4 ms·

Physical modelling has come a long way since it first appeared. Many of the instruments included in the Arturia V Collection (one of the most common collections

Physical modelling has come a long way since it first appeared. Many of the instruments included in the Arturia V Collection (one of the most common collections of instrument plug ins) are made using physical modelling [1]. Modartt's products (even more so than Arturia) offer amazing control over the physical model, with the ability to configure per-note parameters [2]. Speaking of virtual pipe organs, for Organteq to compete with the market leader, Hauptwerk (which is sample-based) there needs to be a good choice of expansions though. With Hauptwerk, you can download hundreds of different organs from all around the world [3]. Modartt has an expansions marketplace for Pianoteq [4], why is the same thing not available for Organteq? [1] https://www.arturia.com/phi https://www.arturia.com/phi [2] https://www.modartt.com/pianoteq_pro https://www.modartt.com/pianoteq_pro [3] https://www.hauptwerk.com/instruments/ https://www.hauptwerk.com/instruments/ [4] https://www.modartt.com/pianoteq_instruments https://www.modartt.com/pianoteq_instruments

Darthy 3y ago
I don't think [4] is a marketplace in the sense that the expansions are from third party creators. They are all created by Modartt themselves, one by one over the past 15 years. After all, you can't just sample another instrument, there is a lot more work involved. That might also be the answer why this doesn't yet exists for Organteq. It just takes time. But your reason why you need good selection of expansions might be valuable market insight for Modartt. I know they appreciate feedback, I've talked to them before, so I'd suggest you tell them your concerns.
- HPsquared 3y ago
  Tuning the model parameters to sound like a sampled piano sounds like an interesting optimization problem.
- mastazi 3y ago
  > But your reason why you need good selection of expansions might be valuable market insight for Modartt. I know they appreciate feedback, I've talked to them before, so I'd suggest you tell them your concerns. Good point. Disclaimer: I'm just a hobbyist but I know a few professional organists. TL;DR: each organ is unique, pipe organs are not produced in series like pianos. The main reason for people use something like Organteq or Hauptwerk, is usually to practice organ at home. Unlike what happens with other instruments, it is very rare for an average person to own their own pipe organ, so digital reproductions are the way to go for pretty much everyone. But here's the problem: organs are very, very different one another. With piano, you could practice e.g. on a Kawai at home then perform a recital on a Yamaha; the two pianos don't have the same sound and the same action, but they are close enough that you can practice on one instrument then perforom on another. This is not the case with organs: a baroque organ and a romantic organ will be completely different in terms of registrations, pedalboard extension and shape, manuals etc. etc. - not only that but each pipe organ ever built, is a completely custom instrument. For example, all Steinway D-274s ever built, despite some differences due to the year and place of construction, are still the same model (model D-274). But with organs, it's almost never the case that the same organ is built twice. Each single instrument is completely unique, and that's why we identify them by opus number, and we have catalogues of all the organs ever built, like for example this one https://ohta.org.au/organs-of-australia/ https://ohta.org.au/organs-of-australia/ So, the main problem for someone setting up a practice organ is to have access to a digital model that is either that exact organ, or at least a very similar instrument to the one they will use to perform in public. For hobbyists like me who don't plan to perform, there is still a need to at least have a certain type of organ available, because of the great diversity of organs.
  - jacquesm 3y ago
    Part of the reason is that an organ isn't 'just' an instrument, it is a building which happens to contain an organ. In order to properly model an organ in a particular setting you'd need to model the building as well and that is getting into a wholly different level of complexity. Pianos are more or less designed to sound 'the same' (with some obvious differences between major brands and some pianos that have been constructed very explicitly to sound different, such as the Klavins instruments). https://klavins-pianos.com/products/ https://klavins-pianos.com/products/ So within reason you can model a particular piano just by specifying make and model and look up a bunch of parameters in a table. With an organ in a 'virtual' church you could specify the exact disposition (the number and kind of sets of various pipes and other complications) of the organ, and you could supply some basic parameters of the space (size, area, placements of the ranks and so on). But I suspect that if you did that you'd not get even close to what the real instrument in its actual setting sounds like. Organs are super temperamental, air pressure, humidity, temperature, intonation, various stuff that is broken (almost no mechanical organ is ever in perfect shape), mechanism sounds (which can appear very different even within the same make and model) and so on. This is a stupendously hard problem and as a happy user of Pianoteq I'm impressed with Modartt even attempting this and I really hope they will succeed in being able to switch seamlessly between different venues and to place an organ of a different manufacturer into an existing church (just to give one example of what that would make possible). But I'm not going to hold my breath for it and sample based organs (including the free GrandOrgue) still hold a significant edge compared to Modartt's organ simulator. But for piano it is now so good I almost prefer it to the real thing for practice. Not that I'm such a discerning user, more that I've grown up with pianos and they're always (unless very recently tuned) a bit out of tune and that can be quite annoying. The software based one is always perfectly tuned and sounds as good or even better than the best sampled instruments (personal favorite from the sample side of the aisle: Yamaha P-515 Bosendorfer sample set is extremely good).
  - PaulDavisThe1st 3y ago
    Another reason for physically modelled or excellently sampled pipe organ synthesis: even when an actual pipe organ exists, maintenance has historically become more and more of an issue for its owners, and thus implementing it digitally (with huge attention to detail) can be a way for an organ's unique character to live on even when its owners can no longer afford to keep it functional.
PaulDavisThe1st 3y ago
The open source Aeolus physically-modelled organ is at least as good as any other pipe organ synthesis engine, unless you insist on glossy UIs (it does have a graphical UI though). https://kokkinizita.linuxaudio.org/linuxaudio/aeolus/ https://kokkinizita.linuxaudio.org/linuxaudio/aeolus/
- chrismorgan 3y ago
  Aww, seems to need a MIDI input, guess I’ll have to wait until I get home in March to try it. Impressive tininess. The Arch Linux package is 879.54 KiB. And yes, there’s dynamic linking with the contents of other packages like libx11, jack and alsa-lib, but I think the total most people will be adding will be 1.11 MiB. But then, I had to replace jack2 with pipewire-jack to get it to work, and that saved 1.18 MiB, so installing Aeolus literally saved me disk space! (Yeah, I know I could have done jack2 → pipewire-jack independently, and yeah, I know I’ve added a few files to /var/cache/pacman/pkg/ so that I didn’t actually save disk space, but hey, I like the narrative.)
- mastazi 3y ago
  Speaking of open source pipe organs, there is also Grandorgue, however this one unlike Aeolus is sample-based https://github.com/GrandOrgue/grandorgue https://github.com/GrandOrgue/grandorgue
  - LegitShady 3y ago
    so many interesting instruments so little time. found some demos on youtube and this one sounds great.
- LegitShady 3y ago
  never heard of this VST before and it seems likely to have predated my interest in music production. I found a few videos online and aside from clunky UI seems pretty decent with some reverb. When I have some time I'll compare it against some sample based virtual instruments.
prvc 3y ago
There is also Aeolus (free software) for those who wish to have complete control over the model!
tgv 3y ago
Why no expansion packs? I guess because the product isn't popular. I tried it, but Organteq just doesn't sound good. Hauptwerk is so much better. It took Modartt many years to get a decent piano model. For a long time, the more critical plugin users disliked Pianoteq. It had a sound best described as plastic (the material), synthetic. Since version 5, the sound has improved, and since Pianoteq can do things even the largest sample library can't (we're talking .5TB for some libraries!) at a fraction of the CPU and memory cost, it has become a lot more popular. Organteq just isn't there: it sounds synthetic, and its reverb is really lacking. It's also not very CPU friendly. It also doesn't really compete on price, so I expect no change unless Modartt invests heavily, but since the market is smaller than for piano sounds, don't hold your breath.

Darthy 3y ago

I don't think [4] is a marketplace in the sense that the expansions are from third party creators. They are all created by Modartt themselves, one by one over the past 15 years. After all, you can't just sample another instrument, there is a lot more work involved. That might also be the answer why this doesn't yet exists for Organteq. It just takes time. But your reason why you need good selection of expansions might be valuable market insight for Modartt. I know they appreciate feedback, I've talked to them before, so I'd suggest you tell them your concerns.

HPsquared 3y ago
Tuning the model parameters to sound like a sampled piano sounds like an interesting optimization problem.
mastazi 3y ago
> But your reason why you need good selection of expansions might be valuable market insight for Modartt. I know they appreciate feedback, I've talked to them before, so I'd suggest you tell them your concerns. Good point. Disclaimer: I'm just a hobbyist but I know a few professional organists. TL;DR: each organ is unique, pipe organs are not produced in series like pianos. The main reason for people use something like Organteq or Hauptwerk, is usually to practice organ at home. Unlike what happens with other instruments, it is very rare for an average person to own their own pipe organ, so digital reproductions are the way to go for pretty much everyone. But here's the problem: organs are very, very different one another. With piano, you could practice e.g. on a Kawai at home then perform a recital on a Yamaha; the two pianos don't have the same sound and the same action, but they are close enough that you can practice on one instrument then perforom on another. This is not the case with organs: a baroque organ and a romantic organ will be completely different in terms of registrations, pedalboard extension and shape, manuals etc. etc. - not only that but each pipe organ ever built, is a completely custom instrument. For example, all Steinway D-274s ever built, despite some differences due to the year and place of construction, are still the same model (model D-274). But with organs, it's almost never the case that the same organ is built twice. Each single instrument is completely unique, and that's why we identify them by opus number, and we have catalogues of all the organs ever built, like for example this one https://ohta.org.au/organs-of-australia/ https://ohta.org.au/organs-of-australia/ So, the main problem for someone setting up a practice organ is to have access to a digital model that is either that exact organ, or at least a very similar instrument to the one they will use to perform in public. For hobbyists like me who don't plan to perform, there is still a need to at least have a certain type of organ available, because of the great diversity of organs.
- jacquesm 3y ago
  Part of the reason is that an organ isn't 'just' an instrument, it is a building which happens to contain an organ. In order to properly model an organ in a particular setting you'd need to model the building as well and that is getting into a wholly different level of complexity. Pianos are more or less designed to sound 'the same' (with some obvious differences between major brands and some pianos that have been constructed very explicitly to sound different, such as the Klavins instruments). https://klavins-pianos.com/products/ https://klavins-pianos.com/products/ So within reason you can model a particular piano just by specifying make and model and look up a bunch of parameters in a table. With an organ in a 'virtual' church you could specify the exact disposition (the number and kind of sets of various pipes and other complications) of the organ, and you could supply some basic parameters of the space (size, area, placements of the ranks and so on). But I suspect that if you did that you'd not get even close to what the real instrument in its actual setting sounds like. Organs are super temperamental, air pressure, humidity, temperature, intonation, various stuff that is broken (almost no mechanical organ is ever in perfect shape), mechanism sounds (which can appear very different even within the same make and model) and so on. This is a stupendously hard problem and as a happy user of Pianoteq I'm impressed with Modartt even attempting this and I really hope they will succeed in being able to switch seamlessly between different venues and to place an organ of a different manufacturer into an existing church (just to give one example of what that would make possible). But I'm not going to hold my breath for it and sample based organs (including the free GrandOrgue) still hold a significant edge compared to Modartt's organ simulator. But for piano it is now so good I almost prefer it to the real thing for practice. Not that I'm such a discerning user, more that I've grown up with pianos and they're always (unless very recently tuned) a bit out of tune and that can be quite annoying. The software based one is always perfectly tuned and sounds as good or even better than the best sampled instruments (personal favorite from the sample side of the aisle: Yamaha P-515 Bosendorfer sample set is extremely good).
- PaulDavisThe1st 3y ago
  Another reason for physically modelled or excellently sampled pipe organ synthesis: even when an actual pipe organ exists, maintenance has historically become more and more of an issue for its owners, and thus implementing it digitally (with huge attention to detail) can be a way for an organ's unique character to live on even when its owners can no longer afford to keep it functional.

PaulDavisThe1st 3y ago

The open source Aeolus physically-modelled organ is at least as good as any other pipe organ synthesis engine, unless you insist on glossy UIs (it does have a graphical UI though). https://kokkinizita.linuxaudio.org/linuxaudio/aeolus/ https://kokkinizita.linuxaudio.org/linuxaudio/aeolus/

chrismorgan 3y ago
Aww, seems to need a MIDI input, guess I’ll have to wait until I get home in March to try it. Impressive tininess. The Arch Linux package is 879.54 KiB. And yes, there’s dynamic linking with the contents of other packages like libx11, jack and alsa-lib, but I think the total most people will be adding will be 1.11 MiB. But then, I had to replace jack2 with pipewire-jack to get it to work, and that saved 1.18 MiB, so installing Aeolus literally saved me disk space! (Yeah, I know I could have done jack2 → pipewire-jack independently, and yeah, I know I’ve added a few files to /var/cache/pacman/pkg/ so that I didn’t actually save disk space, but hey, I like the narrative.)
mastazi 3y ago
Speaking of open source pipe organs, there is also Grandorgue, however this one unlike Aeolus is sample-based https://github.com/GrandOrgue/grandorgue https://github.com/GrandOrgue/grandorgue
- LegitShady 3y ago
  so many interesting instruments so little time. found some demos on youtube and this one sounds great.
LegitShady 3y ago
never heard of this VST before and it seems likely to have predated my interest in music production. I found a few videos online and aside from clunky UI seems pretty decent with some reverb. When I have some time I'll compare it against some sample based virtual instruments.

prvc 3y ago

There is also Aeolus (free software) for those who wish to have complete control over the model!

tgv 3y ago

Why no expansion packs? I guess because the product isn't popular. I tried it, but Organteq just doesn't sound good. Hauptwerk is so much better. It took Modartt many years to get a decent piano model. For a long time, the more critical plugin users disliked Pianoteq. It had a sound best described as plastic (the material), synthetic. Since version 5, the sound has improved, and since Pianoteq can do things even the largest sample library can't (we're talking .5TB for some libraries!) at a fraction of the CPU and memory cost, it has become a lot more popular. Organteq just isn't there: it sounds synthetic, and its reverb is really lacking. It's also not very CPU friendly. It also doesn't really compete on price, so I expect no change unless Modartt invests heavily, but since the market is smaller than for piano sounds, don't hold your breath.