Topic: Possible bug

I'm playing around with the trial version rendering with the 4 keyboards mapping a 3 channel MIDI file of an organ piece not produced by Organteq, so without any imbedded registration (only channels 1, 2 and 3). If I try out different stops and then click somewhere else in the MIDI file, all the stops are pushed, and I lose whatever changes I had made. Of course I could save it before clicking in the MIDI file but this seems like a bug to me. The demo 6 channel file does not show this problem.

I found I can use Undo to bring back my changes but it does not seem correct to have to do that.

Last edited by Gilles (29-11-2019 02:20)

Re: Possible bug

Ive noticed that the 'pop-up' help menu boxes don't close very well after moving away from them(standalone version). If I go from one item to another and the boxes appear, my screen gets rather full of them. They start to close or fragment as I pass over the boxes themselves, but not the item it descibes. I notice this behavior and/or the inability to even read the help hints in pianoteq pro as well. In the vst3 version of both pianoteq and organteq, the popup hints are not visible, tho a transparent shaded box seems to open in the ui. Small issue, does not prevet me from using the products.

Last edited by Shanesplanet (29-11-2019 05:16)

Re: Possible bug

Gilles wrote:

I'm playing around with the trial version rendering with the 4 keyboards mapping a 3 channel MIDI file of an organ piece not produced by Organteq, so without any imbedded registration (only channels 1, 2 and 3). If I try out different stops and then click somewhere else in the MIDI file, all the stops are pushed, and I lose whatever changes I had made. Of course I could save it before clicking in the MIDI file but this seems like a bug to me. The demo 6 channel file does not show this problem.

Hi Gilles,
It is a bug. Thanks for reporting.

Re: Possible bug

I noticed too some small bugs :
When he has to play the 4 quarter note chords with the MIDI player, he plays these notes as a half note, anyway, we hear the bug if you play this music : https://forum.modartt.com/uploads.php?file=uhjb.mid
When I start the software, it is a bit long to load, I think it can be corrected, it's the same when I change the temperament...
The French translation is not complete, hmm...
I hope we can play with the computer keyboard !

Re: Possible bug

When playing fairly full chords on fairly full registrations, with some of the stops on the right-hand side drawn also, such as Great Quint and Great 2'... the music is frequently accented by sharp metallic sounds (like a cricket chirping) when a new chord is struck.  Very distracting!

Listening just to Great Quint alone (Cathedral Organ with Church Reverb), the Eb above middle-C seems to have some of this metallic component as a part of every attack.

Maybe not a "bug" per se, but perhaps an "accurate" physical modelling of a Pipe which itself needs re-voicing.  No Organ Builder would permit such sounds to exist in any Organ after careful and proper Voicing!

For a First Release, this Software is truly wonderful, and shows great potential.  Looking forward to enjoying it more fully when some of these "rough spots" get successfully smoothed out!

Re: Possible bug

In my configuration (Windows 10), the trial version has a very strange behaviour.
1) The application doesn’t recognise the clicks of my mouse. Only the touch screen allows an action.
2) A right-click on a keyboard doesn’t allow the auto-detect function. It works as a left-click and the note playing during this action stays pressed without any possibility to stop it.
3) Touching a pop-up menu doesn’t work, but when I release my finger the menu appears very briefly.
I’ve uninstalled and reinstalled Organteq 3 times, without any success.

Re: Possible bug

Have you checked the CPU load diagram in settings? As written I am running Organteq on a fairly high end system (maybe too new??) and it's a beast in consuming cpu load. Only with 32 note polyphonic your described effect can be eliminated.

Galuwen


OrganoPleno wrote:

When playing fairly full chords on fairly full registrations, with some of the stops on the right-hand side drawn also, such as Great Quint and Great 2'... the music is frequently accented by sharp metallic sounds (like a cricket chirping) when a new chord is struck.  Very distracting!

Listening just to Great Quint alone (Cathedral Organ with Church Reverb), the Eb above middle-C seems to have some of this metallic component as a part of every attack.

Maybe not a "bug" per se, but perhaps an "accurate" physical modelling of a Pipe which itself needs re-voicing.  No Organ Builder would permit such sounds to exist in any Organ after careful and proper Voicing!

For a First Release, this Software is truly wonderful, and shows great potential.  Looking forward to enjoying it more fully when some of these "rough spots" get successfully smoothed out!

Re: Possible bug

galuwen wrote:

Have you checked the CPU load diagram in settings?

Thanks for the tip.  I checked all of my settings... Sample Rate, Buffer Size, Polyphony, while following the CPU load.

No problems with any of these settings.  I can certainly PRODUCE an audio overload, by disabling multi-core support, but that effect sounds nothing like what I am referring to.  I will be posting further discussion in the "suggestions" thread.

Re: Possible bug

I think my first issue is the same as the one the OP is reporting. Here's a procedure to replicate it:

1. Change One keyboard simple to One keyboard split.
2. Record a short piece using the range of the keyboard.
3. Save the file and reopen it.
The change is lost, so playback sounds different.

I tested in Cubase 10 and Cakewalk with the same results.

Re: Possible bug

Once a preset is loaded, if we change anything that is on the second line of buttons (diapason, temperament and so on) the preset is marked as modified but not for changes on the stops, combinations and coupling. I understand that a lot of trial goes on there that are not meant to be saved as preset but this is dangerous since we can save unwanted changes without knowing. Every possible modification to a preset should be flagged in my opinion, as it is in Pianoteq.

EDIT: Another thing that bothers me: If I save a preset containing many combinations and reload it later, it starts in the last combination I made but does not light the associate combination button so I don't know what's the initial state.

Last edited by Gilles (02-12-2019 21:55)

Re: Possible bug

I think that saving and opening custom presets is too broken to use at this point. When you reload a preset, those settings don't overwrite current settings.

1. Arrange some drawknobs and save the preset. Make a screenprint.
2. Load a preset with a different arrangement of drawknobs.
3. Compare the UI to the screenprint and listen.

I've also tried using General Cancel between steps 1 and 2.

Last edited by mitch_i (03-12-2019 00:47)

Re: Possible bug

mitch_i wrote:

I think that saving and opening custom presets is too broken to use at this point. When you reload a preset, those settings don't overwrite current settings.

1. Arrange some drawknobs and save the preset. Make a screenprint.
2. Load a preset with a different arrangement of drawknobs.
3. Compare the UI to the screenprint and listen.

I've also tried using General Cancel between steps 1 and 2.

Mmmm, I've tried this and it seems to work correctly for me on MacOS. There may be a small delay at times though before all knobs are reset, not always. I find though that the hourglass that is supposed to indicate the delay sometimes does not appear and stops move later than expected (quite rare though, usually no hourglass means an instantaneous change). There is probably a lot of small situations where the part of the GUI that was inherited from Pianoteq is not quite in the same conditions (longer adjustments needed).

Last edited by Gilles (03-12-2019 02:19)

Re: Possible bug

For some parameters changes, the pipes data have to be updated. That occurs when:

  • launching the app the first time

  • changing the sound card settings

  • changing the global tuning options (diapason, temperament, global note-per-note tuning)

The internal computations of all the pipes (more than 2500) currently takes about 5 sec. on a regular laptop. During that time the engine still works, but the interface is somehow frozen. As long as the hourglass is shown, the internal update is still running. We are working on speeding that part up and on making the user experience smoother.

Note that when you get back to a previous tuning setting, the change is faster as internal computations are cached.

Re: Possible bug

mitch_i wrote:

I think that saving and opening custom presets is too broken to use at this point. When you reload a preset, those settings don't overwrite current settings.

1. Arrange some drawknobs and save the preset. Make a screenprint.
2. Load a preset with a different arrangement of drawknobs.
3. Compare the UI to the screenprint and listen.

I've also tried using General Cancel between steps 1 and 2.

The live state, i.e. stops switches, couplers switches, pedals, ... is saved in the fxp files (presets). When rendering from the standalone version with Organteq MIDI files, the live state remains coherent. When rendering through a DAW, you have to check that the live state you are starting with is the good one.

That, combined with a long update, might look like a broken preset.

Re: Possible bug

In the meantime, Roman, why not just make the hourglass an entirely contrasting color to the rest of the GUI, like red or blue?  Even when I see it I hardly see it!

- David

Re: Possible bug

Ludu wrote:

In my configuration (Windows 10), the trial version has a very strange behaviour.
1) The application doesn’t recognise the clicks of my mouse. Only the touch screen allows an action.
2) A right-click on a keyboard doesn’t allow the auto-detect function. It works as a left-click and the note playing during this action stays pressed without any possibility to stop it.
3) Touching a pop-up menu doesn’t work, but when I release my finger the menu appears very briefly.
I’ve uninstalled and reinstalled Organteq 3 times, without any success.

Does anyone understand why I can’t run the trial version normally?

Re: Possible bug

Ludu wrote:

In my configuration (Windows 10), the trial version has a very strange behaviour.
1) The application doesn’t recognise the clicks of my mouse. Only the touch screen allows an action.
2) A right-click on a keyboard doesn’t allow the auto-detect function. It works as a left-click and the note playing during this action stays pressed without any possibility to stop it.
3) Touching a pop-up menu doesn’t work, but when I release my finger the menu appears very briefly.
I’ve uninstalled and reinstalled Organteq 3 times, without any success.

Hi Ludu,
Thanks for reporting. Can you send a support request and mention your config including your screen settings?

Re: Possible bug

Roman wrote:

For some parameters changes, the pipes data have to be updated. That occurs when:

  • launching the app the first time

  • changing the sound card settings

  • changing the global tuning options (diapason, temperament, global note-per-note tuning)

The internal computations of all the pipes (more than 2500) currently takes about 5 sec. on a regular laptop. During that time the engine still works, but the interface is somehow frozen. As long as the hourglass is shown, the internal update is still running. We are working on speeding that part up and on making the user experience smoother.

Note that when you get back to a previous tuning setting, the change is faster as internal computations are cached.

One question about optimization (and max polyphony): On my 2013 quad-core MacPro, Pianoteq uses mostly the first two cores even with complex input and I never get audio overload. I notice Organteq uses all four cores quite evenly, something I find normal given the parallel nature of organ pipes. Does that mean that the more core the better for Organteq, contrary to Pianoteq where it does not really matter?

I find I am limited to a bit over 160 voices when using Tutti, does it mean that the maximum 512 voices (on Tutti) is attainable only with over 12 cores (or less with a newer faster cpu than my 3.7 GHz Xeon E5)?

(That seems to be the case since Bruno (bm) reports in another thread 350 polyphony using an i9-9900k which has 8 cores if I'm not mistaken)

Last edited by Gilles (03-12-2019 15:06)

Re: Possible bug

Gilles wrote:

I find I am limited to a bit over 160 voices when using Tutti, does it mean that the maximum 512 voices (on Tutti) is attainable only with over 12 cores (or less with a newer faster cpu than my 3.7 GHz Xeon E5)?

(That seems to be the case since Bruno (bm) reports in another thread 350 polyphony using an i9-9900k which has 8 cores if I'm not mistaken)

My system does fine with 512 polyphony.  Intel Core i7-6900K @ 3.2 GHz, 8 cores.

Re: Possible bug

I think I discovered a bug with the MIDI player... !

Re: Possible bug

That's not a particularly detailed report

Hard work and guts!

Re: Possible bug

I have the impression that key couplings are not all perfectly synchronized...
And, I have the impression the MIDI player reads badly the notes that are repeated...

Re: Possible bug

Not really a bug, but organteq takes a rediculously long time to load an instance, in compare to any and all of my other plugins. i-7quad 4ghz 32g ram  win7-64. I think I read mention of this is going to be addressed, but i cant find that referance. Still having hte popup menu bug, but i am thinking it may be something in my computer cuasing it, as I recall it happening on another vst i own. Perhaps a java or some other outdated driver i have. I am NOT up to snuff on most these things, obviously. Still loving it tho! Using it as the low note on a bass line with a voyager layered is amazing! Loving those noise fx for sure!

Re: Possible bug

Loads in 7 seconds over here.

(Also Java or any drivers in your system in 99.9% of cases have nothing to do with plugin loading time.)

Last edited by EvilDragon (14-12-2019 22:59)
Hard work and guts!

Re: Possible bug

EvilDragon wrote:

Loads in 7 seconds over here.

(Also Java or any drivers in your system in 99.9% of cases have nothing to do with plugin loading time.)

under 4 seconds for pianoteq to open on a blank reaper project, 11 seconds for organteq, blank reaper project.  3 sconds for Kontakt and 6 seconds for 7gb sample load. So I guess Organteq is almost on par with a 7gb sample based instrument. Maybe I'm just so damn eager it seems longer. Maybe a quad core 4.0ghz ssd system isnt what it used to be. Its just obvious when I open a project containing organteq, as the project hangs a bit and its the last to finish. Custom built win7 64 desktop (intel), fwiw.

Last edited by Shanesplanet (18-12-2019 02:57)

Re: Possible bug

Note that Organteq has to initialize and reserve resources for about 1000 virtual pipes upon instantiation. Pianoteq on the other hand has 5 times less strings to initialize. Sometimes, things just need a bit more time

Hard work and guts!

Re: Possible bug

With version 1.0.2, I find that the problem I reported on Dec. 2 about preset sharing is fixed. The keyboard configuration is now saved.

Thanks, Modartt.

Re: Possible bug

HI
Maybe I missed something

On my PC, I use 2 displays
Qhen I activate the full screen mode on the main display. all is ok

If I try the same operation with Organteq GUI on the second display
The right-click on the stops knobs does not work in full screen mode (but it works in normal mode)

Re: Possible bug

I recently acquired Organteq, and use it with four keyboard layout (35 years old reliable analog Heyligers organ with 13 stops over 2m+p, extended with 3rd manual and midi). For study the Heyligers remains my preference, but Organteq is an excellent add-on, with a similar direct sound and small latency.

Because of compatibility between the two sets, the A/D converter for the expression is only used over a limited range, 0.32 to 0.91, rather than the initial setting 0 to 1. I filled in the actual A/D outputs in Organteq, but see that the expression pedal remains stuck at both ends as if 0 to 1 is still held, as it can be moved further with a mouse. Please help or solve.

Analog Heyligers E1 organ, 13 stops, 2 manuals + pedal
December 2018 extension: third manual and midi
Software Organteq, GrandOrgue, Sweelinq

Re: Possible bug

Alex_vD wrote:

Because of compatibility between the two sets, the A/D converter for the expression is only used over a limited range, 0.32 to 0.91, rather than the initial setting 0 to 1. I filled in the actual A/D outputs in Organteq, but see that the expression pedal remains stuck at both ends as if 0 to 1 is still held, as it can be moved further with a mouse. Please help or solve.

Hi Alex,

Thank you for reporting. You can restrict the input range of the MIDI controller: in the MIDI mapping window, click on the controller and choose Enter custom range... in the Restict the value range of Controller # menu.

Restict_CC11_1


You have to provide the new range with an integer MIDI value (i.e. in the range [0, 127]). So in your case the new bounds would be 0.32*127 = 40 and 0.91*127 = 116.

Restict_CC11_2


Please contact support for further assistance.

Roman