Unexpected effect of a setting in Stop assignments.

Thanks again, for the most recent reply, Martin.

I shall be trying your suggestion of turning off the MIDI IN ports shortly (i.e. doing that within Hauptwerk's dialogues), and I'll get back to you with the result. However, in the meantime, I will answer your direct questions below:

Do you know for certain which MIDI OUT message(s) could turn on the problematic stop on your MIDI console?

Answer: I suspect that any MIDI switch/led on the stop jamb configured as "ON by default" in the "first" organ will caused the same switch/led to illuminate in the "second" organ provided that it is not configured otherwise in the second organ.

To satisfy myself that I am not going bananas with this ( !! ) I am going to double-check my reply to your question by setting up some other organs as "culprit-first", and using St. Anne's as the guinea pig "victim-second". I believe that any MIDI switches with OUTPUT configured as "ON by default" will turn on the same corresponding LED if said switch/led is unconfigured in the next organ loaded.

I'll get back to you as soon as I can with the result, and possibly some further information about the issue.

Many thanks indeed!

Ken
--
Kenneth A. Spencer
===============

I had a few moments...

1. I loaded St Anne's.
2.. I autodetected the Grt & Ped Combs coupled to a stop tab.
3. I set the default to be ON.
4, The stop tab works fine. I turned the stop off.
5. I unloaded the organ leaving the stop off.
6. I loaded St Anne's again and the organ loaded with the stop drawn and the console tab illuminated.
7. I drew the Swell to Pedal stop and its assigned stop tab illuminated.
8. I did not do anything else (so the Swell to Pedal and the Grt and Pedal Combs coupled tabs were both illuminated and the stops on.).
9. I unloaded St Anne's..
10. The Swell to Pedal tab went off but the Grt & Pedal Combs coupled tab remained illuminated.
11. I loaded another organ which has no tabs configured - it only uses the touch screens.
12. The Grt & Pedal Combs coupled tab remained lit. It would not go out with the General Cancel for the newly loaded organ.

I think this suggests that "always on stops" are not being "turned off" when the organ is unloaded.

Does this help?

Iain

[mdyde]

Thanks, Ken.

Answer: I suspect that any MIDI switch/led on the stop jamb configured as "ON by default" in the "first" organ will caused the same switch/led to illuminate in the "second" organ provided that it is not configured otherwise in the second organ.

To satisfy myself that I am not going bananas with this ( !! ) I am going to double-check my reply to your question by setting up some other organs as "culprit-first", and using St. Anne's as the guinea pig "victim-second". I believe that any MIDI switches with OUTPUT configured as "ON by default" will turn on the same corresponding LED if said switch/led is unconfigured in the next organ loaded.

It trying that test, it would be helpful if you could do it as follows, please:

- Launch Hauptwerk via a 'spare' 'Hauptwerk (alt confg N)' configuration (desktop shortcut) .

- Use 'File | Revert all settings to factory defaults' within that configuration to ensure that all of its settings are (still) at their defaults. (Hauptwerk will exit after you revert the settings.)

- Re-launch that same 'Hauptwerk (alt confg N)' configuration, and select the audio device and the minimum MIDI IN and OUT ports necessary for reproducing the problem.

- Stick to using that configuration for your testing.

- For the two organs, if possible, stick with just MDA-supplied organs (e.g. St. Anne's, any of the CODM organs that are installed as standard with Hauptwerk, or the MDA Salisbury, since I understand you have the latter). That way, if you need to send us a diagnostic file we could try loading the organs with your MIDI settings to examine and test them, and also we know that those organs don't have any unusual logic within their organ definitions.

- For the organs used in the test, configure/auto-detect as few virtual controls as possible to reproduce the problem, so that it's easier to identify the relevant MIDI messages that are responsible in the log.

Answer: I suspect that any MIDI switch/led on the stop jamb configured as "ON by default" in the "first" organ will caused the same switch/led to illuminate in the "second" organ provided that it is not configured otherwise in the second organ.

It should be absolutely impossible for organ-specific MIDI settings or states from one organ to affect the MIDI output sent by any other (provided that their OrganIDs are different). However, note that also it's possible to control Hauptwerk's master combinations and menu functions on a 'for all organs' basis, e.g. by auto-detecting those functions with the 'for all organs' right-click option on their control panel buttons or via the "General settings | MIDI/key triggers for master pistons and menu functions (for all organs, unless overridden)" screen. (However, even if MIDI settings for master combinations are configured on a 'for all organs' basis, combination sets/files, registrations, and organ's virtual control states will still always be entirely organ-specific.)

It would be key to identify from the MIDI log the specific MIDI output message(s) from Hauptwerk that are causing the problem with your console, preferably tested as above (with minimal settings and MDA organs, so that we can load and look at your settings meaningfully if needed).

[mdyde]

I had a few moments...

1. I loaded St Anne's.
2.. I autodetected the Grt & Ped Combs coupled to a stop tab.
3. I set the default to be ON.
4, The stop tab works fine. I turned the stop off.
5. I unloaded the organ leaving the stop off.
6. I loaded St Anne's again and the organ loaded with the stop drawn and the console tab illuminated.
7. I drew the Swell to Pedal stop and its assigned stop tab illuminated.
8. I did not do anything else (so the Swell to Pedal and the Grt and Pedal Combs coupled tabs were both illuminated and the stops on.).
9. I unloaded St Anne's..
10. The Swell to Pedal tab went off but the Grt & Pedal Combs coupled tab remained illuminated.
11. I loaded another organ which has no tabs configured - it only uses the touch screens.
12. The Grt & Pedal Combs coupled tab remained lit. It would not go out with the General Cancel for the newly loaded organ.

I think this suggests that "always on stops" are not being "turned off" when the organ is unloaded.

Does this help?

Iain

Thanks, Iain. (I see that we replied at the same time.)

Yes -- if a virtual stop/switch is configured to default to on then it won't get turned off via MIDI output when unloading the organ. Hence if a MIDI stop that's auto-detected to that virtual stop/switch for one organ isn't auto-detected to anything in subsequently-loaded organs then that MIDI stop would potentially still be on.

Perhaps it is indeed simply that which Ken's noticing?

Greetings, Martin:

I have more news:

I loaded the Groton organ (simple organ, loads of unassigned Switch/Leds to experiment on).
I changed just one setting: the Gt to Pedal coupler was moved from its previous stop Switch/Led to stop Switch/Led 59 and that setting was amended to "Switch defauts to 'ON'". Then I saved the settings, and of course the virtual drawstop popped out and the LED lit up as it should - defaulting to ON as it should. I popped it back in.

Then I loaded the Little Waldingfield - also simple, few stops/divisions lots of spare Switch/Leds. All was well UNTIL I pulled a stop! Whether I pulled a virtual stop or a Switch/LED the LED in Switch/Led 59 that was previously set as Default to ON in the Groton, but unassigned in the Little Waldingfield lit up. Closing the virtual drawstop or the Switch pulled before and the Led went off.

I accept that I could have made a really terrible error in my wiring, but this issue seems to affect both jambs and any Switch/Led that is assigned and defaults to ON in the first organ, and is unnassigned in the second organ. Of couse it could be the Decoder, but it affects both stop jambs which have their own En+De+coders. Maybe it is an odd feature of the Decoder design. One other possibility is that somehow the "Defaults to ON" value is misbehaving in the Hauptwerk code somewhere.

I would be absolutely grateful beyond expression if you were able to use a console with MIDI encoded/decoded Switch/Led pairs to repeat this pattern. It seems that any organ pair, and any Switch/Led pair on any Note Number of the En-De-coder will suffer this effect if:
1. an organ is loaded with a control set as "Default to ON"
AND
2. a second organ is loaded with that same control unassigned.
Otherwise everything is fine. My thinking is that the effect can now be summarised in just the two above lines.

It is interesting because I noted this "feature" or "buggette" quiite a few years ago, and never really mentioned it. But with larger organs as the use of more features in my console I have been increasingly interested in the possible causes.

ADDENDUM: I should have added: "I will try to carry out your request to disable all MIDI Inputs tomorrow, and run through the same routine as above,collecting the Activity Log, and will report back.

Sorry for the long message, but I am really keen now to get your opinion on it!

Very best wsihes and very many thanks!

Ken
--
Kenneth A. Spencer
===============

[mdyde]

Thanks, Ken.

If Iain's point (i.e. that default-to-on MIDI stops don't get turned off when unloading an organ) doesn't explain it, please try the last test I mentioned above (use a spare 'all-default' configuration with minimal MIDI settings and two MDA organs), and see if you can still reproduce it.

If the problem doesn't occur in the second organ until the moment that you press something on your MIDI console, and if Hauptwerk is indeed somehow responsible, then it must be due to a MIDI output message being sent from Hauptwerk at that point in time, which would show up in the logged MIDI. We need to know exactly that MIDI message is from the log.

If Hauptwerk *isn't* sending a MIDI output message to turn the MIDI stop on at that point in time then I think you must have a MIDI decoder problem (or some kind of MIDI hardware/wiring/firmware problem).

Well....
1. I loaded St Anne (which has the Grt Ped Combs Coupled as always on in my configuration).
2. Loaded Prib (small chamber organ). Grt &Ped Combs remained on.
3. Cleared Gr &Ped Combs via the lit tab.
4. Drew a stop on Prib via the corresponding tab on the console.
5. The stop came on and Grt & Ped Combs Coupler lit and it remained on when all the stops are cleared.
6. Cleared the Grt Ped Combs tab.
7. Tried a stop on Prib which is not autodetected to a tab and the Grt Ped Coms tab did not light. It only lights for stops which are autodetected to a tab.

Iain

[mdyde]

Hello Iain,

Do I understand correctly that you mean that you *can* reproduce Ken's problem?

1. I loaded St Anne (which has the Grt Ped Combs Coupled as always on in my configuration).
2. Loaded Prib (small chamber organ). Grt &Ped Combs remained on.
3. Cleared Gr &Ped Combs via the lit tab.
4. Drew a stop on Prib via the corresponding tab on the console.
5. The stop came on and Grt & Ped Combs Coupler lit and it remained on when all the stops are cleared.

Was your 'Gt & Ped coupler' MIDI coupler auto-detected to any virtual controls in the Prib?

Thanks, Iain, and Martin.

Martin: indeed I will try your routine suggested with MIDI INPUTS disabled. I will try it first with a pair of very small organs to reduce the amount of data to plough through in the Activity log.

Then I will try it with two of the slightly smaller MDA organs as you suggest: The St. Anne's and the St. Eucaire Metz will then be useful candidates to test.

It may seem odd that it took so long for me to ask about this apparent issue. But the reason is that there are quite a lot of Anches and other controls in the Caen which I prefer to set "default to ON" so it has proved irritating to see a series of unnallocated Switch LEDs lighting up all over the place!

Iain: I think that we are definitely homing in on the problem! Thanks for your very helpful comment. I think the evidence does point to a bit of code being missing to ensure that "defaulted to ON" controls are managed (i.e. properly turned off) when the instrument is to be unloaded. Probably a trivial bug-fix.

Ken
--
Kenneth A. Spencer
===============

Martin:
Yes - I can produce what appears to be the problem Ken has found.
No the tab used for the Grt&Ped Combs Coupled in St Anne's was not assigned to any control in the Prib.

Just to be certain ..
1. Loaded St Anne's. It has the Grt&Ped Combs coupled stop autodected to a light tab and is set to be ON when the sample is loaded. It worked as it should.
2. Loaded the Prib organ.
3. Reset the midi for Prib. Autodetected the keyboard.
4. Operated one of the stops from the screen. The tab used for the Grt&Ped Combs Coupled stop did not light.
5. Autodetected one of the stops of Prib. As soon as the autodetect was complete the tab used for the GRt&Ped Combs Coupled lit.
6. Cleared (touched the tab for) the G&PCombs Coupled tab - light went out.
7. Turned on the stop autodetected by the physical tab:: the stop came on and lit its tab and the Grt&Ped Combs tab illuminated.
8. Turned stop off and cleared the Grt&PCombs tab (light went out).
9. Turned the stop on from the touchscreen the stop came on and its tab lit, and the Grt&P Combs tab lit.
10. Turned the stop off and cleared the Grt&Ped Combs tab. No tabs illuminated.
11. Turned on another stop (with no midi assignment), Grt&Ped Combs tab did not light.

Seems that
i. if the previous organ loaded had a stop assigned to a midi illuminated tab which is set to be ON when the organ starts,
ii. on the organ now loaded, turning a stop on, which has a midi tab autodetected to operate the stop, causes the tab assigned, in the previous organ to the stop that is set to be ON when the organ loads, to be illuminated.

I turned on midi logging and operated the tab to turn on the stop and the Grt&Ped Coms tab lit. Turned off logging. Had a quick look at the log , would need to spend some time reviewing it to deduce what was going on. (There seemed to a lot of midi output activity to other devices...). I created a diagnostic file should you want to see it..

Ken: Pleased to try and help.

Iain

Martin and Iain

I am hoping to get time to produce for Martin the definitive "smoking gun" on this issue, later this morning, although I also have some music to prepare for a service on Sunday and a funeral on Monday and I'm out all day tomorrow. But I'd like to repeat (what is very similar to Iain's conclusion) that this little issue occurs under just two conditions:

kaspencer » Wed 31st Jan, 2024 7:45 pm
1. any organ is loaded with a MIDI Switch/LED control for a stop/coupler set as "Default to ON"
AND
2. a second organ is subsequently loaded with that same MIDI Switch/LED control unassigned.
The fault will be shown as soon as any assigned MIDI Switch/LED control is engaged.

I hope to be able to get the offending lines of a MIDI Activity Report, although it should be easy for anyone to replicate the fault. If I had the code I'd look at the code which manages the MIDI data behind the "Default to ON" tick box as it seems to me that that item's MIDI status data is not cleared ready for a different organ to be loaded.

I'll hope to report back later today!

Best wishes, gentlemen both,

Ken
--
Kenneth A. Spencer
===============

[mdyde]

Thanks Ken and Iain.

I'll try again to reproduce it shortly.

Report of progress so far:

1. All MIDI Inputs are currently enabled;
2. MDA SS St. Eucaire loaded;
3. Lt Jamb MIDI Switch/Led on Note 035 control allocated to Recit Anches and set to "Defaults to ON";
4. St. Anne's Moseley loaded - Lt Jamb MIDI Switch/Led on Note 035 is unassigned;
5. A MIDI Switch/LED assigned to a stop coupler was pressed;
6. The unassigned Switch/LED on Note 035 illuminated.
7. The St. Anne's organ was unloaded.

THEN:
1. All MIDI Inputs were disabled;
2. MDA SS ST. Eucaire loaded;
3. Recit Anches MIDI Switch/Led control still allocated to Lt Jamb Note 035 and set to "Defaults to ON";
4. St. Anne's Moseley loaded - Lt Jamb MIDI Switch/Led on Note 035 is unassigned;
5. A Virtual drawstop was selected on-screen (MIDI Switch/LED are unavailable);
6. The unassigned Switch/LED on Note 035 DID NOT illuminate.
7. The St. Anne's organ was unloaded.
COMMENT: MIDI Input (NOT virtual drawstop action) is required to produce the fault.

TO FOLLOW
I shall shortly generate an Activity Log with MIDI INPUT disabled and report back.

I hope this helps, but it only confirms my conclusions drawn some time ago!

Best wsishes,

Ken
--
Kenneth A. Spencer
==============

Next steps performed as previously requested ...

Activity logging set to ON.
1. All MIDI Inputs are disabled;
2. MDA SS St. Eucaire was loaded;
3. Activity logging was continued;
4. Noted: "Defaults to 'ON'" was unset,
though I'm sure I left it set. Corrected setting to ON.
Could be due to MIDI IN disabling?
5. Loaded MDA SS St. Anne's Moseley;
6. Activity logging was continued;
7. I manually activated a virtual drawstop ON SCREEN (MIDI IN is disabled anyway);
8. Switch/LED control on Note 035 on left Jamb ILLUMINATED.
NB.: This was with MIDI Inputs disabled!

I conclude that MIDI Input enablement is not necessary for this fault to occur

I will obtain the Activity Log and trim it to the above session, and see what it reveals. I will pass on any lines of interest in the report.

Ken
--
Kenneth A. Spencer
==============

[mdyde]

Thanks, Ken.

I'm still unable to reproduce this. I don't have a MIDI console with moving/illuminated drawknobs/tabs, but I'm using a Novation Launchpad with its buttons configured to send and receive MIDI note-on/off messages, which should be equivalent to using a MIDI console that sends/receives note-on/off messages.

With all of MIDI settings (globally and for each organ) reset fully to defaults, I auto-detected just two stops in the first organ, with one of those defaulting to on, and in the second organ I auto-detected just one virtual stops to one of the same MIDI stops (Launchpad buttons) that were auto-detected in the first organ (the one that isn't configured to default to on), but left the other MIDI stop/button unassigned, and I then switched back and forth between the two organs multiple times. On no occasions did any unassigned MIDI stops/buttons illuminate, and no spurious MIDI messages were sent, regardless of what MIDI stops/buttons I pressed in the second or first organs, and regardless of what virtual controls I clicked on.

I also tried assigning multiple stops in the first organ, with some defaulting to on and some not, and assigning multiple stops in the second organ, leaving some of the MIDI stops which were assigned and defaulted to on in the first organ unassigned in the second, and everything still appears to be fine (in terms of illuminated MIDI stops and logged MIDI OUT messages). Likewise, defaulting stops on or not in the second organ didn't produce any incorrect results for me.

Activity logging set to ON.
1. All MIDI Inputs are disabled;
2. MDA SS St. Eucaire was loaded;
3. Activity logging was continued;
4. Noted: "Defaults to 'ON'" was unset,
though I'm sure I left it set. Corrected setting to ON.
Could be due to MIDI IN disabling?
5. Loaded MDA SS St. Anne's Moseley;
6. Activity logging was continued;
7. I manually activated a virtual drawstop ON SCREEN (MIDI IN is disabled anyway);
8. Switch/LED control on Note 035 on left Jamb ILLUMINATED.
NB.: This was with MIDI Inputs disabled!

I conclude that MIDI Input enablement is not necessary for this fault to occur

Straight after carrying out that test (please repeat it if needed, ideally with nothing auto-detected except for the relevant stops, so as to minimise spurious logged MIDI data), without touching any other virtual or MIDI controls (so as to try to limit the logged MIDI to the relevant events), please use "File | Backup your settings ..." (that will contain your settings for all organs, not just the currently-loaded one), and send the file to support[at]hauptwerk.zendesk.com with a note asking Francois to pass it on to me.

That way I should be able to load those two organs with your MIDI settings and compare your logged MIDI events with your settings in those organs. Please also indicate which of the four configurations you were using for the test (since a backup will contain all four).

Thanks again.