Daz's unofficial guide to the SoundWall at The Collection, Lincoln.

INTRODUCTION

Built in 2004/5, The Collection is Lincolnshire's flagship Museum and Art Gallery - amongst other things, home to an extensive clock collection, some Grayson Perry vases, a Plesiosaur and plenty of Roman bits 'n' bobs - for more info on that side of things, including the Usher Gallery and the various shows they now produce, please visit their site.

But for the moment, forget all that stuff as this page is dedicated to the Museum's rather remarkable sound system, known simply as "The SoundWall" :


The system is difficult to capture in a single image due to the shape and size of the room - the wall is around 33 metres long (100 feet or so) and embedded behind the Ancaster stone facade are 22 speakers arranged in three horizontal rows.

The design and installation of the wall was a collaboration between Lincoln City and Lincolnshire County Councils, Arts Council England, Architects Panter Hudspith, Sound Artist Susan Hiller and systems designer Cantoris Sound. Additionally, artist Adinda van 't Klooster was commissioned to create an initial piece for the wall.

So a quick declaration of interest : I was a consultant on audio-hardware around playback and control during design and commissioning and have since become the unofficial house-engineer working with artists to execute additional pieces on the system. As there's no official web-presence for the SoundWall, I figure I may as well share a bit of what I know and have worked-out in the process of facilitating new works.

What I gather from others working on the original project is that the overall concept was to use a literal wall of sound linking voices and history, people and archeology to enhance the visitor experience, reconnecting individuals to their scale within history and helping them engage with the artefaces in the permanent gallery (which occupies the space behind three of the four doorways shown above). An embedded piece of sound art as a permanent system in a public building with a design-life of 120 years.

Each morning, the building would fill with the sound of voices - some recorded locally, some from archive - telling a variety of local stories. As the day progresses the voices slowly begin to be processed with archeological data and by mid-day the processing reaches its peak, before slowly subsiding towards the end of visitor hours.

What all-but two or three people didn't realise at the time of the project was the scope of system required to execute the concept, and what it would mean in terms of future audio possibilities. They had effectively created a new and unique audio-playback format.

Sounds great, right ?

At the time of the build, as far as I understand it (and I could be wrong), various funding policies meant that from a producer's point of view the initial system would, despite its size, be a little ... erm ... minimal. The concept required 22 speakers, about a mile of cable, loads of amplifier channels, rock-steady automated multichannel playback and an audio piece to play. However, from the political point of view, if the system had a microphone input it would be classed as a public address system and if it had an editing rig it would be classed as a studio and both scenarios would have disqualified the system from obtaining funding. Adinda's piece was created in the Netherlands and didn't so much require mixing, being composed of algorithmically processed loops, so installing any form of editing or recording system was simply not on the agenda as it was judged to be redundant in the context of delivering the initial concept. So when the system originally went into commission it was effectively a brand-new format, but with only one piece to play - like having an incredibly posh cd player ... with only one cd ... and no convenient means of making another one for it.

Initially, Adinda's piece played for a couple of years before the system began to be opened-up to works from other Artists, the first of whom was Marcus Coates in 2007. Still at this stage a purely playback system - the rig was chugging along in the background; waking itself in the morning, playing through the day, sleeping at night, and proving itself to be one of the most reliable bits of hardware in the building. It was at this stage that the system's lack of mixing facilities became apparent as I had to shift our entire setup into the basement to connect to the hardware driving the wall. This worked fine except the speakers were upstairs and we were downstairs, so after each additional edit or tweak we had to go back upstairs and check the sound on the actual system. Slow. Tiring.

This situation persisted for the next two productions, and was finally remedied in 2010 when some of the Museum's end-of-year finance was made available to the SoundWall. Now the rig sports a powerful edit/mix computer enabling pieces to be created in-situ on an industry-standard platform : no more dashing up and down a long flight of stairs just to check a fader setting as editing now takes place in the room, with the speakers, although the process from artist's files to finished piece still takes about a week - see below for why ...


ARTISTS


Amie Slavin (photo: Sunshine Gray), (montage: Jo Grisaffi), Simo Alitalo

The following artists have produced pieces for the SoundWall

HARDWARE
(those of a squeamish disposition : look-away now)

This image is a composite scanned from photocopies of an early design document. The three layers of speakers have been coloured red (layer A, 6 large-drivers) around 15-18 feet-up, green (layer B, 9 medium drivers) at adult head-height and blue (layer C, 7 medium drivers) at child head-height. The control connection to the hardware in the basement emerges through the floor at point N and can be connected to a wireless-box or other network devices on an ad-hoc basis.

Initial configuration : Recent Additions :

PRODUCTION



Day-to-day playback is handled by the harddisk recorder and iDR system as this uses less power and has fewer moving parts than the workstation, but for creating new pieces, we rely on the full system - PC harddisk and iDR all working together to route the audio and virtualise the playback space.

As you might imagine, productions tend to be a little out-of-the-ordinary - the first question is which speakers to use - the limit of 16 discreet amplifier channels determines the maximum number of speaker destinations, although through the speakon patchbay, all 22 speakers can be targetted - (in which case, sets of two speakers each share amplifier channels) ... and from there its a small matter of navigating the billions of possible routing configurations between sounds 'on-tape' and where and how they appear in the space.

There are effectively 3 matrix layers between raw sounds and speakers - the workstation capable of routing a couple hundred audio tracks to any 16 lines, a further matrix mixer capable of sending any 16 inputs to any combination of 16 outputs, and then the speakon patchbay allowing the 16 amplifier outputs to be routed to the 22 speakers.

In order to leverage all the dsp power of the system, a laptop is used in the main room, which acts as a remote control for connecting to the workstation downstairs - everything on-screen in the basement is visible upstairs and the local user's mouse controls the software as if it was running on the local machine. This happens via VNC - Virtual Network Computing - and when coupled with wifi allows a smartphone (such as the HTC android pictured, or any compatible portable device, ipad, etc) to be used as a mix tool anywhere in the space from the wall right out to the front-door, and museum shop (handy for checking annoyance factor!). Whilst the workstation interface isn't optimised for touch-screen control and detailed editing using a phone-screen is out of the question, the setup performs sufficiently smoothly to be more than adequate when it comes to mixing, EQ'ing, keeping an eye on level meters and invisibly triggering playback as an audience-member at events.

To further augment the setup and provide some hands-on tactile control, MIDI data from virtual mixers and controllers can be routed accross the network and into the workstation to provide some physical knobs in the space. I also sometimes use a midi controller I've built from a graphics tablet to position (and therefore move) sounds accross several speakers at once.


ISSUES



Naturally, producing material for any system on this scale is a little complication, and on top of the added mental-load of working with such a huge number of options comes the very real impact of physics on the system in its space. The speakers are all flat-panels (the only design that would fit in the 10cm gap between facade and fabric) and each one sits in its own chamber. As each chamber is a slightly different size, each speaker has a different sonic profile (as resonant frequency of the chambers depends on their volume). To further compound this, the louvres behind which the speakers sit are also slightly different sizes and depths, and have parallel sides - which results in further colouration. Add to this the fact that due to the thin profile of the speakers, bass response is restricted (because physical movement is limited). Another feature of this type of speaker is the dispersal angle - normally a speaker throws sound in a roughly 60 degree cone from the centre of the driver - as the flat panels work on electrostatic principles their dispersal is through a much wider angle - something close to 170 degrees. This coupled with the louvres makes for an interesting situation !

If you think that complex, we've still not factored-in the room itself ... which is where things get really fun. Shaped like a narrow and very wide cone, every surface in the space is hard - stone, oak, glass, metal - all reflective surfaces that contribute to a reverb-time of between 1.5 and 2 seconds. I say between these times as the space is large enough for more subtle environmental effects to make themselves apparent ...

Okay, a quick primer : Sound travels in air at *roughly* 330 m/s at regular temperature and pressure (21deg C, sea level), and whilst this approximation is close-enough for most classroom examples, if we dig just a little deeper, we find that humidity has an effect too - the denser the medium, the faster sound travels - water, for example transmits sound approximately 5x faster than dry air - so adding water to the air in the form of humidity has an effect on how far the sound can travel before it is fully absorbed by the surfaces objects and air in the space ... and given the hard materials and their tendency to reflect more than they absorb, a small difference in humidity can have a tangible knock-on to the distance the sound travels ... in other words, on a warm dry day, the room's reverb time is shorter than on a cold wet day. Add to this the thermal mass of the porous stone and things become even more compounded - as the space warms (due to sun streaming through the glass roof) the stone releases some of its stored moisture into the air, which has a subsequenct effect on the character of the room. Whilst these effects are too subtle to notice from minute to minute, they are large enough to have a noticable effect during mixing and playback.

I first discovered this particular quirk mixing with Marcus Coates - we'd got his piece sounding lovely - smooth creamy violins and violas filled the space with their lush sound in the warmth of the afternoon, but by the next morning after rain and in the cool damp air - the string parts sounded harsher as they reverberated for longer in the space - subtle, sure, but certainly large enough to have a tangible effect on the mix as we had to adjust to a reasonable best-guess setting in an attempt to have the piece play acceptably accross a range of ambient conditions.

For more background on this acoustic situation, see this 1967 NASA document researching the effects of humidity and pressure on the speed of sound in air (pdf) - possibly in reference to super-sonic rocket exhausts ... or something ...


REMEDIES AND OTHER TOMFOOLERY



Fortunately there is a partial remedy for some of these quirks of the system and its space. Obviously we can't change the rules of physics, but we do have a generous helping of DSP to play with - the iDR unit operates as a post-mix processor (in much the same way as a graphic EQ in a domestic system) and features a fully parametric EQ (and compressor) on each input and another EQ (and limiter) on each output. Using this we can at least lessen some of the effects of the speaker-chamber-resonances by damping some of the problematic frequencies that crop-up. There is no solution to the moving-target of room sound however, so pieces have to be made to a best all-round fit in order to attain as consistent a response as possible. The system as a whole has never been calibrated, and to a large extent, it would be meaningless to even try, for as soon as everything is under control, the conditions in the space change and the issues move to a different part of the spectrum on a different part of the wall - not to mention that each piece uses different combinations of speakers (and of course frequencies) and therefore excites different parts of the wall at different times during playback ... did I mention that some of the stonework is smooth, whilst other parts are rough-cut ?

On-top of the DSP power for several projects I've developed some custom audio software for the SoundWall - multichannel linked compressors, field manipulation tools as well as panning / positioning systems. This occurs on a per-project basis, and depends entirely on what the artist wants to achieve and what the system's perceived capability gaps are at the time of production - with Simo Alitalo's piece, for example, we created a swirling wind effect by processing his stereo recordings into dual false-stereo and then placing the seperate elements of this at opposite ends (and heights) of the wall - the net effect of which is that when the wind blows, it moves around the space creating a vortex-like buffetting experience with different elements of the gusting appearing from different places in the room.


IN CLOSING

So there we have it - one of Europe's largest permanently installed audio difussion system, and certainly a unique instrument. All this text is however meaningless unless you go and see/hear/feel/experience the SoundWall for yourself - it is currently showing Simo Alitalo's piece as part of the Charter of the Forest exhibition. He has mixed original recordings of a local woodland into a glorious ambient environmental piece featuring bird-song, nature-sounds and a very lifelike light-aircraft flying through the space - it'll be playing all summer so if you get the chance do go check it out : it plays at ten past every hour from 10am through to 3:30pm and lasts about 20 minutes - why not take a small picnic and take time to chill-out in the virtual woodland as the summer sun streams in accompanied by the sounds of blackbirds singing, woodpeckers pecking and frogs ... mating.

If you're an artist wanting to join the growing list of those who've created work for this unique system, please either get in-touch via info on my main page, or by emailing the museum's director Jeremy Webster

I'm sure there's probably meant to be various funder logos in this space - will update page as and when available ...

This page first created, 29th June 2011 (all photos © Daz Disley, except where otherwise credited)