Sequential Pro-3 Issues

[gmtta]

I'm wondering if the Pro-3 adaptation was done prior to firmware 1.2.0.4 as importing banks throws lots of errors and many patch names are left blank, 'invalid', or contain a number. Not sure if it matters, but this is with the SE edition.

[RadekPilich]

That' shouldn't affect it. First make sure of this - You must set 'MIDI Rx Prog Chg' to 'ON'.

The current adaption is not ready to handle User / Factory banks correctly.

Add these lines to Sequential Pro 3.py (between install and make_test_data)

def bankDescriptors():
    return [{"bank": x, "name": f"User Bank U{chr(ord('1') + x)}", "size": 128, "type": "Single Patch"} for x in range(4)] + [{"bank": x, "name": f"Factory Bank F{chr(ord('5') + x - 4)}", "size": 128, "isROM": True, "type": "Single Patch"} for x in range(4,8)]

def generalMessageDelay():
    # without delay, PRO3 doesn't manage to process a full bank of program dumps and stops after a handful of patches
    return 50

def friendlyBankName(bank):
    if bank > 3:
        bank = bank
        return f"F{chr(ord('1') + (bank % 8))}-P"
    return f"U{chr(ord('1') + (bank % 8))}-P"

def friendlyProgramName(patchNo):            # Pad with leading zeros to make it match the display
    program = patchNo % numberOfPatchesPerBank()
    bank = patchNo // numberOfPatchesPerBank()
    return "%s%s" % (friendlyBankName(bank), str(program + 1))

[christofmuc]

Oh wow, the message delay might be the reason why @gmtta has the problems getting the bank dump correctly done?

[RadekPilich]

Yes, it gets stuck on an xx patch without delay, didn't 6 go the through all of them. At least on my unit.

[christofmuc]

Implemented Radek's suggestions on master in d56c90d. Thank you @RadekPilich!

• Added 50 ms message pacing and setup guidance for MIDI Rx Prog Chg / MIDI Program Receive and the SysEx connection.
• Added eight bank descriptors: writable U1-U4 and read-only F5-F8, 128 programs each.
• Simplified the naming implementation using GenericSequential's existing friendlyBankName/friendlyProgramName callbacks; no shared-module changes were necessary. Display examples: U1-P1, F5-P1, F8-P128.

The F5-F8 labels are intentional. Although the v1.2 manual says F1-F4, Sequential's firmware notes for 1.2.0.4 explicitly describe F1-F4 on the Morph screen as inconsistent with F5-F8 elsewhere: https://forum.sequential.com/index.php?topic=5201.0

Validation: 48 passed, 21 skipped including the Pro 3 generic/dedicated tests and Fourm regressions. Tests cover the existing 512-patch file, all eight bank request sequences with mock responses, bank labels/ROM metadata, and outgoing 50 ms pacing. Mock tests do not establish real hardware timing or prove the original import failures are gone.

Leaving this issue open pending a hardware retest of full user/factory bank imports on the affected firmware. These changes will be in the next build/release, not the existing 2.9.0 installer.

[christofmuc]

Good analysis by Codex here, I will close the issue on the upcoming 2.10 release, feel free to reopen if it doesn't work as expected!

[christofmuc]

KnobKraft Orm 2.10.0 now includes Radek's Pro 3 improvements: U1-U4 / read-only F5-F8 bank metadata, 50 ms message pacing and setup guidance. Thank you @gmtta for reporting this and @RadekPilich for the investigation and proposed changes!

A retest of full user/factory bank imports on the affected firmware is still needed to confirm the original missing/invalid-name problem is gone. Keeping this issue open for that confirmation rather than treating the automated tests as hardware verification. Please report your firmware version and whether a full import now completes.