of a jack. The changelog dates them to firmware 1.2.0.2 and records a fix to both at 1.3.1.0 - "FIXED: MIDI Portamento (CC# 5 and 65) was not responding in certain configurations" (changelog), an entry the manual's own change log has not got.

THE OTHER TWO ARE DEFAULTS, AND SAYING SO IS THE POINT OF RECORDING THEM. "By default, this is assigned to 'CC 2 (Breath Controller)', though you can select another value" (manual p. 18) of the MIDI CC jack, and "By default, this is assigned to 'CC 1 (Modulation Wheel)', though you can select another source" (manual p. 21) of the MIDI MOD jack. Either jack can instead be driven by any number at all - "Selects which MIDI CC# is the source of the control voltage sent out Cascadia's MIDI CC jack" (manual p. 109), over the full (0 - 127) - or by channel pressure. So these two are what a Cascadia answers to as it leaves the factory, and a consumer offering them should say as much rather than present them as fixtures. akai/mpc_key_37 carries its single control on exactly this footing. The Cascadia's two say assignment: default, the format's word for a factory number an owner can move; until 2026-10-13 only this paragraph said so.

SENDING EITHER OF THEM TO A NEW CASCADIA MOVES A VOLTAGE THAT IS CONNECTED TO NOTHING. Four of the eight jacks are patched internally when the instrument arrives and four are not, and the two that carry controller data are in the second group: the manual gives a DEFAULT ROUTING for pitch, velocity, gate and trigger, and none for the MIDI CC, MIDI MOD, MIDI LFO or MIDI CLK jacks. So controller 1 and controller 2 are audible only after somebody patches a cable, and a panel that offers them without saying so promises a sound that will not arrive.

THE FIFTH NUMBER IS ADVICE AND IS DELIBERATELY NOT A CONTROL. For expressive play the maker says to teach the MIDI CC jack controller 74: "CC Number = 74 Sound Control 5 (MPE Timbre)" (manual p. 102), and "CC74 (MPE Y/Timbre) and Pressure (MPE Z/Pressure) are learnable and output from the active note" (manual p. 118). Learnable, which is the opposite of assigned. It is also the trap in the sweep that found these five: the words Sound Control 5 end in a digit, so a pattern that accepts control 5 reads the name of controller 74 as a mention of controller 5. A number is only a controller number when the word beside it says so, and control does not say so.

EVERY CONTROL BELOW IS receives, AND ONE SENTENCE SETTLES IT FOR ALL OF THEM. "If enabled in the Intellijel Config app, the MIDI OUT jack transmits Cascadia's internal MIDI Tap Clock (as generated by Cascadia)" (manual p. 78). A clock, and nothing else: no sentence in any of the four editions names a controller, a note or a program this instrument sends. The DIN output is otherwise a repeater - "All MIDI data appearing at the MIDI IN [14.G] jack is transmitted out this jack" (manual p. 78), with the qualification that "MIDI data arriving at the USB MIDI [14.J] port is not thru'd" (manual p. 78), which is a fact about a cable rather than about this instrument's own messages. dreadbox/artemis reached the same direction at the previous rank by inference from its settings and said it was the one judgement not resting on a sentence. This one rests on a sentence.

THE DOCUMENT IS THREE FIRMWARE RELEASES BEHIND THE INSTRUMENT, AND THAT WAS CHECKED RATHER THAN ASSUMED HARMLESS. The manual's cover reads Firmware Version: 1.4.1 | Revision: 2026.04.13, while the maker's firmware index gives 1.4.4.0 (firmware_index) and the product page prints the same number at the end of its Firmware row (product_page). The three releases in between are a USB product identifier, a velocity timing fix and an LFO freeze (changelog), and not one of them touches anything recorded here - which is why the model above names the current firmware rather than the manual's.

THE PITCH BEND RANGE IS CONTRADICTED INSIDE ONE DOCUMENT, AND THE CHANGELOG DECIDES IT. The configuration chapter prints Pitch Bend Range (1 - 24), which "Sets the max number of semitones that an incoming MIDI Pitch Bend message will affect the MIDI PITCH" (manual p. 106). Twelve sheets later the same manual says "TWEAK: Pitch Bend Range (Sensitivity) max increased from 24 to 96" (manual p. 118). The second one is firmware 1.4.0.0's release note, printed with its version heading removed directly above a change log whose newest entry is 1.4.1, so it reads as though it were current; every line of that block is in the changelog's 1.4.0.0 entry, including its misspelling of Ableton as Albeton, which is what proves the match rather than suggests it. The manual's own 1.4 entry lists that release's features and drops both of its tweaks, which is how the stale table went unnoticed. So the maximum is 96 and the table is out of date - but no edition anywhere prints the shipped value, so pitch_bend records that the range is settable and not what it is. In the MPE modes it is a separate setting again, whose default the changelog does give as 48.

AND THE TAP CLOCK OVER USB HAS TWO ANSWERS IN EVERY EDITION, SO NEITHER IS CARRIED. Of the USB port: "Tap Clock output is enabled, by default, in the factory settings" (manual p. 78). Of the same port's Off setting: "Cascadia does not send its own internal Tap Tempo MIDI Clock to the USB MIDI port. This is the default setting" (manual p. 113). Both sentences are in the current edition, in the edition before it, in the one before that and in the first (manual_1_3, manual_1_2, manual_1_1), so there is no moment at which one replaced the other. clock: both does not rest on either: it rests on the DIN output, whose entry in the configuration chapter has never carried a default marker in any of the four, leaving sheet 78's sentence about it uncontradicted. What a consumer should do about USB is check the instrument.

CLOCK GOES BOTH WAYS AND THE INSTRUMENT DECIDES WHICH WITHOUT BEING ASKED. "By default, Cascadia is set to power on using Tap (Auto Detect MIDI) mode" (manual p. 20), and "if you send MIDI Clock into Cascadia, then it will automatically switch over to MIDI clock mode" (manual p. 20). So it answers to an incoming clock, generates one of its own, and the arrival of MIDI is itself the switch. "This is the factory default" (manual p. 112) of that automatic mode.

WHAT THIS MAKER CALLS A MIDI MODE IS NOT WHAT MIDI CALLS ONE, and midi.mode is omitted for that reason rather than for want of looking. The configuration chapter's MIDI Mode offers Mono, Dual Mono, MPE (Lower Zone) and MPE (Upper Zone) (manual p. 105), with "standard Mono mode (factory default)" (manual p. 119) - and the front panel calls the first of the four Single where the chapter calls it Mono. None of them is a reception mode in the sense the specification means, and no sheet of the four editions names one or uses the word omni.

DUAL MONO BUYS A SECOND VOICE WITH THE TWO CONTROLLER JACKS. "Cascadia works as a 2-voice synth, and responds to two independent MIDI channels" (manual p. 105), the second channel's "pitch appearing on the MIDI CC output jack" (manual p. 100) and its gate on the MIDI MOD jack, "thus disabling these two jacks as MOD sources on the first channel" (manual p. 100). So the two controllers below and the second voice are alternatives, not additions - which is the sharpest thing in this definition for anybody writing a panel, because nothing in the format can express it.

IT ANSWERS TO MPE WHILE SOUNDING ONE NOTE, SO midi.mpe IS receives AND per_voice_channels IS LEFT UNRECORDED. "Although a monophonic instrument, Cascadia is capable of responding to MPE data" (manual p. 101), across up to fifteen member channels, with pitch on bend, timbre on a learned controller and pressure on channel aftertouch. In this corpus that field means the voices are spread across channels, one note to a channel - which is what modal/carbon8m and waldorf/iridium set it for. One voice cannot be spread, so true would misdescribe the instrument and false would hide the feature. moog/messenger left the same field unrecorded on the same kind of machine, and consistency with it is deliberate. "Cascadia is a monophonic device per MIDI Channel" (manual p. 106).

VELOCITY ARRIVES AND A PANEL SWITCH DECIDES WHETHER IT DOES ANYTHING. "MIDI VEL OUT - CV output with a range of 0V - 5V. The voltage is proportional to the velocity of the last played MIDI note" (manual p. 21), and that output is one of the four patched at the factory: "If nothing is patched into the CTRL IN [4.B] jack, then MIDI VELOCITY [1.E] is used" (manual p. 32) by envelope A. But the switch beside that jack has three positions, and at one of them "The envelope ignore any signal present at the CTRL IN jack, and always outputs a full (+5V) envelope" (manual p. 32) - the maker's own missing letter. The manual even names a silence an owner can meet, with the switch at LEVEL after a velocity of zero: "it will have no audible effect until a higher voltage is patched into the jack" (manual p. 32). No edition prints which position that switch ships in, so velocity is recorded as received: gated is for an instrument whose documents say it ships ignoring velocity until a setting is changed, and where this switch stands is the owner's. Until 2026-10-14 this said gated, with an empty gated_by, as erica_synths/perkons_hd_01 did.

AFTERTOUCH IS CHANNEL PRESSURE, SAID TWICE AND IN THOSE WORDS. "The MIDI CC jack outputs a voltage derived from Channel Aftertouch, rather than from a MIDI CC value" (manual p. 109), and the MIDI MOD jack's chapter says the same of itself. No form of polyphonic aftertouch or key pressure appears anywhere in 122 sheets.

CHECKED ABSENCES, AND WHAT WAS SWEPT FOR EACH. Every sheet of the current edition was searched with its text squashed, and every hit was read rather than counted, which for two of the terms is the whole of the work:

NOT RECORDED: the note range. The manual maps "MIDI note 0 (C-2)" (manual p. 17) to the bottom of the pitch output and note 120 to the top, over "a 10 octave range" (manual p. 17) - but a note above 120 is not thereby silent, which is what this field would claim, and no sheet says what becomes of one. The same sheet also gives that top voltage as a figure its own range contradicts, which is a third self-contradiction and the only one that changes nothing here.