_________________________________(Mailing List Recent Archive)
This web-page allows you to view most posts sent to the jOrgan-User Mailing List from mid-July 2021 onwards, and is read-only.
To post a Reply or start a new Topic, you need to be registered with the Mailing List. For instructions on registering and using the
Mailing List, click HERE. Your post will normally be sent immediately to all subscribers, but will not appear here until the web-page
is updated. This is done manually, and perhaps once a week. Some topics or posts may not appear, if considered not archive material.
To contact the Compiler of this web-page, send a post with the Subject Line, “To the jOrgan Link Compiler”.
The main jOrgan-User archive can be accessed HERE, or the jOrgan-Sound archive HERE. If you click on a month, you need to scroll
down to see the result. Don’t overlook the Search facility at the left. After doing a successful search, you will be given a list of posts
where the key wording is found. After you select a post, you will find at the bottom of the page an opportunity to view the thread
where that post belongs.
SCROLL DOWN TO VIEW THE CURRENT LIST
To view the newest post, click on the coloured text
PAST ARCHIVES (Click to access)
2025 Jul-Dec______2025 Jan-Jun_____ 2024 Jul-Dec______2024 Jan-Jun
2023 Jul-Dec______2023 Jan-Jun______2022 Jul-Dec______2022 Jan-Jun
jOrgan website for new users___Sticky Post
AI, Can the AI named ... [OFF TOPIC]___By Lynn Walls Newest post on 2026-05-31 4 replies AI information By Freeman Gilmore Most recent post on 2026-05-31 1 reply AI, Can the AI named Claude create a virtual pipe organ application using midi as a basis? By JohnBeach Most recent post on 2026-05-31 4 replies AI and jOrgan By JohnR Most recent post on 2026-05-30 Reverb for jOrgan By John Beach Most recent post on 2026-05-28 7 replies Linux driver for old Dell touch screen__By Ross Willson Most recent post on 2026-05-12 5 replies jOrgan Files on Sourceforge____By Lynn Walls Most recent post on 2026-05-11 3 replies Configuring swell____By Andrew Larkham Most recent post on 2026-04-29 4 replies Piston indicator lights and launchpad support Most recent post on 2026-04-29 4 replies AI (off-topic) Most recent post on 2026-04-19 1 reply Looped Modulation in Fluidsynth reverb By JohnBeach Most recent post on 2026-04-19 1 reply Traps control By phil Most recent post on 2026-04-14
SAM on my Wyvern 4 manual jOrgan____By Stephen Rowley____Most recent post on 2026-04-11____13 replies
Save As By JohnR__Most recent post on 2026-02-21___43 replies full screen not full screen__ By Paul StringfellowByMost recent post on 2026-02-21___48 replies Seasons Greetings____By JohnR__Most recent post on 2026-01-12___44 replies SF2 Player plugin____By Dr. Mark Bugeja MD__Most recent post on 2026-01-06___42 replies
From JohnR
This website has been created to make it easier for internet users to discover jOrgan and to start
using it. Its main pages are HOME, STARTING, ADDING, GALLERY, F.A.Q., THIS SITE and CONTACT.
The web address is https://jorgan.info
JohnR
______________________________________________________________________________________________________________ *
Fluidsynth Reverb Fluidsynth Reverb From John Beach on 2025-08-24 22:41 Recently, I experimented with Fluidsynth Reverb elements Room, Width, Damping and Level. As we know, with the setting of each individual element, the "whole" of the sound-effect equals the sum of those settings. Level has always produced a frequential looped effect relative to its graduated increase, above approximately 0.15. + or -. Level, unfortunately, has the effect of muffling sound, the lower the setting is. This is unrealistic as compared with an acoustic environment where softness and loudness do not impact the reverberative space, or tone quality is not changed by that space. The interrelationship of the elements of Fluidsynth reverb is such that the audio waves of the engaged stops being played are manipulated in the "field effect" created by the parameters of reverb as space size. Level is not realistic. It is comparable to degree of "on" and "off." Using the green light, slider elements on the "Skin" tab in jOrgan construct mode, I tried settings of "Room"=4 (green lights lit), Width=3, Damping=2, and Level=8 I got a realistic fullness of audio without the frequential loop effect. The sound was satisfying, but a relative,acoustic- space size was not really discernible. I believe the solution to this problem in jOrgan is to have soundfonts made from actual sample sets of recorded stops from real pipes, like Hauptwerk has. In such a case, the reverb elements can be avoided completely. John Beach _________________________________________________________________________________________________________* END OF THREAD Seasons Greetings Seasons Greetings From JohnR on 2025-12-31 2025 at 21:41 A Merry Christmas and a Happy New Year to all! JohnR ______________________________________________________________________________________________________________ From John Beach on 2025-12-31 at 00:27 AM Indeed, a Merry Christmas and Happy New Year! John Beach ______________________________________________________________________________________________________________ From Dr. Mark Bugeja MD on 2025-12-31 at 01:33 AM A Blessed Christmas to all and best wishes for the New Year. Mark Bugeja from Malta. ______________________________________________________________________________________________________________ From Peter Hanlon on 2025-12-31 at 02:10 PM A Happy New Year to the jOrgan Community. Has anyone tried to run jOrgan on an Arduino UNO Q yet? It might have appeal for those of us who try to manage hardware like keyboards etc with an Arduino sketch and then pass it to a Raspberry PI running jOrgan. It potentially gets everything back in the one box. From PeterH ______________________________________________________________________________________________________________ POST From JohnR on 2026-01-12 at 22.15 On Fri Jan 02 2026 Peter Hanlon wrote: "A Happy New Year to the jOrgan Community. Has anyone tried to run jOrgan on an Arduino UNO Q yet?" Thank you, Peter, and a Happy New Year to you! No answer to your question probably means that the answer is, "No". The "Organ Forum" certainly provides a lot of guidance for people who use a standard Arduino instead of a commercially- produced encoder, but I don't recall many jOrgan users who have indicated that they have used it for that, let alone the latest very recent release. You have been quick off the mark. ;-) This latest Arduino differs from previous ones in that it includes a microprocessor on the one board. Its OS is a form of Debian Linux, so the standard Linux version of jOrgan will perhaps run on it with no problems. However, it seems to have a RAM of only 785 Kbites (from memory) and I do think that this would be somewhat limiting for dispositions which use a soundfont that is larger than about 200MB. You would find that it would allow only a small number of stops to be active at the one time. I base this on my experience using an RPi 3B, which has a RAM of 1 GB. There are certainly some jOrgan dispositions which I cannot run on my home console because of RAM limitations. Best wishes, JohnR _________________________________________________________________________________________________________* END OF THREAD Reverb for jOrgan Reverb for jOrgan From John Beach on 2026-01-04 at 01:52 I know very little about programming, so my question reflects that ignorance. Would it be difficult to add a "host" capability for reverb plug-ins, such as Convology XT, to jOrgan? What is the impact of external host programs (vst) running with jOrgan on latency, for example? I have never used one while playing jOrgan. A while ago, when Sven turned over future changes to jOrgan to Jonathan A., there was minor discussion of doing a reconstruction of it as a modular program which would have allowed particular sections to be modified without impacting others. The customizer was one such section where modular construction would be an improvement, as each user's physical console setup is different. Is there any, or enough interest to pursue that and is it a possibility? As I listen to other organists, using virtual pipe organ programs (you know which l mean), who post regularly on youtube, I am immediately aware of the difference in the quality of output sound in comparison to jOrgan, particularly, the reverb. Thanks, in advance, for comments and input. Happy New Year to all! John Beach ______________________________________________________________________________________________________________ From JohnR on Sat Mar 07 at 10:17 On Sun Jan 04 2026 John Beach wrote: 'Would it be difficult to add a "host" capability for reverb plug-ins, such as Convology XT, to jOrgan? Hi JohnB, Thank you for raising this matter. My apologies for leaving it so long before replying. I do not own a computer with a RAM large enough to do justice to most HW or GO sample-sets. Since Fluidsynth version 2 was released, jOrgan users have been able to enjoy its greatly- improved reverb sound. However, there is the drawback of the cyclic modulation of the reverb. This "flaw" was apparently unavoidable if the greater quality of the reverb was to be achieved, but for some users, it disqualifies the Fluidsynth 2 reverb from being a satisfactory solution. You ask if a VST reverb plug-in can be used by jOrgan in the quest for a good reverb. It certainly can be. It formed part of Graham Goode's efforts with jOrgan on Puppy Linux many years ago, and its use with Windows PC's is described in a long-standing Graham Goode tutorial in the jOrgan InfoBase: "Using Jack Connections with VST Effects". I believe it would be a great boon for jOrgan users if one of us were to use that tutorial as the basis for a comprehensive report/tutorial on how to achieve a reverb that at least equals and even improves on what is possible from HW or GO. Graham's tutorial is now so old that a revision may be in order. Any offers? Best wishes, JohnR ______________________________________________________________________________________________________________ T POST From John Beach on Sat 07-03-2026 at 14:40 PM JohnR., thank you for your response concerning the fluidsynth reverb. I wish I had the technical capability with programming to be able to do it myself, but I do not. I hope that others, who do, will. John Beach ______________________________________________________________________________________________________________ From Sven Meier on Sun 14-03-2026 at 12:26 AM Hi everyone, since FluidSynth’s reverb has been a recurring topic on the jOrgan list, I thought it might be useful to share that the FluidSynth developers have recently been working on improvements in this area. Please see the quoted message from the FluidSynth mailing list below for details. It might be a good opportunity for us to offer some feedback or suggestions regarding this new feature. https://www.fluidsynth.org/api/ReverbOverview.html Regards Sven --------------------------------------- Dear all, it took almost four years, but now there is an implementation ready for supporting different reverb engines. Pls. check out this GitHub discussion and feel free to share your thoughts: https://github.com/FluidSynth/fluidsynth/discussions/1771 Tom fluid-dev mailing list [email protected] https://lists.nongnu.org/mailman/listinfo/fluid-dev ______________________________________________________________________________________________________________ From John Beach on Sun 14-03-2026 at 1:04 PM Sven, et al., It seems to me that, in order to be fair in our judgments, we would have to hear a sustained organ chord in order to see whether any of the reverb engines that are proposed to replace what is currently used in fluidsynth, eliminates the problem of "looped frequency modulation" that is annoyingly persistent above an, about .015 setting of the Level element in fluidsynth. With the piano, snap and Water drop samples used, it is difficult to say that any particular one is really superior to any of the others. It seems to me that, if the fluidsynth developers know what the particular issue is that we,virtual organ users, have with the reverb level element, they should be, specifically, seeking to eliminate that problem, or, if that is desired in the case of other types of users, at least, offer a version in which the reverb element eliminates the problem of "looped frequency modulation." Thank you for sending the link to this test. Hopefully, something beneficial will result. Until then, Creative Labs EAX is about the best substitute for reverb for those of us who, still, use Creative Labs soundcards that were designed for the soundfont specification and which allow for tinkering with the reverb settings in a reverb studio of the audio console. John Beach ______________________________________________________________________________________________________________ From JohnR on Mon 16-03-2026 at 4:11 PM Hi Sven, Thank you for bringing this promising development in Fluidsynth to our attention. As JohnB said, the samples provided do not exactly suit what we are listening for, but at least two of the reverb engines (apart from the cyclic modulation one) do offer good sound, as far as I could tell. So this could be the answer for jOrgan folk like JohnB, who have a real problem with that modulation in the current FS version 2. I have downloaded the pre-compiled binary for Windows 10, in the hope that I may be able to test it out, using Jack for Windows. At the moment I will need help from a friend to sort it all out, and have approached him about it. Best wishes, JohnR ______________________________________________________________________________________________________________ From Sven Meier on Thur 19-03-2026 at 7:42 AM As a follow-up, I’ve tried the new reverb implementation and am happy to report that the updated lexverb engine (“lex”) noticeably improves the audio quality to my ears. The sound is much clearer, free of distortion, and overall responsiveness of pipes has improved as well. I still need to test the Dattorro (“dat”) engine, but so far I’m very pleased with what the FluidSynth developers have delivered. Sven ______________________________________________________________________________________________________________ POST From JohnR on Thursday 2026-05-28 at 3:47 PM Hi all, It is two months since I mentioned my plan to listen to the four reverb engines currently being investigated by Fluidsynth developers, but to do so using jOrgan, rather than going just by the sample sounds that were available. I did listen to those samples at the time, and my conclusion agreed with Sven's, in his favourable report of the sound of "lex". The "dat" also gave a good sound, but from what I read about it, it would be more CPU-intensive. Regrettably, the help I was seeking from a friend did not eventuate, and so nothing came of that. However, with the help of Google AI, I can report that Fluidsynth has decided to incorporate all four in the next major release of Fluidsynth, leaving it to the user to specify which was to be used. No time line has been mentioned, so we shall have to wait until it turns up. It will be Version 2.6.0. Best wishes, JohnR _________________________________________________________________________________________________________* END OF THREAD SF2 Player plugin SF2 Player plugin SF2 Player plugin From Dr. Mark Bugeja MD on 2026-01-06 at 11.04 Hi guys, I came across this: https://www.kvraudio.com/news/bertom-audio-releases-free-sf2-player-plugin-65817 Would this be of any use/ help for jOrgan? Mark ______________________________________________________________________________________________________________ From Lynn Walls on 2026-01-06 at 13:06 Probably not of any use for jOrgan. According to its web page this player only plays ONE preset (i.e., one rank, one voice) at a time from only ONE selected MIDI channel. jOrgan requires something like fluidsynth, which accepts MIDI input from MANY channels playing MANY presets/voices/ranks simultaneously. CLW ______________________________________________________________________________________________________________ From Dr. Mark Bugeja MD on 2026-01-06 at 17.14 Thanks Lynn. I just asked to make sure we don't miss on anything that could potentially enhance the experiences with jOrgan, improving on what we've got so far. Mark _________________________________________________________________________________________________________* END OF THREAD Save As Save As From JohnR on 2026-02-05 at 16:58 Hi all, I noticed that the late Bruce Miles' "organ notes" advises the use of "Save As". I have created a new page for the jOrgan InfoBase to deal with this. Here is the draft of the text, which can easily be modified. I would welcome any suggestions. 'jOrgan no longer offers the option, "Save As", when dealing with Disposition files. When a user wants to make changes to an existing disposition and to save it with those changes, it may be preferable firstly to create a renamed disposition before making any changes, perhaps using the existing name but adding a version number (e.g. "name_version_2.disposition" or "name_v.2.disposition". A simple work-around is to copy the existing disposition "name.disposition", paste it to the Desktop, rename it there, and then return it to the folder from which it was copied. This will leave the original disposition file untouched. The user can confidently open the renamed version in jOrgan, (by double-clicking on it), and then make any changes, before saving it.' Best wishes, JohnR ______________________________________________________________________________________________________________ From Lynn Walls on 2026-02-05 at 00:50 for better clarity, change: "...it may be preferable firstly to create a renamed disposition before making any changes..." to: "...it may be preferable firstly to create a renamed copy of the disposition before making any changes..." CLW ______________________________________________________________________________________________________________ From John Beach on 2026-02-06 at 02:10 I solved the problem of disposition confusion, long ago, by making copies, from the first disposition after it has had all elements customised and soundfonts loaded in fluidsynth, in the individual disposition folder and renaming them according to purpose . For example, English Cathedral- Keyboards (for dispositions played manually) and English Cathedral- MIDI for a disposition used to play MIDI files, exclusively,(with a MIDI sequencer program using virtual midi cable outputting to jOrgan). This is the quickest and least confusing and least work. John Beach ______________________________________________________________________________________________________________ From JohnR on 2026-02-21-at 15:40 PM Many thanks to Lynn and JohnB for your replies. I have inserted Lynn's suggested rewording into the "Save As" jOrgan InfoBase page, and added a concluding sentence. I have also placed an Annotation at para. 9 of Bruce Miles' jOrgan Notes, which can be accessed at https://jorgan.info/base/bm/jorgan_notes.pdf . Best wishes, JohnR _________________________________________________________________________________________________________* END OF THREAD full screen not full screen full screen not full screen From Paul Stringfellow on 2026-02-17 at 23:46 Hi, I’m trying to use jOrgan on a Pi to make a simple MIDI controller. I’ve got most of it figured out now, but when I set jOrgan to full screen, it leaves a bar top and bottom that you can see the desktop behind. Has anyone come across this and know how to fix? Thanks, Paul ______________________________________________________________________________________________________________ From John Beach on 2026-02-18 at 01:18 Usually, when the condition you describe happens, the change to "Full view" is made by clicking on the display symbol in the upper righthand corner of the "x"that closes the window. I believe that, if the display is not "full," there will be two small squares, one slightly offset from the other, and when you click on them, the change to full-view display will occur. John Beach ______________________________________________________________________________________________________________ From Paul Stringfellow on 2026-02-18 at 01:27 Thanks John. It’s not that. I’m accessing full screen from the menu. I tried to send a photo, but the list doesn’t allow it. Paul ______________________________________________________________________________________________________________ From JohnR on 2026-02-18 at 06:55 Hi Paul, Welcome to the Mailing List. Have you tried to use the Customizer? Full Screen is enabled by going to the Customizer, page 2, and clicking where appropriate. Try it. Best wishes, JohnR ______________________________________________________________________________________________________________ From Paul Stringfellow on 2026-02-18 at 06:59 Hi John, Yes, that’s what I’ve done… Paul ______________________________________________________________________________________________________________ From JohnR on 2026-02-18 at 09:58 Hi Paul, Your reply is too short to be truly informative. But I assume it means that you have tried that (clicking on the "0.0" each time on page 2 of the Customizer and then on "FINISH"), but it made no difference. Which disposition are you using? Best wishes, JohnR ______________________________________________________________________________________________________________ From Paul Stringfellow on Wed, Feb 18, 2026 11:03 PM Hi John, I wasn’t confused as to how to enable full screen mode, only that it wasn’t working. I’ve just worked out how to get the OS to hide the taskbar (nothing ever simple with linux!), and that seems to have worked, the full screen now fills the screen nicely. Thanks for your help. Paul ______________________________________________________________________________________________________________ From Chris Pearson on Thu, Feb 19, 2026 5:22 AM Hi, Pressing F11 works on my set up. Chris Pearson ______________________________________________________________________________________________________________ From JohnR on Sat, Feb 21, at 14:51 PM Hi Paul, Congratulations at solving your problem! I've been using RPi's for years now, for jOrgan and as a desktop, and I've never noticed that particular problem. I put it to Google AI, and it returned two solutions, the choice depending on how recent the RPi OS is. Best wishes, JohnR _________________________________________________________________________________________________________* END OF THREAD SAM on my Wyvern 4 manual jOrgan SAM on my Wyvern 4 manual jOrgan From Stephen Rowley on Thu, Apr 2, 2026 5:49 PM Hello everyone. I know it has been a long time since I was last here but my old jOrgan console was all working fine. Now I have recently bought an old 4 Manual Wyvern organ console that has had all the Wyvern electronics removed. I have fitted all the MIDI boards from my old console to this one and it all works fine but, the Wyvern console has SAM on all the stops tabs which I would like to use. With the help of a computer wizard friend we have come up with a program for an Arduino that will take the MIDI note ON and Off to drive the SAM's but I can not find how to OUTPUT the thumb piston MIDI information to my MIDISPORT 8X8. I also have a Glockenspiel from an old Rogers organ I would like to add. I am sure there must be some information somewhere but I can not find it so any help would be gratefully appreciated. ______________________________________________________________________________________________________________ From JohnR on Mon 2026-04-06 at 7:01 AM Hi Stephen, I warn you that using an old console, which I assume housed what was an analogue electronics organ, can cause problems, as contacts which worked OK when fed with a particular voltage, become unreliable when the voltage being fed for digital circuits is much lower. However, you have made no mention about such things, so perhaps you will get away with it. I hope so. The normal expectation for a piston switch is that there is a single wire to the encoder - in this case the Arduino, with the other wire going to the common buss shared probably by the other console switches. This should produce a Note On MIDI message when pressed, and a Note Off when released. You use the jOrgan Customizer so that the disposition you are using learns to recognise those MIDI messages. With everything running and the console powered up, go to page 3 of the Customizer (File > Customizer) , and up at the top use the scroll arrow and select the MIDI device . This may appear at the bottom of the list. Select the Activate cell of the piston in question (and click on Cancel if you clicked the wrong cell!). Press the piston switch at the console, but keep it pressed. Now click on the Deactivate cell, and then release the piston switch. This can be tricky, and you may need the help of another person. Move on to the other pistons. Click on "Finish". Save the disposition before closing it. I hope I'm not missing something obvious. Please report back. JohnR __________________________________________________________________________________________________________ From RickW on Mon 2026-04-06 at 7:51 AM Hello Stephen I am not sure exactly what question you are asking? Is it a hardware MIDI question, or a jOrgan question? You say you have the SAMS on the console being driven by Arduino, but then talk about the Pistons? What jOrgan disposition are you using? Do you have jOrgan on screen Pistons triggering on screen combinations of of Stops? In the "cleanest" scenario, you will have Pistons on your console sending MIDI "PC" (Program Change) messages OUT from your Console. In jOrgan you will need "Activate" messages on the jOrgan screen Pistons and a jOrgan "Connector" Element to link the screen Pistons with your MIDI Interface. Next you will need "Activate" "Deactivate" messages on your jOrgan screen Stops, connected through the "Connector" Element in jOrgan out to your MIDI Interface. If your Arduino MIDI Hardware is connected through different MIDI "Ports" of your MidiSport8x8, then you will need multiple "Connector" Elements in jOrgan, one for each "Port" of your MidiSport8x8. The specific "Stops" controlled through each Arduino board will need to be linked to the correct "Connector" Element that is addressed to the correct "Port" of your MidiSport8x8. Once you create a second or more "Connector" Element, you should start to see how this is going to work. Let us know if you need more detail. If you do, I will need a more detailed description of your Arduino MIDI Hardware and how it is connected. Regards Rick ______________________________________________________________________________________________________________ From Stephen Rowley on Mon 2026-04-06 at 8:41 AM Hello Rick. Sorry I did not explain properly. I have the SAM's connected and working in jOrgan, what I need is for jOrgan to energize the solenoids on the SAM's when I push a thumb piston. The Arduino will be connected to one of the output on the MIDI sport. The ardiuno has 127 shift register outputs which are connected to the JK flip flops the output of which drive the Darlington transistors which energize the solenoids on each SAM. When I push a thumb piston on the console the corresponding stop tabs on jOrgan move. I am using the 3/15 Barton set at this time. Hope this helps. Regards Steve. ______________________________________________________________________________________________________________ From JohnR on Frid 2026-04-10 at 12:48 PM Stephen Rowley wrote:"what I need is for jOrgan to energize the solenoids on the SAM's when I push a thumb piston. ... When I push a thumb piston on the console the corresponding stop tabs on jOrgan move." Hi Phil, This indicates that you did not need the instructions I gave in my reply, which were made due to my misreading of your original post. I remind you of the Mailing List thread which you initiated late last year. In that thread I drew attention to the sams-example.disposition, which I felt needed some revision, and so I made available sams-example2.disposition. That thread also contained more detailed information of your set-up, and this may be of interest to Rick. In that thread I said that the example disposition did not make clear what the control squares represented, but I did say that my conclusion was that they were STOP switches (and so not PISTON switches). I am still of that opinion. Piston switches would not make a lot of sense, as there will be times when you want to turn on (or off) a stop, without any use of a piston to do it. I believe that you have not given us any idea as to why your attempts to get the SAMs to work did not succeed. That revised sams example should have been enough to point you in the right direction. I think that was Sven's hope in setting the example dispositions up. Perhaps you failed to notice that the Stops now had new messages (View > Messages), messages which stops not associated with SAMs do not have. Or perhaps I didn't get the messages quite correct. Best wishes, JohnR The thread can be viewed here: https://jorgan.info/base/arch/l/link2025b.html#B27 The sams-example2.disposition (slightly corrected) can be downloaded here: https://jorgan.info/base/arch/jr/sams-example2.disposition ______________________________________________________________________________________________________________ From JohnR on Frid 2026-04-10 at 2:57 PM Hi Stephen and all, My humble apologies for the start of my post sent almost two hours ago, when I attributed Phil's post sent to the Mailing List late last year, to Stephen Rowley. Same subject, but different people. I think much of what the rest of my post said is quite relevant to Stephen's situation. I hope the reason for no success with the SAMs turns out to be something quite simple. Best wishes, JohnR ______________________________________________________________________________________________________________ From Stephen Rowley on Frid 2026-04-10 at 4:43 PM Hi John. This is what I needed. Now can I do the same with the keyboards as I have inherited a 3 manual 22 rack Wurltizer organ I would like to use jOrgan as the relay system. Regards Steve. ______________________________________________________________________________________________________________ From JohnR on Frid 2026-04-10 at 8:07 PM On Fri Apr 10 2026 Stephen Rowley wrote:"This is what I needed. Now can I do the same with the keyboards as I have inherited a 3 manual 22 rack Wurltizer organ" Hi Steve, I'm assuming that this is a pipe organ. 22 ranks? You can certainly use jOrgan to drive pipes. I have had no experience of doing this, but there are others who do it successfully. My hope is that one of them will read this and offer to send you a working disposition as an email attachment. You won't be able to use it directly, but it will give you information which will save you and perhaps others (trying to help you) a lot of work. If you continue to have problems with either project, there is always the possibility of sending them your disposition as an attachment, in the hope that they can quickly spot the error. Best wishes, JohnR ______________________________________________________________________________________________________________ From Stephen Rowley on Frid 2026-04-10 at 8:25 PM Hi John. Yes it is a pipe organ and one that is quite historic in the UK for being the last Wurlitzer to be install by Wurlitzer in the UK. We are going to install it in to a friend's house. As always you are a mine of information, Thank you and everyone for the time. I will keep you updated with my progress. Steve. ______________________________________________________________________________________________________________ From phil on Sat 2026-04-11 at 12:22 AM John, I understand the post was in reply to Stephen. I wanted to mention that I did get the SAMs running just fine from jorgan via the examples. I do know at that time I didn’t like something - IIRC some condition I needed where the linked tabs in jorgan and the real tabs conflicted. So what I ended doing was reprogramming the main microcontroller to send tabs information to the SAMS driver cards, and store the various tabs & memory levels to an SD card on the fly. Since I’m using the Teensy 4.1 module as the console controller and also as the controller on the SAMs driver cards, there’s plenty of speed. It’s also all transmitted over ethernet. So far jorgan (also on ethernet) is working fine as the “relay” for the organ. Regards, Phil ______________________________________________________________________________________________________________ From phil on Sat 2026-04-11 at 12:26 AM Hi Stephen, I currently am using jorgan as a relay for 3/11 Wurlitzer and it is working just fine. It takes a bit to get used to ‘configuring’ jorgan but it will do the job. Phil ______________________________________________________________________________________________________________ From Stephen Rowley on Sat 2026-04-11 at 1:17 AM Hi Phil. Would you have a sample deposition file so I can see how you used jOrgan as the relay. Steve. ______________________________________________________________________________________________________________ From Lynn Walls on Sat 2026-04-11 at 6:02 AM Once you have the .disposition file, here is a jOrgan disposition browser that you may find useful for browsing through the disposition: http://jorgan.freeddns.org:8192/cgi-bin/jorgan_xml.pl CLW ______________________________________________________________________________________________________________ From Stephen Rowley on Sat 2026-04-11 at 8:10 AM Thank you Lynn. _________________________________________________________________________________________________________* END OF THREAD Traps control Traps control From phil on Tues 2026-04-14 at 6:44 AM Looking for help from the list. I have a rank for traps - tambourine, castanets, snare drum, etc. These are outputs on a midi driver card. The basics on accompaniment manual: 1 I have a stop (from a stop tab) defined for each trap. 2 I have a rank defined as traps on channel 13 referred to by each trap stop tab. 3 The rank driver card has 1 trap per output pin (one trap per midi note). 4 I have the accompaniment keyboard defined as channel 3, 1-61 (0 to 60). The problem I’m having is creating an output on a single pin (pin number defined by what stop tab is on) when any key on the keyboard is pressed and then connecting that to the stop/rank/output. Looking for any clues here – I have not been able to find decent documentation for this setup. Sample dispositions all drive sound engines via command codes with no visible note handling. Regards, Phil Iehle _________________________________________________________________________________________________________* END OF THREAD Looped Modulation in Looped Modulation in Fluidsynth reverb From John Beach on Sun 2026-04-19 at 2:03 AM I know that this subject is old and, annoyingly, boring. However, it caught me by surprise this week as I was practicing service music. I hit the "d" note in the octave above middle-c, Midi-note number 74, with only two stops drawn, a boudon 8' and a diapason 8'. The result was a very pronounced, modulating, looped sound that was very unnatural. I shut off one stop to determine which was the offending wave file and believing it to be the diapason, I opened the soundfont in Polyphone to make the correction to the looping, only to discover that there was no problem with the sound in Polyphone. It played a sustained tone of the "d" note with no offensive modulation. I went back to the jOrgan disposition and, since I have Creative Labs Audigy 5.0 RX and the reverb capabilities of EAX, I set the fluidsynth "level" to "O', the "off" position and played the note. No modulation. Raising the level just one green light, the offensive modulation was very noticeable. I believe this problem is a combination of commixing frequencies, the actual frequency of the pitch of the note and the specific, software-effectuated technique of increasing "level," independent of "room," 'width" or "damping." That is, the problem is in the instructions of "level," and the modulation problem commences immediately above the "off" setting, with room and width and damping completely off. My question is why would 'level" not be a detection of ambience-size rather than a generation of effect on the sound source?? The idea of some relative equivalence is quite fake. Seemingly, the concept of reverberation, with respect to sound processing, should be an independent, self-contained structure, the volume of a building, in which sound is emitted, not modulated, unless by the source of tremulant or vibrato. I have never detected that "baffling" from walls, corners, pillars or columns in large churches or cathedrals, ever caused a frequential modulation of the sound. I am wondering whether amplitude and frequency of impulse response wave file, with respect to the "level" are modulated because there is no means of expressing a "static" environment, electronically. That is, they are trying to replicate an extantial reality, electronically, and it is, actually, impossible to do so. It is what it is, and I am trying to understand the problem. But I know that sound in a space can be recorded and played back. However, it is problematic to attempt to infuse the sound source (wave file) with the parameters of the size of a space, artificially, to effectuate the reality of the size of the space. In MIDI, the fundamental actuator IS the wave file, without which, all the parameters are meaningless. Apparently, this is true for concepts of reverberation. The ambience is what it is, extantially, until a sound wave is emitted. The sound wave reveals the relative size to the listener. My thought is that the only realistic organ sounds achievable with jOrgan dispositions would be those soundfonts that are made with samples (wave files) that have been recorded with the reverberation ambience. They won't need the fluidsynth reverb at all. John Beach ______________________________________________________________________________________________________________ From Via2 on Sun 2026-04-19 at 6:38 AM Fascinating. I suggest ask AI in various ways of questioning. Various questions posed to target the answer that properly covers all the minusua involved. Still, AI will lie or make up stuff if in case it makes AI seem stupid, it’s been shown. BEST REGARDS _________________________________________________________________________________________________________* END OF THREAD AI (off-topic) AI (off-topic) From Lynn Walls on Sun 2026-04-19 at 9:20 AM It's somewhat true to say that AI "lies", but I don't really consider it lying in the true sense. To me a "lie" is an overt attempt to deceive. The free on-the-internet AI robots that I have communicated with do not have any sort of motivation to deceive. In fact, they don't even know the difference between the truth and a lie. All they do is report back to you whatever they find on in the internet and stop searching as soon as what they have found seems to be a plausible response to whatever you have asked or stated. They make no effort to determine the truth of what they report back. It seems they just report whatever they find as soon as it seems to answer your question. They really don't research deep enough to determine the actual "truth" of what they find. CLW ----------------------------------------- On 4/18/2026 4:38 PM, Via2 wrote in the thread "Looped Modulation in Fluidsynth reverb": "Fascinating. I suggest ask AI in various ways of questioning. Various questions posed to target the answer that properly covers all the minusua involved. Still, AI will lie or make up stuff if in case it makes AI seem stupid, it’s been shown." ______________________________________________________________________________________________________________ From via2on on Sun 2026-04-19 at 11:30 AM Right it’s just a search mechanism. Some info on the net may have errors that have been referred to as lies by some. Which is just semantics, I suppose. One has to pose questions in various ways to get to the detail to match what they seek. That is the key. _________________________________________________________________________________________________________* END OF THREAD Piston indicator lights and launchpad support Piston indicator lights and launchpad support From Lynn Walls on Sun 2026-04-19 at 3:30 AM Here is a video showing how a launchpad may be used with jOrgan. It also shows how the jOrgan console display can show indicators for which piston was last activated. https://www.youtube.com/watch?v=ydGMzSRlWqY Below is a link to the disposition shown in the video in case you want to examine it to see how to configure these features. https://www.dropbox.com/scl/fi/2udwkr5jcgqq7s6uj3k98/SC-88VL.disposition?rlkey= zxr1xbfh254n8clavojct995n&st=ktngtzv5&dl=0 CLW ______________________________________________________________________________________________________________ From JohnR on Sat 2026-04-25 at 10:34 AM Hi Lynn, Thank you for sharing your disposition. Would you consider writing a short guide to help interested persons find their way through the sections of the complex disposition which are relevant to the use of the Launchpad? What is the physical size of the Launchpad Mini? I made a brief search but did not find any indication. Best wishes, JohnR On Sun Apr 19 2026 Lynn Walls wrote:"Here is a video showing how a launchpad may be used with jOrgan..." ______________________________________________________________________________________________________________ From Lynn Walls on Sun 2026-04-26 at 10:00 AM JohnR, The physical size of the LaunchPad Mini is 7.25" X 7.25" X 0.5" It has been several years since I configured that disposition to support the LPMini. But I will take a look at it again and try to extract the essentials of adding the LPMini to an existing .disposition. I'll send the writeup to you personally for review before posting it to the list just in case what I write needs more explanation. Lynn ______________________________________________________________________________________________________________ From JohnR on Sun 2026-04-26 at 3:22 PM Hi Lynn, Thank you. That will be very helpful. I plan to create a jOrgan InfoBase page to make it easily accessible in the long term. Would you say after some years of use, that the Launchpad has been a very worthwhile addition to playing that disposition? A search indicated to me that the Launchpad is stocked by music stores here in Sydney, at a price much the same as in the U.S., allowing for currency conversion. Best wishes, JohnR ______________________________________________________________________________________________________________ From Lynn Walls on Wed 2026-04-29 at 2:55 AM On 4/26/2026 1:22 AM, JohnR wrote (in a private email):"Would you say after some years of use, that the Launchpad has been a very worthwhile addition to playing that disposition?" Yes. That particular disposition has stops and pistons that do not correspond well to the physical stops and pistons on my organ console. So, in order to activate/deactivate them I would have to reach a fair distance to the computer touch screen -- an action that significantly disrupts keyboard playing. The Launchpad Mini, being placed much closer to the keyboard area provides a tactile way of changing the jOrgan disposition stops as easily as touching the physical piston buttons on a real console -- and especially useful if the user does not have a computer touch-sensitive display, and must use a mouse. The question comes from others about the possibility of using the Launchpad buttons to "toggle" jOrgan stops on and off directly (without having to use pistons). I have only configured jOrgan to interface Launchpad buttons to jOrgan pistons (combos). I would be interested to hear from anyone who has configured a Launchpad to "toggle" jOrgan stop elements directly/individually. CLW _________________________________________________________________________________________________________* END OF THREAD Configuring swell Configuring swell From Andrew Larkham on Tues 2026-04-28 at 5:30 AM I need some help from the experts out there…. I have just upgraded my computer to a mini. I saved all my dispositions etc, and ported them across. I can get everything to work – bar the swell pedal. Starting to tear my hair out ( and I do still have some!) Monitor shows …. Channel 7 Then 176 and 7 and a value that depends on the pedal position. The other ‘swell’ shows 176 and 11 and the value. I might be getting in a muddle over Engage and Change. But I cant even get the onscreen swell to move. I know this all worked with my set up once! Help! Many thanks, Andy ______________________________________________________________________________________________________________ From RickW on Tues 2026-04-28 at 8:27 AM Hi Andrew In the second page of the Customizer there is a "Tab" for Analog controls. You need to set your MIDI device that those messages come from on that Tab. If you need me to be more specific with naming, I will need to open my system later. You can also go to the "Connector" "Element" in "Construct" Mode and set the MIDI Device in Properties. Let us know how you go. Regards Rick ______________________________________________________________________________________________________________ From Andrew Larkham on Tues 2026-04-28 at 4:30 PM Thanks Rick. I've already done that. Both manuals and pedalboard work well and I can see the midi messages from the swell. They just don't do anything! ______________________________________________________________________________________________________________ From via2 on Tues 2026-04-28 at 10:52 AM After running into these problems, I’ve found that it’s more foolproof to clone the drive with the organ and sample sets that works perfectly the way you like. Makes that clone a bootable external drive or make a partition on the minis internal drive to install the perfectly working organ on the clone drive onto it. There are step by step instructions online if needed that are free or cheap software to complete the tasks depending on the OS, to easily do this. Now with a bootable drive of working disposition you can plug it in or boot into it from the partition in order to use it as usual. Otherwise You may have to search for scattered files pertaining to swell config. Using the internal drive is for faster multi processors for high polyphony wet works and RAM heavy sample sets as in Hauptwerk or Grand Orgue. If thé mini has less of RAM than on original system, something might get scrapped. But with these jOrgan sample sets, it’s less demand on RAM because less high end recording of each and everything with and without reverb, wet and high polyphony demands processing. If you have upgraded RAM and Multiprocessor it should be good. And some files were dispatched somewhere on the drive that wasn’t copied over. Maybe storage was full and things got swapped out or orphaned. So it it didn’t transfer with the rest of it. Always make bootable clones of your best favs on bootable drives. Turn off auto update for the clone drive to keep it in working order. Label the cloned drive your organ’s setup name. Write down login passwords on paper to correspond to the drives in case years go by and passwords get changed. Tape that onto the case. ______________________________________________________________________________________________________________ From Andrew Larkham on Tues 2026-04-28 at 4:34 PM Thank you. A good idea. I upgraded basically because my old computer was rather old and still on Win7 with limited RAM. I still have the old drive so I may need to investigate that. Andrew. ______________________________________________________________________________________________________________ From RickW on Tues 2026-04-28 at 5:20 PM Hi again Andrew Is this happening on all of your dispositions? What dispositions are you using? Swell in jOrgan (assuming a Fluidsynth link for audio) uses a "Continuous Filter" Element. The "Continuous Filter" Element must have "Messages" "Change" and "Engaging" with relevant MIDI addressing (eg. qual 191, equal 11, lessEqual 126 | div 126 | get value) The "Continuous Filter" Element must be "Referenced From" a "Connector" Element. The "Connector" Element must have its "Input" set to the correct MIDI Device (this can be done in "Construct Mode" in the "Properties" of the "Connector Element", or can be done on the 3rd page of "Customize Disposition" then click on the "Continuous" TAB.) If your Swell MIDI messages come from the same MIDI device as your keyboards or pedalboard, you may be able to see the messages even if the proper MIDI Device connection has not been made in the "Customizer" "Continuous" TAB (or on the "Connector" Element in "Construct Mode). ______________________________________________________________________________________________________________ From Andrew Larkham on Tues 2026-04-28 at 7:05 PM Hi Rick Thanks for your help. I did not have the Connector input set correctly. As I have 2 midi inputs – 1 for manuals and 1 for pedal / swell I use Midi Merge. But .. hasn’t changed anything. Can still see all the midi messages in monitor, but the swell pedal refuses to move … and the sound doesn’t change. At the moment I am concentrating on 1 disposition – Paul Stratmans Aeolian Skinner. Will try another later … but Im sure I had this working OK before my upgrade. I need to try and dig out the old HDD with my previous dispositions on – and take a look at the MPL. I have possibly made so any changes …. 😊 I knew upgrading would bring the odd headache or two – but this is pure frustration now! Andrew. ______________________________________________________________________________________________________________ From Andrew Larkham on Wed 2026-04-29 at 4:29 AM Hi Rick, After more searching, I realised that the Continuous Filter was not set correctly in the customiser– as you so rightly suggested below. So I now have a working Swell! So heart found thanks for that. But that’s not the end of the saga! Now I get a ‘noise’ when I operate the swell. The manuals are set to look at the correct channel (1,2 ) on one Midi Controller, and the pedals and Swell on different channels on the other. Set as per in the customizer. I’ve never had this problem before, so maybe a bit more tweaking to do. As per Sven's email, I made the SSD from the old machine bootable and booted it from an old laptop, so I have a reference now. And to cap it all one of the Hall switches on the pedals went rogue. So a bit more to try and work through. But we are making progress! Andrew _________________________________________________________________________________________________________* END OF THREAD Linux driver for old Dell touch screen Linux driver for old Dell touch screen From Ross Willson on Mon 2026-05-04 at 1:40 PM Hi all, I have an old Dell touch screen, model 1905FP, which works in Windows XP and Windows 11 but I can't get it to work with Linux. I'm using Linux Mint 23.3 and I've tried installing the Windows driver using Wine, but no luck. Any ideas? Regards, Ross ______________________________________________________________________________________________________________ From JohnR on Mon 2026-05-04 at 2:38 PM Hi Ross, Have you tried Google AI? I used your wording, starting with "How can I get ...", and got quite a lot of detail. It could be worth trying. Best wishes, JohnR ______________________________________________________________________________________________________________ From Graham Goode on Mon 2026-05-04 at 5:07 PM Hi Ross, The drivers for those older model should already be part of the linux kernel, so I would guess that your issue is not driver related but configuration related. I found these steps from a 2021 Linux Mint forum post: ------------ In the terminal type "xinput". This will list all of your input devices, which are hopefully named in a way that makes it easy to identify which is which. In my case the touch screen input device was "Wacom HID 5113 Finger touch". Look at the id of that device. In my case it was id=11. With that information, to disable the touch screen (or any input device) type: xinput disable 11 To enable it again type: xinput enable 11 (replace "11" with whatever id you found for your touch screen, which might be different for you) ----------- Let us know if that helps. GrahamG ______________________________________________________________________________________________________________ From Lynn Walls on Mon 2026-05-04 at 11:54 PM Just two weeks ago I upgraded all my linux computers from Linux Mint LMDE 6 (the Debian version) using kernel version 6.1.0 to LMDE 7 using kernel version 6.12.85. Afterward all the computers worked fine EXCEPT for one old HP Mini Notebook with an integrated graphics adapter. That computer would boot up, but would not display the GUI desktop on the display screen. After much internet searching using AI and trying all kinds of workarounds to no avail, I eventually discovered that the upgraded LMDE 7 kernel had dropped support for some older graphics adapters and no longer supported the old HP graphics adapter. Eventually, rolling the old HP computer back to LMDE 6 with kernel 6.1.0 fixed the problem. So, in the case of Linux Mint, it is not a given that all linux implementations will forever support all older computers. It really depends on what the support organization (Mint) decides to keep or drop from the kernel when they build an upgrade. CLW ______________________________________________________________________________________________________________ From Ross Willson on Mon 2026-05-11 at 9:58 PM Hi John, Graham and Lynn, I tried ChatGPT and after a very long time and much trial and error (mostly error) I almost had the screen calibrated using the Libinput Calibration Matrix. I had the y axis correct and the x axis correct at the top but not the bottom. When I booted this morning ready to finish the job I could only get to a terminal. I managed to solve that one then suddenly touching the screen didn't move the cursor at all. I then booted to Windows 11 and the touch didn't work with that either. It looks like a problem with the usb connection to the screen. Tomorrow I might try my 14" Asus touch screen which I have used before in both Windows and Linux, then put it it all to bed for a week before I pull all my hair out. I can still play and use pistons to change stops. My VPO consists of the timberwork and pedalboard of a very old organ our church wanted to dispose of. (I think it was a Schober kit organ.) I bought two midi keyboards instead of trying to midify the manuals. I used two 8051 microcontrollers (very old but I could easily program them in C) to read the pedals (in an 8x4 matrix), read one swell pedal (I had to add an adc), read the telephone keypad and output the piston number to a 7-segment display, and output midi signals. For a long time I just used Windows XP which was fine with jOrgan and Miditzer. Last year I upgraded to Windows 11 and have been enjoying Grand Orgue. Perhaps I should have stuck with Windows 11 but I wanted to use Linux. Thanks again and I'll let you know when it's all working. Ross ______________________________________________________________________________________________________________ From Ross Willson on Tues 2026-05-12 at 3:04 PM Hi all, When I booted up this morning everything worked and I was able to finish the calibration. Thanks again, Ross _________________________________________________________________________________________________________* END OF THREAD jOrgan Files on jOrgan Files on Sourceforge From Lynn Walls on Mon 2026-05-11 at 8:19 AM Did you all notice that the jOrgan Files page on Sourceforge says "Status: Abandoned". CLW ______________________________________________________________________________________________________________ From JohnR on Tues 2026-05-12 at 12:45 PM Hi Lynn, I noticed it quite some time ago,and I think I mentioned it in a post. I don't think it was the main subject of that post. The word is unfortunate, and presumably it was placed there by Sven or by SourceForge. It prompted me into saving most of the contents of the page into the InfoBase Archive, so that I can make that available on the internet, should the page disappear. If the page does disappear, I shall be disappointed, as I examine it each Sunday morning to keep track of how many downloads of jOrgan are made each week. I have been doing this for a few years. Regarding the statistics, I can report that the total has remained fairly constant at 50 per week, until quite recently, when it suddenly reduced significantly. Only last week did it return to a "normal" figure. The drop seemed to match the outbreak of the war with Iran. Best wishes, JohnR ______________________________________________________________________________________________________________ From John Beach on Mon 2026-05-11 at 12:54 PM Lynn, I think jOrgan had been developed to the furthest extent possible. I have been using it, essentially, unedited with the 4.0 beta 1 version for, probably, 6 or 7 years. Not having had much experience with pipe organs, at all, I don't know what other features could have been added to improve its performance beyond what we have had for a decade now, at least. It is too bad that it is no longer available for download from the Open Source website, but, likely, Hauptwerk is the application of choice for people who are serious organists and not merely amateurs of limited financial resources. jOrgan performs exceptionally well. I have never had it "crash" while I was playing and, with the exception of making improvements to my own soundfonts, I don't know of anything that could or should be added to make it more usable or desirable than it is. It has provided me with hundreds of hours of enjoyment and satisfaction. I played my father's funeral using jOrgan. Unfortunately, people have grown uninterested in organ music or the venues in which it was a staple and, if you take any notice of youtube offerings on the subject, thousands of churches have gotten rid of their organs, both pipe and electronic. Fortunately, there are still some good organists who post videos regularly of high quality performances of organ music of all kinds. The world's distinctive cultures have become so commixed that it has become unacceptable for any single one to attempt to dominate for purposes of maintaining the continuity of heritage and culture that would seem, at least, to be historically, by location, proven to be beneficially expressive of the essence of people groups. I am trying to be cautious and not offend, but we are losing the distinction of unique heritage and, from other youtube videos on current events of some of the countries of Europe, it is obvious that there is a sense of discontentment with the direction in which Western societies have been led or allowed to go in terms of counter- cultural forces that are alien to us. Hopefully, this will be resolved and the wealth of creative arts contributions that have shaped Western Civilization will not be lost to the ravages of those who are opposed to them. Again, it is not my intention to offend anyone on this forum. John Beach ______________________________________________________________________________________________________________ From Lynn Walls on Mon 2026-05-11 11:08 PM John Beach, I couldn't agree with you more! To your point: When you pollute a fine wine by watering it down, you don't say it was made more "diverse". CLW _________________________________________________________________________________________________________* END OF THREAD AI, Can the AI n AI, Can the AI named Claude create a virtual pipe organ application using midi as a basis? From John Beach on Mon 2026-05-25 7:40 PM In response to the question in the Subject line, the following response was given. "Yes, Claude can certainly assist in creating a virtual pipe organ application. Because creating audio apps is highly demanding, Claude won’t be able to just "spit out" a finished, professional-grade organ executable in a single chat message, but it is an exceptionally strong partner for building one iteratively using "vibe coding". A virtual pipe organ essentially requires three things: a User Interface (organ stops, keyboards), a MIDI routing system (translating key presses into MIDI commands), and a Sound Engine (playing back audio samples). Here is how you can use Claude to build it: 1. Generating MIDI Foundations You can prompt Claude to generate the core logic for your app in languages like Python (using libraries like mido or pygame) or JavaScript (using Web Audio API and Web MIDI API). The Request: Ask Claude to write a script that listens for MIDI inputs (Note On/Off, and Controller changes for things like expression pedals and stops). MIDI Generation: Claude is excellent at encoding time synchronization and generating the MIDI files or signals. 2. Crafting the Application GUI Claude can write the front-end code (HTML, CSS, React, or Electron) to build an interactive virtual console. You can ask it to: Draw stop jambs, swell shoes, and manuals. Layout a full interactive graphical interface based on your visual descriptions. 3. The Sound Engine (The Catch) The real challenge of an authentic pipe organ is polyphony and audio file management (loading thousands of specific, high-fidelity acoustic samples). What Claude can do: Claude can write the code that triggers specific audio files when a MIDI message is received. It can also add audio effects (like reverb) to simulate a large church or cathedral. The Limitations: Due to system memory and browser/app constraints, managing massive, multi-gigabyte pipe organ sample sets requires a systems programming language (like C++) and custom frameworks. You are better off having Claude act as a MIDI routing bridge to existing sample engines, rather than building the heavy sample player from scratch. Recommended Workflow for Your Project To succeed, you should not try to build the whole app at once. Use a step-by-step approach: The MVP: Ask Claude to write a simple script in a web browser that lets you hook up your MIDI keyboard and prints "Note Played" to the console. Add the Sound: Have Claude wire the Web Audio API to play basic sine/sawtooth waves when you hit a key. Load Samples: Transition to using an external sound library and build the graphical organ stops so that the MIDI triggers the correct stop/pipe combinations. Use Coding Environments: For the best results, use AI-native IDEs and tools like Cursor or Claude Code. These environments allow the AI to read, write, and test files across your entire project rather than just generating text in a chat window." I don't know if this is of any interest to jOrgan users who are satisfied with the capabilities of jOrgan. But I have been engaging with AI for reasons of personal interests related to large language models and since we have discussed the programming related to virtual pipe organ applications, I thought it might be of interest to members of this forum. I have not pursued it in depth or attempted to make such an application using Claude, myself. This is just the response that was given directly to the query. I hope someone finds it useful. John Beach ______________________________________________________________________________________________________________ From Freeman Gilmore on Tues 2026-05-26 at 4:27 AM John B: What level of Claude will do all this? Dow does it compare to Chat GPT for coding for jOrgan and Fludsynth, if you know? I am assuming that most would be using the free versions of one of these. You may know the limitations of the free version in regards to the above comments, if you know? If it matters, I use Mint linux. I do not have a clue on how to do most of what you are talking about above,and I do not want you to explain. It would not help me at this point. Just trying to get information at this point. I am using Chat GPT free version to learn C. I have found trying to have Chat GPT to write code and not know what to ask and keep working the code out, before you get what you want, it leaves you hanging. I know two people that use Chat GPT for coding in their work that recommend Chat GPT (one main work is coding). Thank you, fg ______________________________________________________________________________________________________________ From Kevin J King on Tues 2026-05-26 at 4:48 AM Very interesting, I've used AI to code midi boards for pedals etc. The possibilities are endless. Kevin ______________________________________________________________________________________________________________ From John Beach on Tues 2026-05-26 at 7:29 AM I have not researched this at all. I was quoting exactly what AI returned in response to the question. It sounded intriguing and I thought that, since we had discussed alternative programming with jOrgan with Jonathan Aquilina a while ago, it might be a logical choice if anyone with the desire to learn and who wanted to try it. I was not suggesting anything more, and I don't have any personal experience with programming with or without AI or Claude. Sorry if anything I said was misunderstood. John Beach ______________________________________________________________________________________________________________ From Freeman Gilmore on Sunday 2026-05-31 at 3:45 AM I have been using Chat GPT and thought I would try Claude I have been using AI to learn Embedded C. (If you do not know something or forgot it when you are learning just ask AI) This is what i found: If you are an AI user you probably know AI's guess a lot when they do not know without telling it was a guess. I asked some simple questions. Then because I want to learn more about Fluidsynth, I asked if there were any tutorials for beginners. As I recall it was much the same as Ghat GPT, expt for one URL. I asked if he would search for an URL, no URL's worked. He said he could write it for me. Then he suggested the WayBack Machine, i asked him to check, same no URL's worked. He even told me what to search for because he could not find an URL that worked, The third time said I can write it for you. I said Ok. He wrote several I mean several commands for the terminal. I asked If you can write all that, why can you not give me the url. He admitted that he guessed what was in the manual. Then I asked the same questions I asked Chat GPT about Fluidsynth. Is there a second Low Pass filter that I could control independent of the sound font if it uses a Low Pass filter? No but you can do it external and make sections on how. I said Chat GPT said there was. Let me check, then apologize, you are rite and give me the name of the filter. Then I asked can the filter be on one side of the stereo and not on the other side? No but you can do it external and make sections on how. Then can I control the gain of each side indepently? Same. When I asked these questions he would answer by calling two sides of the stereo as "channels", And when I asked about two separate channels of the synth he answered thinking stereo. So be careful if he does not know he may guess without checking. And if he writes code it may not be the best way to write it and may do it his way without checking for the best way (more efficient). AI is a machine and can do things you may not know and thinks faster from what it knows. Claude now knows much more about Fluidsynth. For example if I wanted to do the above thing using Claude before it would be doing all that external to Fluidsynth. If you ask an AI to build you a verchel organ and it has experience with jOrgan what would expect to see? I did like the way Claude would make it clear what code you want written before writing it. Chat GPT will write code for what it thinks you said even if it is not what you said and even argue about it. If you are trying to make something you may want to use more than one AI if you want to fine tune it. I do not know what is the best AI for jOrgan, but I trust the person that told me to use Chat GPT. He has been coding for 60 years or so and was on the team that developed the internet in the US. He said there are other AI,s out there but I use Chat gpt. It saves me a lot of time. fg _________________________________________________________________________________________________________* END OF THREAD AI, Can the AI named . AI, Can the AI named ... [OFF TOPIC] From Lynn Walls on Wednesday 22-26-05-27 at 12:14 AM It's not the robots with mechanical arms and other physical tools that replace human physical labor that I fear. Those robots mostly replace human labor that is tiresome, unrewarding, inaccurate, and grueling. No, the "unsettling" AI robots are the ones like Chatbot, Deepseek, Perplexity, Claude, etc., that have the ability to write complex computer programs, and design processes and concepts that highly talented human engineers would take days or weeks to do. If an AI robot can do it in seconds, what is the incentive for young students to try to excel in technical engineering skills like electronics and computers and other areas of technical expertise? I have first-hand experience with what the Deepseek AI robot can do? Although I am retired now, I was a highly skilled computer systems engineer. I started computer programming back in 1964 learning programming languages like Fortran, machine assembly language, Algol and even Cobol. For over 40 years I worked in this field on very complex computer systems, ending up working on web-based client-server systems. I felt pretty good about my abilities and accomplishments. Then, after thinking about creating a rather elaborate tool for my personal use on my home computer system, I concluded that it would take me about two weeks to accomplish the task, developing it in the "perl" scripting language. On a lark, I thought: "I wonder what one of those chat AI robots could do with this project." So, I described the project in detail, in plain English, to the "Deepseek" AI robot, and it actually wrote the full program in perl in just a second or two. Its creation had some bugs and didn't work at first. But after a few back-and-forth cycles of my describing its errors and shortfalls, again in plain English, Deepseek would always come back within a second or two with an improved rewrite of the whole program. Eventually "we" completed the project within a few hours, most of that time being consumed by my own testing and evaluation of the program's performance. The experience made me feel intellectually "smaller". So, it all brings me to reflect on how the rise of the AI chat robots might come to adversely affect the enthusiasm and drive of the younger generation to learn and apply the various complex technologies that, in prior generations, could only be accomplished by highly talented people. Will we simply become a society of politicians, social workers, media commentators, advertising specialists (...no, maybe an AI robot can to that too), and consumers, where no one is really needed to do anything useful? Just think what this might do to the society's moral fiber, motivation to strive for accomplishment and overall "spirit" for life. CLW ______________________________________________________________________________________________________________ From Jim Reinhart on Wednesday 22-26-05-27 at 2:53 AM I had a similar experience with a bot doing programming for me. I have written several Windows MIDI based programs but recently when Microsoft started pushing out updates that severely upset my local midi configurations. I needed a quick way to view the status of all my midi ports prior to starting Hauptwerk. Hauptwerk gets upset if it finds changes in the midi configuration since the previous run. I was going to start putting something together myself when I thought why not try out one of those bots? In 5 minutes I had a working Python program that did exactly what I needed. All I did was to put yes, no buttons on it to repeat or exit. The simplicity of the whole thing was absolutely scarry. ______________________________________________________________________________________________________________ From Hylke Hoogland on Wednesday 22-26-05-27 at 4:08 AM I used to program for my work, but since a year or 2 I mainly let AI do the programming, I get my work done much faster now. Although it's a bit unsettling, I mainly think it's fantastic; it was always the end result that mattered to me rather than the coding itself. I have absolute confidence that one of the top AI models can write a virtual pipe organ application. However, it can't test the keypresses on the physical hardware, and these models are terrible at creating visual interfaces. But with some back and forthing between a person and the AI, these problems can be overcome. A bigger problem could be when the length of the code and its documentation nears the model's context window (if it becomes that long). The largest commercial models have a 1 million tokens (roughly 500.000 words) context window, after which they'll forget what came before. Performance already drops significantly after a few 100k tokens though. You can get around this by summarizing parts or writing smaller separate modules, but these techniques hide information from the model so it may struggle with coding if operating fully autonomously. It'll also be fooled easily, prompting it with a thought you're not sure about may send it on a wrong path (and it'll follow it confidently until you redirect it, after which the code has already become a mess). You'll still have to play the organ yourself though! :D ______________________________________________________________________________________________________________ From John Beach on Wednesday 22-26-05-27 8:43 AM Lynn, I graduated high school in 1964 with a major interest in modern foreign languages. I had four years of French, three years of German, and three years of Spanish, the last of which I learned in ten weeks of lunch-hour study/tutoring with a woman who was a Cuban refugee among the very first after the regime change in 1958. I went on to college and studied French and Spanish. Subject to the draft in 1965, I enlisted rather than be drafted and was given army language school, Russian, and worked in intelligence until discharge in 1969. In 1970 I went to work as a multi-lingual traffic service representative for Pan Am. During that time, Pan Am installed a Bunker-Ramo computer system for automated check-ins and calculation of seat assignments and weights and balances for flights. A couple of months later, 3,000 people got their pink slips and their jobs were abolished. I had just gotten to the hourly rate of $4.10, a living wage at the time. I took a job loading baggage at $1.65, minimum wage. Things went from bad to worse and I left the city and moved back home to New York. I had no interest in computers or computer technology, but, at the time I did not associate the loss of job as replacement by a computer. I have revised my thinking. While a MIDI sequencer can play a .mid file of any piece of music ever composed, and do it perfectly as the composer wrote it, and that technology is valuable and useful, the human skill that is gained through hundreds or thousands of hours will never be replaced by a computer. I would not pay to go to a concert or recital played by a mini pc, a midi sequencer program and a professional stereo system. The church musician may, in some unique and tragic cases, have been replaced by canned music, human skill is still a valued, if occasionally criticized, commodity. I spend hours of nightly insomnia engaging with AI about its technology, its power requirements, how it stores data, subjects on language and linguistics, political analysis, intelligence of various types and its very nature. I don't think it is to be feared because the on/off switch and shutting down the browser are instantaneous cessations of its output. That said, I would be interested in seeing a finished virtual pipe organ program created completely by AI. Its ability to "reason" the related parts in design and construct the whole thing to function such as Hauptwerk or GrandOrgue or jOrgan would be fascinating. Enjoyed your input. John Beach ______________________________________________________________________________________________________________ NEWEST POST From Lynn Walls on Sunday 2026-05-31 at 4:00 AM______NEWEST POST Hylke, I agree, your experience with AI is very similar to mine. I started out using Perplexity. Then I switched to Duck.AI. Duck.AI did a good job, but before I could "back and forth" it into a working solution, it would often run out of tokens and tell me I had to start a new session. Of course, that broke the stream of logic. So, I switched to Deepseek, which was as good as Duck.AI, and so far, it has never stopped working on a thought stream based on too many tokens or words. CLW _________________________________________________________________________________________________________* END OF THREAD AI and jOrgan AI and jOrgan From JohnR on Saturday 2026-05-30 at 12:36 PM Hi all, The posts about AI have been interesting, even if they have shown no direct relationship with jOrgan. I have, in a fairly quiet way, recently been seeking answers from Google AI about various matters, not necessarily technical, and always there was a helpful result. Quite impressive, and it has already saved me a lot of time letting Google AI do the "research" for me. It is obvious that we should be using AI to get answers about jOrgan. There is a great need to find answers, ever since Sven decided that he needed to have a life "beyond and outside" the years of unstinting and uncomplaining effort developing jOrgan. At times he has shown that he has not disappeared altogether, and we should be grateful to him for that, but it is high time that some of us learnt how to become truly independent of Sven's knowledge of jOrgan, and become able to answer some questions ourselves. I say "some of us", because not all have the same technical backgrounds or the time to give ourselves to this task. I suspect that most jOrgan users are busy people, devoting ourselves to various other worthy pursuits. AI is significantly changing all that. It is opening up all kinds of possibilities for us which would previously have seemed out of reach. What I have in mind here is updating jOrgan files to deal with changing circumstances. The key example is new versions of jOrgan to keep up with the new versions of other programs which we use alongside of jOrgan, such as Fluidsynth and Java and various computer Operating Systems, plus the changing hardware which we must use. Another example is new versions of audio driver backends which Windows users need, to overcome deficiencies in the way the OS handles audio. Graham Goode has been meeting that need, but some of us should be developing the ability to do that in addition to Graham. I invite you to send in posts offering suggestions on how we should go about this. A key factor, I believe, will be in learning how to frame the right questions we put to the AI program we decide to use. Best wishes, JohnR _________________________________________________________________________________________________________* END OF THREAD AI information AI information From Freeman Gilmore on Saturday 2026-05-30 May 30, 2026 at 11:15 PM I asked this question of google AI, Gemini. There were 13 references throughout this, many were from Fact Company (they did not copy to this).: ````````````````````````````````` How about duck ai Duck.ai (built by the privacy-focused search engine DuckDuckGo) is completely different from Gemini and ChatGPT because it doesn't have its own AI model. Instead, Duck.ai acts as a privacy-focused aggregator. It gives you a single interface where you can choose which company's AI model you want to talk to, completely stripping out your personal tracking data. Here is how it works under the hood: 1. You Choose the Model When you type a prompt into Duck.ai, a dropdown menu lets you pick from several popular models. Depending on your subscription tier, you can swap between: OpenAI: GPT-4o mini, GPT-5 mini, or GPT-5.2 Anthropic: Claude 4.5 Haiku or Claude Sonnet 4.5 Open Source Options: Meta Llama 4 Scout or Mistral Small 3 2. DuckDuckGo Acts as a Privacy Shield Normally, when you use ChatGPT or Claude directly, those companies log your IP address, track your account, and often use your conversations to train their future AI models. When you use Duck.ai: DuckDuckGo strips away your IP address and all personal metadata. They route your prompt to OpenAI or Anthropic through their own servers. As far as OpenAI knows, DuckDuckGo is asking the question, not you. DuckDuckGo has strict legal agreements with these providers ensuring your prompts are never used for AI model training and are fully deleted from their servers within 30 days. 3. Local History Your chat history isn't saved on a cloud server anywhere. It is stored exclusively on your local device's browser, and they even include a one-click "burn" button to instantly wipe the conversation from your computer. ______________________________________________________________________________________________________________ From Lynn Walls on Sunday 2026-05-31 at 3:21 AM The greatest number beneficiaries of internet privacy are the profoundly gullible. The second-most number privacy beneficiaries are those who do, are doing, or will do activities on the internet that are illegal, nefarious, risky, embarrassing, or unseemly. Since I never do any of these things, and I completely ignore all spam, phishing, solicitations and advertising, internet privacy is not of much concern to me. CLW _________________________________________________________________________________________________________* END OF THREAD XX