The organ as cultural heritage

Dear Hauptwerk colleagues and friends,

My dear colleague Rüdiger Loos and I have taken the liberty to open this new topic. We would like to stimulate a discussion regarding the long-term preservation of the organ as cultural heritage using the tools provided by advanced sampling techniques, such as the Hauptwerk software.

This initial post will be a somewhat longer one, and we will outline the key ideas and problems. We will conclude by kindly asking you to contribute to this issue with your comments and ideas.

Many pipe organs are cultural assets and are maintained by a lively church and concert cultural life, which includes particularly a demanding organist training and a high-quality organ building culture. In 2017 organ music and organ building was recognized as Intangible Cultural Heritage by UNESCO. Therefore, preserving this cultural heritage is an important task. Organs as complex technical devices are subject to the influence of time. They can be destroyed by weather conditions, poor maintenance, woodworm infestation, vandalism or even fire. They are subject - as often the most serious factor, at least over longer periods of time - to the respective taste of the time and changing ideas of an ideal organ, as they were expressed, for example, particularly strongly in the so-called “Orgelbewegung” in the last century. We therefore consider it useful to document not only the technical apparatus of an organ, but also its specific sound, perhaps even as a function of the respective era, and to preserve it in the long term for the next generations.

Today, with tools such as the Hauptwerk software, it is possible to create a digital replica (or sometimes called digital twin) of the sound (sample sets), which comes very close to the original sound. With such a system, an organist can experience the original sound and to a large extent also the functionality of an instrument in a good approximation and use it for exercise, performance, or even for scientific purposes.

In this forum, we all know about the very encouraging results. Digital replica of the sound of a whole series of historically important instruments are available in high quality. Details such as audio-multichannel techniques, sensitivity to key velocity, approaches by means of mathematical deconvolution procedures with the acoustic impulse response of the respective room housing the organ, shall not be addressed here. Nor will we go into the degree of sophistication in the construction of the digital replica - by way of example, let us just mention that some digital replicas take into account wind consumption in a physically high-quality simulation and are thus able to reproduce the changes in pitch and timbre during full-fingered playing with many stops, the changes in sound caused by the degree of opening of swell shutters, and, more recently, even the coupling of sounding pipes.

We guess it is fair to say, that the “Hauptwerk” product by Milan Digital Audio is the leading software. In the meantime, there exists quite a number of very good quality sample sets of historically relevant instruments for Hauptwerk.

What is missing is the sustainability behind it. The sample set manufacturing companies, and also Milan Digital Audio, are mostly very small companies. They can disappear from the market at any time. The Hauptwerk project therefore does not fulfill on its own the conditions one has to set for a long-term preservation of sound libraries.

We did a local brainstorming with several organists and people interested in this subject. Many questions arose - here is a first list, which might be amended by your input:

• Which sufficiently high-quality sample sets already exist? What criteria agreed by the community exist to evaluate the quality of a sample set? Which other real instruments should be recorded with priority?
  
• How does one deal with the fact that both the software producers and the sample set producers are usually very small companies, often one-person enterprises with owners of advanced age, which can disappear from the market any time?
  
• How does one anticipate the changing software standards over time to ensure that current sample sets remain usable in the future, maybe 20, 50 or 100+ years from now?
  
• How can commercial interests of software and sample set providers be reconciled with the goal of open, long-term archiving? For example, can open standards and a secure institution for long-term storage be defined that sample set providers agree to use?
  
• How can the problem of long-term storage be organized? How would it be financed?
  
• What research needs to arise from these questions?

Our wish would be that we can have a fruitful discussion about this here in this thread. Please feel encouraged to provide your input.

Burkard Hillebrands & Rüdiger Loos

Thank you for your thoughts on this matter. Clearly the instrument that we all love is facing daunting challenges. I applaud your effort to ensure the future viability of the organ. One thought that occurs to me is that this discussion might benefit by soliciting opinions from people who may not read this forum. One person who may have some valuable insights is Michael Barone.There are others as well.

Dear voet,

Your suggestion is very important. Please forward our contribution to anyone you think might be interested in this topic. I am also happy to communicate through other channels.

Thank you for your thoughts on this matter. Clearly the instrument that we all love is facing daunting challenges. I applaud your effort to ensure the future viability of the organ. One thought that occurs to me is that this discussion might benefit by soliciting opinions from people who may not read this forum. One person who may have some valuable insights is Michael Barone.There are others as well.

Michael Barone is aware of Hauptwerk, and also a friend of mine - we got to know each other first when I agreed to publish some music he’d commissioned (Henry Martin’s 24 Preludes and Fugues for organ)and then later much better as I eventually moved to the Twin Cities (Minnesota, USA) and we regularly cross paths.

Without meaning to speak for him, I think my impression of his opinion of Hauptwerk is mixed… he’s always tried to be polite when I talk about it, but I wonder what he really thinks. Wouldn’t hurt to ask.

In the USA at least, Charlie and I are committed to preserving at-risk organs via HW whenever possible - to date we’ve done several: all 3 organs at Immanuel Presbyterian Church in Los Angeles - a small Hradetsky in one chapel, medium-sized Schlicker in another chapel, and most importantly the IV/62 EM Skinner in the sanctuary, the instrument played by composer and organist Clarence Mader. All of these are at risk, as the church no longer has a viable congregation, and is being kept afloat by endowments. Some of the campus is rented out to outside groups, and the church hosts a food bank. Inevitably, someday the property will need to be sold.

Last summer we recorded a small 1893 Steere organ down in Owatonna, MN. That church also was in financial trouble, and this spring it was sold. The new congregation, while respectful of the heritage and history, has no desire to use the organ, and hopes to sell it and preserve it that way. We recorded all of the pipes in 4 channels (close and at the back of the balcony - it’s a small room).

This summer we recorded another at-risk organ, the 1877 Johnson at Our Lady of Good Counsel sisters chapel. The sisters membership has aged, and declined so fast (with little or no new sisters joining) that the house is no longer viable. The property will be sold eventually, and one hopes the organ (the largest surviving Johnson) can be moved somewhere else and continue to be used. Due to the stunning acoustic, we recorded in 8 channels (although it remains to be seen if all 8 will be used)

Whe do the historic organs as a labor of love - some of them do make a little money, but that’s not the main point - we do it to preserve the unique sound of them in the room.

We are happy to consider other historic and at-risk instruments for documentation, and hope in some small way to preserve the sounds for future generations to explore and enjoy.

An interesting discussion, but not an easy process! One problem is who is to say which organ or organs are culturally important? If each of us had to choose which 10 Hauptwerk organs had to be preserved, I'm sure each of us would come up with a different answer. So the solution for me would be not to judge - if a database were to be made, then every organ should be included on it if possible.

Then the problem becomes how will they be stored? The prevailing answer for storage of photos is to have them printed. All digital storage media, whether discs or hard drives or what have you, has a finite shelf life. The .wav file that Hauptwerk uses has been a pretty long lasting medium. But 100 years from now, will there be computers that can play it? And more to the point, will there be a Hauptwerk like program that can load sample sets created now?

The loss of the people involved in Hauptwerk is a very real concern as well. If Martin or any of the sample set producers were to fall ill and die, it would fall to their heirs to continue on the legacy. But what if the password to the computer isn't written down? How will future computers access iLok encrypted files if iLok no longer exists?

So, the ideal would be to create an unencrypted database that is double backed up, and continually updated to ensure the original samples are protected and at every moment are able to be played on the most recent technology. But then who will pay for this to be done? It would require payment to the sample set producers and Hautpwerk, as well as legal agreements as to how the files may be able to be distributed in the future. Then it needs payment for the staff who will manage the computers, sample storage, and continual actualization. Where will this money come from?

So, it is a difficult question, but one worth asking. In the present, individual Hauptwerk users are storing those samples and keeping them alive by being users of the system. In the case of loss of a sample set producer, they will still have access to the work as long as they have Hauptwerk and the original files. As long as iLok exists, then unencryption of the files will be possible. So the more we can do to preserve these organs the better, but there are always risks. But to some extent though, I think it is just as it is for photos. The best preservation is that of the original organ, not its likeness in digital media.

Interesting discussion. Here is a technical point of view.

All the audio files in a sample set are heavily processsed (denoising etc). Processing and modeling is never perfect. Standards (what is 'good enough'?), tools and techniques can evolve and improve over time.

Therefore, a long-term storage project should ask for raw, unprocessed recordings. And preferably a lot of metadata that can be used as inputs for processing and modeling:
- Brand & model of each recording microphone and all other audio equipment
- Recent calibration file for each mic
- Exact recording location and height of each mic.
- Temperature and humidity traces during the recording sessions
- Date/time and mic channel of each recorded audio file

I think a good wind model is extremely important, if we want to capture the real sound signature of an organ. If possible, the following details should be included (in consultation with the organ builder/restorer)
- default pressure in each bellow, and approximate weights of the bellow top boards
- dimensions of all windchests, wind channels, bellows etc
- a map of the location of each pipe on the windchest
- pipe mensuration

We need reference recordings to validate any modeling. These reference recordings should be made with the same microphone setup as used for the recording of samples:
- Full organ
- Common stop combinations for ranks that are on the same windchest
- Swell box open/closed
- Tremulants on/off

Finally, a collection of impulse responses should ideally be included, because those can be used in a surround setup to model how the sound signature changes when you move through the building.

Dear voet, Jonathan, Iarason2, and josq,

thank you very much for your replies and kind advice.

It would be my pleasure to get in touch with Michael Barone - could you send me a PM on how to reach him?

Also, as you said, Iarason2, and I think we all agree on this, preserving the original instrument, if at all possible, is the top priority. And I very much appreciate the work that you are doing. We have the problem of a constant loss of instruments.

Overall, we would have to address a whole range of issues. There is the question of how we select instruments, for one thing. And then there is the issue of long-term storage.

As you said, Josq, for long-term storage, we need unprocessed raw data including metadata, information about the wind model used, and even impulse responses from the building and reference recordings, etc. So we need to discuss how to define the appropriate data structure that will last over a very long period of time. And there is the issue of hard- and software: we don't know what operating systems and even Hauptwerk or a Hauptwerk successor software will look like in 50 years or so - maybe we will be using ubiquitously available quantum computers, who knows? However, if we stick to open formats, this problem can probably be solved in the future. Please remember that we have emulators for Ataris and Z80 processors and the CP/M operating system - that was the mainstream hard- and software when I started my career. And we need a kind of suitable institution, trusted by the sample set producers and software producers. Perhaps national organ societies such as the AGO in the U.S. or the GDO in Germany could be enlisted for this task, at least in the longer term (I fear this is a difficult task).

Probably a more critical problem is funding. I can't suggest any solutions here. Clearly, financial resources need to be secured, and this over a long period of time. However, I believe there is some reason which allows us to be optimistic. The issue of long-term preservation of digital data in the area of cultural assets seems to be coming more and more to the fore. I can speak for Germany: here the government has put a substantial amount of money into a consortium called "NFDI4Culture," which is looking at long-term data preservation in the context of tangible and intangible cultural assets, see https://nfdi4culture.de. This initiative aims to build an on-demand research data infrastructure that serves the community of interest, ranging from architecture, art history, and musicology to theater, dance, film, and media studies. I know of a funded project on the "Corpus Vitrearum Medii Aevi.“ It is about taking high-quality digital photographs of all the medieval stained glass windows in Germany and providing long-term storage and access to that data - very reminiscent of the topics we are discussing here.

So there is hope. But we have to be active.

Dear voet, Jonathan, Iarason2, and josq,

thank you very much for your replies and kind advice.

It would be my pleasure to get in touch with Michael Barone - could you send me a PM on how to reach him?

Also, as you said, Iarason2, and I think we all agree on this, preserving the original instrument, if at all possible, is the top priority. And I very much appreciate the work that you are doing. We have the problem of a constant loss of instruments.

Overall, we would have to address a whole range of issues. There is the question of how we select instruments, for one thing. And then there is the issue of long-term storage.

As you said, Josq, for long-term storage, we need unprocessed raw data including metadata, information about the wind model used, and even impulse responses from the building and reference recordings, etc. So we need to discuss how to define the appropriate data structure that will last over a very long period of time. And there is the issue of hard- and software: we don't know what operating systems and even Hauptwerk or a Hauptwerk successor software will look like in 50 years or so - maybe we will be using ubiquitously available quantum computers, who knows? However, if we stick to open formats, this problem can probably be solved in the future. Please remember that we have emulators for Ataris and Z80 processors and the CP/M operating system - that was the mainstream hard- and software when I started my career. And we need a kind of suitable institution, trusted by the sample set producers and software producers. Perhaps national organ societies such as the AGO in the U.S. or the GDO in Germany could be enlisted for this task, at least in the longer term (I fear this is a difficult task).

Probably a more critical problem is funding. I can't suggest any solutions here. Clearly, financial resources need to be secured, and this over a long period of time. However, I believe there is some reason which allows us to be optimistic. The issue of long-term preservation of digital data in the area of cultural assets seems to be coming more and more to the fore. I can speak for Germany: here the government has put a substantial amount of money into a consortium called "NFDI4Culture," which is looking at long-term data preservation in the context of tangible and intangible cultural assets, see https://nfdi4culture.de. This initiative aims to build an on-demand research data infrastructure that serves the community of interest, ranging from architecture, art history, and musicology to theater, dance, film, and media studies. I know of a funded project on the "Corpus Vitrearum Medii Aevi.“ It is about taking high-quality digital photographs of all the medieval stained glass windows in Germany and providing long-term storage and access to that data - very reminiscent of the topics we are discussing here.

So there is hope. But we have to be active.

I think there should be contact information for Michael Barone on Pipedreams.org

A few more thoughts…

- funding… this is a bit of a puzzle. What we do is bear the cost of doing the recordings ourselves, but ask the venues to allow us to sell the product after editing. That nearly always covers our costs (and sometimes even makes a bit of profit!). That is fine for the short term, but someone needs to hear the cost and responsibility of archiving the data after our demise

- We would certainly do more historic instruments if we can get access. We’d even work with remote partners to do the recording to our specifications.

- And, I think now I want to make some written instructions as to where the data will go when I’m gone. That would need to be someone willing to preserve the history…

Thanks for raising this topic!