Auto-detected piston links disappearing

This is a version 4 system, and I’m working with Neil Jensen’s 3-35 theatre organ; this is an old set (originally produced for Hauptwerk 3 I think) so I wonder whether that’s the cause of my problems?

I’ve been auto-detecting pistons on a new console; these work as expected for a while, then suddenly disappear. I’m mainly using the scoped pistons, as on a theatre organ I need access to tremulants – for instance, if I want to use the Orchestral Oboe as a solo stop I often want it’s tremulant off; scoped pistons are perfect for this, whereas I can’t think of a way to do it with divisionals. Also scoped pistons enable those on the console to stay lit, which I find helpful.

When my console pistons lose their connection to Hauptwerk, the scoped settings are still there, and work if I click on them on-screen; but if I use the “Manual adjustment” option, said scoped pistons are shown as “No Input” for both Input 1 and Input 2, whereas they ought to show the MIDI code sent by the console piston.

At the moment I’m unable to reproduce this problem reliably – it seems to happen at random. I just needed a Windows (10) update, which involved restarting the PC; after that the piston links I set up most recently are still there. I’m using “Save combination set” regularly – I’d expect that to preserve the console links?

I’ve checked with MIDI Ox and the (Hoffrichter) console is sending the expected MIDI messages; Hauptwerk’s own MIDI activity indicators are also showing what I’d expect. I have “Off-the-shelf MIDI organ console” set to Hoffrichter.

My understanding is that Hauptwerk’s Master pistons operate “outside” the sample set, and, so to speak, tell it what to do; in that case it wouldn’t matter which organ I have loaded. This seems an appropriate point at which to seek advice on the forum . . .

Could Windows have changed the name of the MIDI port(s) it presents to Hauptwerk after the update, or randomly? That happens sometimes, and HW's MIDI mappings can be lost.

I don't remember if HW4 did this, but recent versions give you a warning screen with options for how to handle the issue.

[mdyde]

Hello organsRgreat,

Briefly (since we no longer officially support Hauptwerk v4):

This is a version 4 system, and I’m working with Neil Jensen’s 3-35 theatre organ; this is an old set (originally produced for Hauptwerk 3 I think) so I wonder whether that’s the cause of my problems?

No -- version 3 sample sets should work fully in v4.

When my console pistons lose their connection to Hauptwerk, the scoped settings are still there, and work if I click on them on-screen; but if I use the “Manual adjustment” option, said scoped pistons are shown as “No Input” for both Input 1 and Input 2, whereas they ought to show the MIDI code sent by the console piston.

If any MIDI settings existed (e.g. had been auto-detected) for a given virtual control then you would always be able to see them via "Adjust MIDI/trigger settings manually ..." from right-clicking on the control; it isn't possible for a given virtual control to have MIDI settings configured/present/working yet for them not to be visible via that screen.

My guess is that:

- You don't actually have the scoped combinations auto-detected directly to the relevant Hoffrichter pistons, but instead have some other virtual control auto-detected to them, which is usually getting triggered indirectly (e.g. as a result of something else changing state as a consequence of the scoped combination being triggered), or:

- Some kind of MIDI feedback/loop is occurring between the MIDI console and Hauptwerk which is resulting in things changing states unintentionally, or:

- The MIDI driver has dropped the connection to its MIDI port(s), or renamed the ports. (Hauptwerk will warn about missing/renamed ports when it launches, but may not know if it happens whilst Hauptwerk is actually running and using them.)

If needed, I'd suggest turning on the "General settings | General preferences | Advanced ...: Diagnostics: log MIDI messages received and sent" option then using "Help | View activity log" to examine the incoming MIDI messages (if any), which MIDI ports they're from, what virtual controls they're triggering, and what Hauptwerk is sending in response.

My understanding is that Hauptwerk’s Master pistons operate “outside” the sample set, and, so to speak, tell it what to do; in that case it wouldn’t matter which organ I have loaded.

It's correct that Hauptwerk's master pistons (scoped, etc.) are separate and additional to any provided natively within the sample set. However, in Hauptwerk v4 their MIDI settings were always per-organ, i.e. auto-detecting them to MIDI pistons in one organ wouldn't affect others. Also, combinations stored to them are stored in the organ's combination sets (files). Hence effectively they are entirely per-organ.

[Edit: P.S. Mark and I replied at the same time.]

Thank you gentlemen for so much helpful advice. There’s a lot to think about here, and it will take me while to work through all the possibilities. Since I first posted it’s occurred to me that if I, or Hauptwerk, was loading the wrong Combination file, that could produce this sort of problem. I save each time I’ve auto-detected console pistons; if I loaded an earlier Combination file, that wouldn’t have the recently set links to the console. So I’ve double-checked that Hauptwerk is set to load the most recently-used Combination set when it starts up.

While I investigate, a few more points. The console does all the MIDI merging; it has only MIDI In, Out and Thru ports. The keyboards have always worked perfectly; if there was a problem with the MIDI port assignments I’d expect the keyboards to misbehave, as well as the pistons.

MIDI reaches the PC via a MOTU 828 Hybrid Mark 3 sound card, connected over Firewire, so that I can run two on the ASIO driver in Windows. The “primary” MOTU is #2, and is being seen correctly as the MIDI input:

“If needed, I'd suggest turning on the "General settings | General preferences | Advanced ...: Diagnostics: log MIDI messages received and sent" option then using "Help | View activity log" to examine the incoming MIDI messages (if any), which MIDI ports they're from, what virtual controls they're triggering, and what Hauptwerk is sending in response”.

This produced pages of data, so I’m just pasting in the final section below. Piston messages from the console are on MIDI channel 7 or 8.

2023-11-24-17-51-50: INF:2561 Diag: OUT: MIDI note on: port: 828mk3 Hybrid #2, channel (1-16): 08, note (0-127): 040, velocity (0-127): 127. (Raw hex bytes: 97 28 7F.)
2023-11-24-17-51-50: INF:2561 Diag: OUT: MIDI note on: port: 828mk3 Hybrid #2, channel (1-16): 16, note (0-127): 040, velocity (0-127): 127. (Raw hex bytes: 9F 28 7F.)
2023-11-24-17-51-50: INF:2561 Diag: OUT: MIDI note on: port: Hauptwerk VST Link, channel (1-16): 16, note (0-127): 040, velocity (0-127): 127. (Raw hex bytes: 9F 28 7F.)
2023-11-24-17-51-50: INF:2551 Diag: IN: MIDI note off: port: 828mk3 Hybrid #2, channel (1-16): 08, note (0-127): 040, velocity (0-127): 000. (Raw hex bytes: 87 28 00.)
2023-11-24-17-51-50: INF:2658 Diag: IN (derived): virtual switch turned off by MIDI/key input event: 000830 'Registration: scoped comb. A01'.
2023-11-24-17-51-54: INF:2550 Diag: IN: MIDI note on: port: 828mk3 Hybrid #2, channel (1-16): 08, note (0-127): 040, velocity (0-127): 127. (Raw hex bytes: 97 28 7F.)
2023-11-24-17-51-54: INF:2656 Diag: IN (derived): virtual switch turned on by MIDI/key input event: 000830 'Registration: scoped comb. A01'.
2023-11-24-17-51-54: INF:2551 Diag: IN: MIDI note off: port: 828mk3 Hybrid #2, channel (1-16): 08, note (0-127): 040, velocity (0-127): 000. (Raw hex bytes: 87 28 00.)
2023-11-24-17-51-54: INF:2658 Diag: IN (derived): virtual switch turned off by MIDI/key input event: 000830 'Registration: scoped comb. A01'.
2023-11-24-17-51-55: INF:2550 Diag: IN: MIDI note on: port: 828mk3 Hybrid #2, channel (1-16): 08, note (0-127): 037, velocity (0-127): 127. (Raw hex bytes: 97 25 7F.)
2023-11-24-17-51-55: INF:2656 Diag: IN (derived): virtual switch turned on by MIDI/key input event: 000832 'Registration: scoped comb. A03'.
2023-11-24-17-51-55: INF:2562 Diag: OUT: MIDI note off: port: 828mk3 Hybrid #2, channel (1-16): 08, note (0-127): 040, velocity (0-127): 127. (Raw hex bytes: 87 28 7F.)
2023-11-24-17-51-55: INF:2562 Diag: OUT: MIDI note off: port: 828mk3 Hybrid #2, channel (1-16): 16, note (0-127): 040, velocity (0-127): 127. (Raw hex bytes: 8F 28 7F.)
2023-11-24-17-51-55: INF:2562 Diag: OUT: MIDI note off: port: Hauptwerk VST Link, channel (1-16): 16, note (0-127): 040, velocity (0-127): 127. (Raw hex bytes: 8F 28 7F.)
2023-11-24-17-51-55: INF:2562 Diag: OUT: MIDI note off: port: 828mk3 Hybrid #2, channel (1-16): 08, note (0-127): 037, velocity (0-127): 127. (Raw hex bytes: 87 25 7F.)
2023-11-24-17-51-56: INF:2551 Diag: IN: MIDI note off: port: 828mk3 Hybrid #2, channel (1-16): 08, note (0-127): 037, velocity (0-127): 000. (Raw hex bytes: 87 25 00.)
2023-11-24-17-51-56: INF:2658 Diag: IN (derived): virtual switch turned off by MIDI/key input event: 000832 'Registration: scoped comb. A03'.
2023-11-24-17-51-59: INF:2550 Diag: IN: MIDI note on: port: 828mk3 Hybrid #2, channel (1-16): 07, note (0-127): 001, velocity (0-127): 127. (Raw hex bytes: 96 01 7F.)
2023-11-24-17-52-00: INF:2551 Diag: IN: MIDI note off: port: 828mk3 Hybrid #2, channel (1-16): 07, note (0-127): 001, velocity (0-127): 000. (Raw hex bytes: 86 01 00.)
2023-11-24-17-52-00: INF:2550 Diag: IN: MIDI note on: port: 828mk3 Hybrid #2, channel (1-16): 07, note (0-127): 002, velocity (0-127): 127. (Raw hex bytes: 96 02 7F.)
2023-11-24-17-52-00: INF:2551 Diag: IN: MIDI note off: port: 828mk3 Hybrid #2, channel (1-16): 07, note (0-127): 002, velocity (0-127): 000. (Raw hex bytes: 86 02 00.)

What I can’t account for is the presence of messages on MIDI channel 16 – could that indicate a problem? Pedals are on channel 1, Keyboards channels 2 3 4 5, pistons channels 7 or 8.

It is the Hauptwerk VST that seems to be producing a message on channel 16; Hauptwerk is linked to Reaper via the Hauptwerk VST in the usual way, so as to add reverberation, which is not otherwise available in version 4.

Hmmm . . . .

There might be a misconception there,. A combination file doesn't contain MIDI assignments (mappings), which is what I think you mean by "links", so loading a combination won't affect those.

Thank you so much Mark! I’ve been assuming – without any evidence really – that once I’ve auto-detected my console pistons to one of Hauptwerk’s Master Scoped Combinations, that “link” would be stored as part of a combination file. By correcting my misconception you’ve saved me a huge amount of time and stress!

So – where are they stored then – the things I call links, which you are telling me should properly be referred to as “MIDI assignments (mappings)”? Can I save them by going to “File > Save any unsaved changes (Hauptwerk auto-saves anyway)”?

If those MIDI assignments disappear – which is the heart of my problem – should I be able to restore them by using “File > Restore your Hauptwerk settings or personal data from a backup”?

The MIDI mappings are saved by File | Save and should persist unless something goes wrong.

They are also written to a backup file by File | Back up and by automatic monthly or weekly backups, which can be restored using File | Restore later if the mappings are lost. So you can autodetect your pistons, make a backup, and recover them later as long as you keep your backups up-to-date with any recent changes (including registrations) since the restore includes all your user settings.

That doesn't help with why your MIDI piston mappings are going away, though. That can be caused by either Windows renaming your MIDI ports out from under Hauptwerk, or a problem with the MIDI interface or console. I don't have HW4, but I don't recall it losing settings when I used it.

“That doesn't help with why your MIDI piston mappings are going away, though”.

The sample set I’m working with - Neil Jensen’s 3-35 theatre organ – became corrupted and wouldn’t load. I had to reinstall it from the .rar files, and it turned out that one of the rar files was itself corrupted. Fortunately I managed to get the whole organ back, and I suppose it ought to have checked for settings once re-installed? The whole process was so upsetting that I can’t remember the details, but very likely that was why the piston mappings got lost. You also contributed a vital insight to that discussion:

viewtopic.php?f=4&t=21070

I will now set up some more mappings, saving as I go along for safety. If they disappear again I’ll try the restore process as discussed. I think there’s a very good chance that will solve the problem, but if it doesn’t I’ll look into the other possibilities you and Martin have explained.

Please also run File | Backup as soon as you get your MIDI mappings done, and copy the backup file and your RAR installation files onto a USB drive for safekeeping off the computer.