It's rare these days to experience sound incorporated as an element into Public Art Works. I'm so very inspired by the impressively large-scale public sound installation by Tony Myatt, and Max/MSP wizard, Oli Larkin of the Music Research Centre (University of York, UK). The Morning Line is a huge outdoor sculpture, concealing a fifty-three speaker sound system driven by Max, which performs multi-spatial compositions by some of the world’s top sound artists. The Morning Line is touring internationally, currently in Vienna. Oli and Tony were kind enough to take the time during their massive set-up schedule, days before opening, to sit down and go into detail about the creation and technical management of such an ambitious and wonderful public art piece.
Can you give us a brief description of your project?
Tony: The Morning Line is a very large project, part of a series of art pavilions commissioned by art collector and patron Francesca von Habsburg’s Thyssen-Bornemisza Art Contemporary Foundation [T-B A21].
This is the second of four pavilions that she’s planning. It was designed by Matthew Ritchie, and architects Aranda/Lasch, from New York. They also worked with the Advanced Geometry Unit of Arup Engineering.
It began as a project to establish a collaborative structure that could be the basis for a whole variety of new art works. From its very conception — years before we got involved in the project — it was always designed to have a sound element of some sort that would try to unite sculpture with music in some way.
We were approached by T-B A21 in May 2008, who asked us to put together a system that could be a platform for the composers to realize spatial works on The Morning Line, which is a 17-ton outdoor, modular aluminum structure.
We’ve done a lot of work over the years in surround-sound technologies, methods and approaches, particularly on authoring techniques, ways in which people hear surround, and in techniques that can be used to enhance spatial audio perception.
And this is the system that travels with the sculpture structure?
Correct. We showed it first in Seville at the Biennial of Contemporary Art in 2008. It was on display for a year. This is an outdoor sculpture, by the way, so we’re very grateful for the Meyer MM4XP speakers, which are waterproof. The sculpture traveled to Istanbul last year, as the centerpiece for the European City of Culture festival and was there for six months. We’re just about to open an exhibition here in Vienna, for the next six months, with a music festival of newly commissioned works.
The organization, T-B A21, have commissioned 28 composers now to write works specifically for The Morning Line. Oli, myself, Peter Worth and Dave Malham from the Music Research Centre, have been working with these composers to help them develop their ideas and concepts to be realized on the system and also to help them with the technicalities of working with spatial sound in general.
So we’ve got 10 new composers here whom we’re working with. I think that Francesca von Habsburg and T-B A21’s idea is that the piece will continue to tour and build a body of work over many years. And we’re also hoping to take it to North America at some point.
It’s all so exciting. Who curates the composers?
Tony: There have been a series of curators. T-B A21 is an art organization that collects contemporary art, so it has its own curators, but it has also commissioned sound curators to put together the music program for each venue.
For the first installation in Seville, Bryce Dessner — who plays guitar in the band The National — curated four works, and also an artist called Florian Hecker, who curated another four.
Russell Haswell curated the Istanbul international program of works, and four Turkish composers were also chosen by the MIAM institute in Istanbul for that exhibition.
For the Vienna installation Franz Pomassl has curated a series of artists including Carsten Nicolai and Christian Fennesz, and several from Eastern Europe.
We’re thrilled to be working with all these people.
So all the material is on the Mac Pro, and then going through Max. Is that correct?
Tony: Yeah. It’s all generated by our Morning Line patch.
That must be a really big Max patch!
Tony: Maybe one of the largest you’ve ever seen. [Laughs] Certainly one of the largest I’ve ever seen.
And how is the material mixed? Do you use pre-mixed multitrack files?
Tony: Because we want to make the pieces a multi-polyphonic experience in spatial terms, we don’t want to work with stereo files really. We want to work with all of the independent tracks and channels that a composer might use to put together that piece, so that we can separate out all the elements spatially across the sculpture.
Everything gets mixed on the fly, and also gets spatialized live as the pieces are playing back. So the Mac Pro contains all of the individual sound files that constitute the works. The Max patch sequences all of these sound files, spatializes them, pans them around, locates them, moves them and so on whilst the piece is being played. It’s a key feature of our approach that we can do that.
How long have you been working in Max?
Oli: Well, I’ve been using it for I’d say about nine years. And you’ve been using it for quite a long time as well, haven’t you Tony?
Tony: Well, actually, a few months before we began this project, I moved from PD to Max and since then, I’ve used Max exclusively. So we went straight in with Max 5 for this project.
Are there features in Max 5 that were beneficial to this project?
Oli:. One of the best things about Max 5 for us is the fact that it looks very good on Windows and Mac. We can build standalone applications very easily that work like a normal application on both operating systems.
As well as the software that we have that runs on our Mac Pro, we’ve also developed lots of individual authoring tools in Max. These are applications that allow the artists to import sound files and attach some spatial information to those sound files. Depending on what sound system setup they have in their own studios, they can monitor the sounds in different ways and get an idea of how it’s going to work when it’s actually on the sculpture.
So they can roughly preview their ideas…
Yes. Of course the sculpture has a completely unique sound system, so it is only a rough idea, but it really helps both us and them to be able to use these tools to quickly generate and audition the spatial data. Once that’s done, when the composer comes to fine tune their work on the actual system, we can then put the files into the Morning Line Engine - the big Max patch that runs on the main system, and it’ll play back all those trajectories alongside the sound files. If the spatialization isn’t quite right we can quickly edit the data in our authoring tools.
Do you write any of your own objects?
Tony: We’re using Jitter now, as well.
How are you incorporating Jitter into your work?
Tony: We’re working on a piece with Carsten Nicolai called “Morning Line Quanta”. The concept for it is an auralization of solar wind data. The patch downloads the latest solar wind image from NASA’s SOHO satellite, which is constantly pointed at the sun. Every few hours we get an image of the solar winds that are currently emerging from the sun. Then that data is auralized.
Oli: We use Jitter to download and map a 360 degree radial scan of the image to a new matrix, where each radial scan line from the original image is a column in the matrix and each row indicates a different scan angle.
How is the audio side presented?
The pixel data is traversed and auralized using impulse generators, and the dynamics of the impulses are controlled by the content of the pixels. It simultaneously spatializes the impulses, almost as if the sun were at the center of the Morning Line.
Tony: It produces a granular cloud of impulses that rotates around the sculpture and changes in texture every time it is played. It’s a great benefit just to be able to use Jitter to do that sort of thing.
In addition to the things that Oli mentioned, as far as I’m concerned ⎯ as the person running and managing the project in addition to being a programmer ⎯ the speed with which we are able to develop in Max is very important to us. Quite often we have to write patches very quickly in response to requests from artists to do some things that are not regular features of our software, so it’s really important for us to be able to make patches quickly and have a lot of flexibility.
When you’re dealing with something like a 50-channel sound system, being able to just manage all of those audio channels, is really crucial. So the way that we employ poly~ objects, the way that we deal with multi-channel matrices, for example, is really vital for us.
Oli: Carsten’s piece is an example of one way we can play back sound on the sculpture, but we have this section of the software where we can actually generate a score to sequence the sound files that have trajectories associated with them (what we call spatial audio objects). So that’s one other way in which some of the pieces are realized.
Because everyone has a different approach, using Max allows us to actually write each artist a separate patch to do something specific, which is something that wouldn’t be possible if we ran the system from a conventional DAW. Some of the pieces are algorithmic, some of those are playing back samples more like a typical quadraphonic system, and perhaps certain sounds that have been recorded with Ambisonic microphones and things like this.
You must be discovering a lot from approaching the surround experience from so many points of view.
Tony: I think one thing that this project has really brought home to me is that the world of surround is in a real mess at the moment. There are no real standards. There are a whole load of different approaches and different methods. Even though DAWs claim to be surround ready, their handling of multichannel formats is often specific to their interpretation of a surround format or method and can be inflexible, only accepting certain “types” of quad file, for example, restricting how multichannel files can be routed, not allowing tracks to have different number of input and output channels etc. DAWs, editors, plug-ins, panners, surround decoders are not joined-up; it can be a nightmare to simply create a good workflow and software environment for surround sound production in a studio – particularly if you want to move outside 5.1 and include height, like with Ambisonics.
What we’re finding is not just about how people work with these different methods in their own studios, but also that some methods are better or worse for certain types of applications. One of the artists we work with, Chris Watson, who’s a UK-based wildlife sound recordist [ed: some may remember him from the band Cabaret Voltaire], has made a piece called Snaefellsnes, which is based on an area in Iceland. He’s recorded quadraphonic sound underwater using hydrophones through to double MS recordings on the giant rocks, moving down the beach, wind through the grasses, mono sounds from beneath glaciers and from inside the volcano at the top of Snaefellsjokull, which, of course, is the volcano in Jules Verne’s Journey to the Center of the Earth.
Just that one piece uses a number of different recording approaches to capture surround. It needs different panners, encoding and decoding techniques and ultimately to be converted and coherently represented on the Morning Line system.
So the fact that we can run patches in Max that do all these things is a really good way of dealing with the integration of different surround sound approaches.
Are there existing Max objects, as opposed to your custom ones, that are predominately used in this project?
Oli: I’d have to say that poly~ is probably the most important object.
Why is that?
Oli: Well, for our score-based system, we need to be able to have lots of simultaneous polyphonic sounds in multiple areas of the sculpture, and some of the pieces, I think Florian Hecker’s piece in particular, uses a vast number of different sounds polyphonically, all around the sculpture. So, yeah, without poly~ we’d all have really, really bad RSI [Repetitive Strain Injury].
Tony: And a lot bigger Max patches, that’s for sure!
I hope it comes out here to San Francisco.
Tony: We’d like that, but we’ll have to wait and see where T-B A21 can take it next. We’re also linking it with another project now, which is to look more at the authoring end of the surround-sound spectrum with a project that’s going to be based on a ship.
That’s going to begin this summer, heading over to Iceland and Greenland, and then it’s going to circumnavigate the Americas, and we’re hoping to be doing projects all the way around there, many of which will be in surround sound. We’ve equipped the boat with a similar surround-sound system and surround-sound microphones. The software for the boat sound system is also implemented in Max/MSP.
We’re just about to begin a program of commissioning new works from artists, and also projects from scientists who are interested in marine life, environmental studies and so on.
In addition to the Morning Line, T-B A21 are launching a project in Vienna called the Temporäre Kunsthalle, which is a huge, temporary — by temporary I think they mean five or six years, not a few months — art gallery. The current plans are to focus this exhibition space on sound art works in some of its larger spaces.
So that’s also going to be a very exciting venue to reuse all the things we’ve learnt on The Morning Line, and hopefully generate more works of this sort.
Is your Max patch talking to any sequencing or commercial software packages?
Oli: No. We do all the sequencing within Max.
Are you using any type of hardware controllers?
Oli: Well, there is an interactive system built-in to The Morning Line as well, which is based on a Make controller (a microcontroller similar to Arduino). For some of the pieces the artists wanted to detect movement around the sculpture, essentially, to detect when people were near to speakers, and we needed to do that in a way that was weatherproof.
So we developed a system based on the ultrasonic sensors that you get in car-parking-detection systems. We read the digital signal that comes from the sensor units, process that in the Make Controller, and send the data over Ethernet to Max.
That allows us to have the computer quite far away from the sculpture. We thought about doing it with video, but this was the best solution, really. Because they are mass-produced for the auto industry, these sensors are really inexpensive, and replaceable.
Tony: Also robust. I mean, we rely on the auto industry to do all of our testing for us!
Oli: As well as the interactive system, the Morning Line sound system is actually designed to be controllable remotely via OSC messages. For instance, in some performances artists spatialize live audio input from inside the sculpture using their own computers or hardware controllers.
Have you ever wanted to throw out your patch and start from scratch?
Tony: Well, we’ve never completely done that. We were very careful to think about our software-engineering approach when we designed the patch, because we knew it was going to be very complicated.
Having said that, there were several aspects that we had to revisit, so that the patch we made for Seville was something where we were experimenting and trying out a lot of these ideas. We then developed it, and started to refine it.
Now it’s pretty much the same as it was when we worked in Istanbul, but we had to address issues of timing, phase coherence and things like that, particularly with poly~, before the Istanbul installation.
We have worked on it for a couple of years, and evolved it rather than throwing it away completely. Quite a lot of it has changed in that period of time… gradually.