[Darthy]

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.

[mastazi]

> 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/

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.

[weinzierl]

> "it is very rare for an average person to own their own pipe organ"

Rare, but not completely unheard of. Donald Knuth has one in his home.

https://www-cs-faculty.stanford.edu/~knuth/organ.html

[jacquesm]

I know more than one person that has one. Usually these are home-built by the person playing them or their spouse. Every now and then they are up for sale but they usually are large enough that buying one is a pretty major decision.

[mastazi]

also, I wrote _average_ person and DK certainly isn't one ;-)

[PaulDavisThe1st]

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.

[jacquesm]

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/

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).

[weinzierl]

> "organs are very, very different one another."

Anna Lapwood addressed this in a recent TikTok. She said that for every new organ, she takes a whole day before the concert to familiarize herself with the instrument, then a night of sleep and a good part of the day the of the concert. She is a professional organist who toured the word, so for almost everyone else it will certainly take longer.

[HPsquared]

Tuning the model parameters to sound like a sampled piano sounds like an interesting optimization problem.