Logistics

The exhibition is nearly over: five days to go. So, mentally, I’m winding down now, and there are rather fewer things to think about. Which means I’ve a little time to reflect on the process, and what I’ve discovered and, hopefully, learned.

So, this page will probably be tedious for most people (and I bet you it turns out to be long, too, which will make it worse) as I’m using it to lay out some of the logistical and practical elements of the work. Why? For my own sake, so I’ve a reference point for the next project (I’m determined there will be one – probably several at the same time, knowing me) and in the hope that anyone going along a similar road can make use of the roadsigns I’ve knocked up.

Collaborative work

I’ve worked collaboratively in lots of academic contexts, but never artistically. Much of “Exploding Poetry” and in particular “Speakers” was a collaborative enterprise. I think there are two areas of learning: managing the thing, and balancing the creative inputs.

In the case of “Speakers” most of the issues have been logistical: how do you get 100 people to contribute poems and to record them, how do you combine them into an workable whole, how do you deliver several audio tracks through multiple channels without just creating noise, how do you construct a piece of work from other people’s work? Lots of very interesting questions, which required skills in project management and lots of compromises in the practical side of things.

Originally my idea was for eight channels, each with around twenty to thirty texts, of which I’d provide about half, delivered through randomised “shuffling” on mp3 players (4 players using the two stereo channels as separable mono channels). This became, in the current version, “Speakers”: four audio-visual channels, on DVD players,  delivered through TVs.

How did this happen? Partly it was purely pragmatic. What can you afford? What do you have space for? How many texts can you write or obtain? How long have you got?

As each of these questions received clearer and clearer answers, changes were made. Sometimes they were compromises, sometimes new opportunities came along.

But the practicalities were also interrelated with the design and content decisions. For example, in the October workshop I had three channels running from MP3s and computers (actually, it was four for a while, but one of the MP3s decided to misbehave). They ran the random collection of recordings I’d assembled at that time. The effect was occasionally quite interesting, but often disconcerting: long periods of total silence suddenly interrupted by two or more voices competing with each other, without any sort of preamble or hint that someone was going to speak. This startled a couple of visitors, and also me, even though I knew it was there.

Now, in a different artistic context, this could be interesting: the reactions of people to apparently arbitrary and random sounds. I can see how that might be a creative idea. But it wasn’t what I wanted. I wanted people to move between texts and listen to them at will, moving around. Moreover, because everything was random, I had to create playlists which included recorded silences, of various intervals, so that the machines were actually playing randomised recorded silences. Whilst I’m sure John Cage would have approved of this, and I could claim a very avant garde experience, in fact, it was no experience at all. It might’ve had a point if I’d had a point in doing it, but I’d only done it in order to separate the spoken tracks – and this involved a fair degree of working out: how many tracks of what length of silence would make for decent random intervals, but which were typically not too long. And this was complicated by the fact that you could not use identical filenames, so each silence had to have a different name: so I had a range of tracks called things like “15 seconds silence v2”, which gives you a really weird playlist.

This was also run in a comparatively small room, where the distance between speakers was quite small, and therefore issues of relative volume and relative speech could become critical.

Out of this, and a brief conversation with Michael (who at that time I hardly knew) it seemed to me obvious that the practical setup (1) could not be purely random, but needed a little more planning; (2) could include silence, but really most of the time needed some audio “links” between texts, so that audiences were primed to hear something; (3) either needed to be in a larger room or could not be the planned eight speakers. I reduced the idea to six at this point.

I also then looked at the possibility of using a computer for control with multiple audio outputs. Now, this is perfectly possible, given four things:

– the right hardware

– the right software

– the right technical expertise

– the money to pay for the other three.

It turns out, for example, that sound surround boards can give you eight tracks, but controlling those tracks independently is a very complex technical exercise (beyond my competence). Alternatively, you can go down the MIDI route, but this means connecting your computer to several MIDI devices, such as sampling keyboards, and suddenly the cost goes through the roof. My budget was very limited, because of my original plan.

Another constraint was the sheer logistical burden of gathering the texts. Six channels of thirty texts means 180 texts. If I wrote 60, which is a lot of poems to write, especially if you want them to be good poems (I’d say it’s about a year’s work, for some poets rather more), then I’d need 120 from other people, and I had collected about 30 at this point. It did not seem likely I’d hit this ambitious target in a decent timescale.

Moreover, every text had an organisational overhead. Essentially the process went:

1) think of a way to invite contributions (e.g. discover a new writers’ website)

2) invite contributions, probably by negotiating by the owner of the website. This would also involve creating or adapting messages about the project, the idea, what would happen to the work, copyright and so on.

3) get three or four responses, some of which would be actual poems, but some of which would be enquiries for more info.

4) check the poems. They’d be of varying quality, in different timescales, perhaps one recorded already, probably not. One or two might benefit from editing, so there might have to be a subtle suggestion or two made. One or two might be too long to use in full, so negotiation would be needed to allow for only a part of it.

5) Find someone to record it. I wanted only women’s voices, because of the nature of the project. (That in itself is an issue, of course, if a man starts sending out messages asking if women will record themselves for him. The nature of the message needs careful thought, and also who it is addressed to. I decided I could only use people who knew me at least moderately well, and so would be reasonably sure I wasn’t doing anything reprehensible, or else paid actors: but, again, my budget for this was limited.) Moreover, I’d discovered in the workshop that it was disconcerting to hear the same voice from separate speakers, and I didn’t want any given voice dominating the piece: it was supposed to be about “woman’s” voice, so I wanted maximum variety, accent, age, cultural background etc. I also didn’t want to ask too much of people.

So although some readers were wonderful enough to record more than three pieces, essentially I settled on asking people to record no more than three pieces. So, for 180 pieces overall I’d need at least 60 readers.

The scale of the project was just too large, given my budget, time and, indeed, energy. To be effective in making this work, I would either have needed an assistant to do lots of administrative work, or to stretch the project by another 6 to 12 months.

So, I reduced it again. I would aim for only around 100 texts (privately I was hoping for 120, including my own) which would be 20 texts on six speakers.

6) Organise, advise on, administer, facilitate the various recordings. Some people were experienced. Some had no idea. Some had technology available. Many didn’t. Some needed instructions. Some needed loan of equipment. Some wanted to come into studios. Every one of the readers needed some sort of specific support to enable them to record: light touch or intense.

7) I knew that some recordings would be poor quality, and I wanted to make sure each text had a good chance of a good recording. So I had to arrange a schedule whereby each text was recorded at least twice, by different voices, whilst ensuring that no-one read more than three things, and that all texts were recorded. Meanwhile other texts were arriving. So I had various spreadsheets recording who had written what, whether it was recorded, what the quality was like, and also where it might fit in the overall scheme of things.

8 ) Meanwhile, try and write my own stuff.

9) Meanwhile keep people informed of what was happening, if anything was happening, and try to think of other ways to attract texts. I used contacts, websites, friends, groups I belong to. I asked on person to pursue her own lines of enquiry on my behalf. I organised a small competition with a prize for the best contribution on the theme. I kept my eye open for published work that might fit the bill (but did not dare approach the big names, such as Carol Ann Duffy, who has pieces that could fit the project quite well). I did invite some published poets whose work I rated, and a few responded positively. Mainly, though, amongst the well-published they either didn’t respond or made a promise they didn’t keep. (However, this is true for all kinds of contacts, I found.  I would say that about a fifth to a third of people who make a promise to contribute will not actually do so. The number gets better if you remind them occasionally of their promise but, of course, this is more logistical work).

So I’ve engaged in hundreds of communications, mainly emails, to make this project work, kept up around a dozen spreadsheets, organised various hardware and software.

Now, once you’ve set such a ball rolling, you can’t stop it. I had a further offer of a possible contribution just last week, and the day before yesterday someone offered. So there’s continual communication even when you think you’ve finished.

The need for an audio link between texts and the technical problems of delivering multitrack audio, coupled with the realisation that I needed some coordination between the different channels if the impact was to be effective, convinced me that I needed an expert. Fortunately, I had thought ahead well enough to realise there’s be a major need like this, so I was able to bring Michael in, as composer, audio artist and, to some extent, technical guru.

Soon after this Bank St began firming up the exhibition programme, so we had to decide how we were to use the available gallery spaces. Again, the logistics of what was possible in the different spaces had an impact. I’d been thinking we would use the largest space for the audio work, to maximise distance between speakers. but this room is long and thin, so speakers would not be equidistant, and it was also the gallery used as thoroughfare to the other spaces, so a bad room for audio work, which would be constantly interrupted.

Which meant a smaller room. Therefore fewer speakers. Which meant that the pressure to generate massive numbers of texts went away, just as the outcomes of the competition delivered 20 more poems! So now, in fact, I had the opposite problem: too many texts for the time we’d planned on. How would we get them all in? Or did I need to institute a more draconian editing process?

From the start I’d wanted to be as inclusive as possible, so my plan was to include everyone who contributed at least a reasonable piece of work although, of course, I wanted the highest quality I could find. Fortunately even the simplest pieces worked quite well, and some of my readers had extracted from the works quite interesting  versions of even relatively straightforward material. And then, of course, there was the intrinsic interest of the subject matter in many cases: even where the text might be a little on the weaker side, the content is often superb, painful, revelatory, fun, surprising. So there were few of the pieces I’d been given that I was happy to lose.

I wanted to incorporate Vaughan Williams Fantasia on a theme by Thomas Tallis in some way, because of its spirit and some personal associations. We thought about using a phrase, but were worried about the issues of sampling other people’s work (and, in fact, I think the original piece is still protected by copyright, too). But Michael had the brilliant idea of using the shape of the Vaughan Williams piece, its waveform, as the shape of ours. This made our piece 15 minutes long.

Michael produced the first version of the piece, which pretty much astounded me. His audio enhancements,  the cadences derived from the Vaughan Williams piece and, most of all, the interacting effects of the texts themselves seemed to me something I couldn’t have envisaged or imagined – almost as if the work was making itself. Which is an experience you might be familiar with, if you’re an artist or a writer, where the work somehow seems to have a life of its own, working through the author, rather than belonging to her or him.

Along the way, though, there was another logistic issue: keeping track of what texts, and which readers, went into each channel. We decided we wanted to keep all the texts by the same reader on a single channel, and yet each channel was also to have its own structure; and the four channels were to echo and relate to one another, as a composed piece, yet we knew that the four channels would be running semi-idependently because they would be started by invigilators in the gallery in their own times with uneven intervals between them. So we needed to keep records of each piece, where it fitted, which recording of it was used (where there were multiple recordings) who was on each channel and so on. Moreover, some pieces were recorded very late indeed (the last piece was recorded after two of the channels had been completed) so the actual conception of the piece was in flux as well.

So to get to this version I had to make some hard decisions: rank the poems by my view of quality and value in the overall piece. This included my own work. We also assessed the recording quality and reluctantly decided to drop some good pieces badly recorded. A few we decided to use in a “treated” way, so that they were deliberately to sound as if scratchy or heard at a distance.

But there was a further problem. When we heard the four channels in the actual space, it worked, and, at times, it worked superbly – but those times were few and far between. Instead there were too many instances where the texts simple clashed with each other. Occasional moments of bedlam were fine – in fact, desirable – but not half the piece. We needed to use fewer texts. But we’d been through a hard process of decision-making, and didn’t wnat to have to do it again (and time was beginning to look short, as well).

So, instead, we made a different decision. We decided to stretch the piece from 15 to 30 minutes. It would still keep the Vaughan Williams structure, but elongate it. This way we would double the amount of silence in the whole piece, meaning that there would be only half the instances of conflicts between speakers. (Even this, it seems, is too much for some people, judging by some of the feedback I’ve had – although I think it is about right – you need to work hard at some moments, and others are relaxed and gentle).

This gave us the final working piece. Except for one thing.

We could have delivered it using CDs and CD players, or even the MP3 players I’d first imagined. But this would’ve meant a room that contained nothing other than four loudspeakers. Imagining people wandering into a room with nothing to look at, and staying there for 30 minutes, seemed pretty unlikely. Moreover, we didn’t have four identical loudspeakers. What Bank St did have was four TV monitors. So we’d decided on DVDs, simply because we had the technology. But this meant we needed images on the TVs. And not merely images, but images which did not distract from or interfere with the audio. And we would need either four lots of 30 minutes animation (ie. a feature film length of animation) or the equivalent in stills. We could, of course, place the same sequence on all four monitors, meaning we would only need 30 minutes, but this seemed both to miss an opportunity and to be a little defeatist.

So, roughly a month from the Private View, I began the task of creating slide shows of stills to cover 2 hours of viewing. Now, the mathematicians amongst you will see quickly that the key factor is how long you show an image for. If, for example, it’s on screen for fifteen seconds (which is quite a long time) then for a thirty minute sequence you need 120 images, meaning I’d need to create 480 in four weeks, or 12o a week, or about 18 a day.

In the end I produced a little over 200. I used all sorts of mechanisms to make the images easy to produce, but I was also conscious that I wanted images which would do a job, not merely fill the screen. I realised I could use the same images on several screens, if I placed them in different orders; and that I could start from a given image and alter it to create a new one, which, when placed in sequence, led to gradual changes of appearance. I buils some images to be about woman and war, some to contain text presented in intriguing ways, designed to draw a viewer towards a particular monitor in order to decode it, and text which simply had a connotation or association which seemed to fit, a few of which were personal (such as a letter written by my Dad en route to his station in Burma during WW2) and some of which simply seemed expressive.

This created another logistical task. With so many images, but not enough to place a different image in every “slot” in the system, you need to keep a record of which ones are used for which channels, and, to some extent, how they fit together. Primarily this was another practical issue: trying to ensure the maximum variety and impact across four tracks, using broadly random organisation. But also there was an aesthetic element, as I tried in part to build sequences which either progressed, or reflected the audio.

And this in turn has aroused other ideas. I don’t think many of the images are created stand alone as having particular quality. A few I’m pleased with, and some are good ideas, but a little rough in delivery. However, I can see how a series of images and words, properly thought about, arranged in parallel with speech, properly thought about, can create a poem (or at least “artefact”) which communicates through several channels at the same time. I’d think of it as a poem, but it might equally be called a film, animation or composition, as the different media are intended to work together. Now that I understand the tools and considerations rather better, I’m pretty sure there is a work here for me to produce.

Finally, we had the issue of whether to produce an independent piece or not. Once we knew how good the four track version was, we wanted to see if we could make something workable that most people could listen to, outside the gallery, if they wanted. I thought first of a package which contained all four tracks (two stereo systems), so people could reproduce the full experience at home. In principle it’s perfectly possible, as we simply mix two pairs of mono tracks into two stereo mixes.

But who has a home setup that allows for four speakers to be played at the same time? I can imagine a few people wanting to do it, but it’s not most people’s idea of home entertainment.

The next thought, then, was a DVD, to reproduce at least a core experience. The problem here is that there’d only be one image, not four, and this would probably be centre screen on a TV in the lounge as the audio played away. In other words, the image would have a huge role, probably bigger than the audio, and none of them had been designed with that in mind. As the whole point is the poetry, the voices, I didn’t want people to be distracted by second-rate images. If I’d the time to construct a more meaningful video, then possibly – but, obviously, there was pressure to get a product completed before the end of the exhibition.

But, in the case of “Speakers” I’d also not secured permission from everyone to use their material in a different context, so I had to engage in a further round of communication getting that, and handling any queries that arrived. And Michael couldn’t being te new piece until I could tell him we had permission to use the works he wanted to use (and once again there was another record keeping job, as we had to cut the number of pieces down by more than half for the new medium.)

And now I have the task of keeping everyone informed about the CD, and checking on who has had it, who wants it, who wants to buy it, and whether I have enough, or too many, for future purposes.