Currently, the sound archive is made open to the public in two ways: selected recordings are available on SoundCloudLink opens in a new window, and a very carefully curated subset has been made available as part of the book chapter “Stirrup Notes: Fragments on Listening” in the book “Writing the Field Recording: Sound, Word, EnvironmentLink opens in a new window.”

The book chapter contains essay fragments and poems that integrate listening to sounds into their compositional process. The associated sounds are available as audio recordings online through the publisher's websiteLink opens in a new window (see links for Chapter 11 under the Resources tab) and referenced in the appropriate place within the chapter. One section of the chapter is based more specifically on a collection of sound recordings created as part of a workshop that now forms a subset of the overall dataset. This collection consists of 40 objects (sound recordings on cassette tapes) and 40 poems, each derived from one of the sound objects. For each recording, a specific type of spectrogram was created and used as a model for the form of the associated poem. Some of the spectrograms and poems are included in the chapter together with the reference to the associated audio recording. At the end of the chapter, all recordings referenced are listed together with associated metadata.

A wider range of recordings are available on SoundCloud, sometimes organized in albums. Most of these recordings feature metadata, often using metadata suggested by SoundCloud. The recordings on SoundCloud are standalone and not connected to other types of data (e.g., poems, essays). Making the data available through SoundCloud is motivated primarily by the ease of sharing of the recording via a link. Moreover, it makes the recordings available to the public, so they have the potential to serve as inspiration for artists (e.g., sound samples used by musicians) and to inspire us to listen to our surroundings – especially sounds that one has learned to tune out – and thus, creating a renewed awareness of our environment through its associated sounds.

As a data archive that has grown over a long time with varying motivations behind the original data collection, there is no consistency in the metadata. While, for example, the metadata chosen for the book chapter is internally consistent, it is not consistent with the metadata available on SoundCloud, nor with the metadata encoded in the audio files themselves or in text documents present within the private archive. Moreover, there is no standardization for cataloguing, referencing, or archiving audio files. This leads to open questions around what metadata is desired, especially in terms of technical data regarding method of recording, encoding, and decoding of the sound and how to archive the data.

Various methods of recording and various audio formats have been used to create the dataset. Some audio files are no longer accessible due to lack of software to decode them. This suggests that a common file format is best suited to keeping audio data accessible. Depending on the purpose, one can choose different audio file formats. A compressed format such as mp3 is easy to share but compromised the fidelity of the sound, whereas high fidelity sound recordings are difficult to store and share (large files). Using a compressed audio file format inevitably means removing information. However, in many contexts, this can be considered negligible due to limitations of the devices used for listening.

In terms of online sharing of audio data, there is a particular challenge due to the dominance of sharing visual data on social media (pictures, videos). On many platforms, it is not as easily possible to share sound and no “Instagram for sounds” exists. This raises a question raised about the potential of creating a more open online platform for sharing audio files than SoundCloud.

To address some of the above challenges of opening up the Ecopoetics archive, the Curating Data Cultures project proposed to replace the traditional publication format optimized for print - the book chapter - with a website. One of the key features of Ecopoetics is the intimate connection between sound and the written word. Publishing texts and poems associated with specific audio recordings in traditional print media such as journals or books has the obvious limitation that audio cannot be embedded within the printed text. Hence, the audio recordings have to be provided separately and can only be referenced within the text, leaving it up to the reader to make the necessary arrangements to experience the connection. On a website, a more immersive reading experience can be created by embedding the audio files directly within the text. To show how this could work, we used the University of Warwick’s content management system called SiteBuilder to embed some audio recordings directly within the text of this webpage.

Taking advantage of the fact that much of the data archive is already publicly available on SoundCloud, we can use the embed option from the sharing options on SoundCloud. This generates a short code snippet that can be inserted into the html code of a website and creates a player for streaming the desired track on SoundCloud.

By pressing the “play” button, the reader can now listen to the track while continuing to read. Careful placement of the audio tracks within the text will thus create an immersive reading experience. SiteBuilder also allows us to upload an mp3 file directly to the University website and embed this into the webpage with a very minimalistic audio player. This option may be more suitable for avoiding distractions from the immediate content of the page.

We have deliberately chosen a solution that can be easily implemented with very little specialist technical knowledge. By relying on SiteBuilder and the embed-code automatically provided by SoundCloud, no web-development expertise going beyond a basic understanding of html is required. However, this convenience comes with some notable drawbacks. While it does enable an immersive reading experience, it provides little opportunity to explore the underlying data set itself. For example, there is no automated way to create a list of the audio tracks that have been embedded on the webpage, although this could be added manually for a smaller dataset. Moreover, there is no search or browse functionality to traverse the data archive.

An alternative would be to organize the data archive as a relational database and to design a purpose-built website that showcases the archive. With this approach, a webpage that creates an immersive reading experience can be complemented with pages for browsing and searching the sound archive as well as views of individual items with their associated metadata.

Example item view: SoundCloud track

This example item view of a SoundCloud track was created with Omeka S, a web-publishing platform that was designed for easy creation of digital collections and online exhibits.

Example item view: poem

This item view of a poem was also created with Omeka S and belongs to the same overall dataset as the SoundCloud track.

Within the metadata, the item views link to each other as well as back to a webpage that they are included on. A visitor to the website can therefore freely explore the connection between the individual items as well as enjoying an immersive reading experience on the webpages that the items are included on.

By opening up the dataset of sound recordings via a website, digital formats come to structure the experience of reading and listening. A key question is: how does this organise experience? What are the consequences of this for the interpretative capacity of audiences, and for the integrity of the set?

One of the projects that informed the creation of the Ecopoetics sound archive and the "Stirrup Notes" was a box - an actual box - that was created for a literary magazine called Elevator that is edited in Buffalo, NY. For this project, an artist made several of these boxes and then the contributors were tasked with going out and finding 40 objects. Authors were then asked to write 40 poems in response to these 40 objects.

Opening the box, we find different material objects - such as a compass, a piece of metal, or an old sort of artillery thing, a piece of moss, a part of shoe brush - and envelopes with the poems. However, for the Ecopoetics version of this box, Skinner recorded 40 sounds which were put on a cassette tape. Each of the 40 recordings has a spectrographs of print, and the box contains the forty associated poems are printed on cards. They are called directions for the ear.

The box structures a material, sensory process of "opening up" the data set - if it can be called that - which engages the senses in ways that the digital format might seem to flatten out. That is, there may be a certain forgetting implied in the equation of opening up with digitisation, a forgetting of what we could call the integrity of sensory-material experience.

Similarly, we can ask what modes of participation are afforded by different media. The Ecopoetics project has been presented to many different audiences - to birdwatchers, children, and people who are not usually readers of poetry. For these participants, hearing the sounds provides a way into the poems: sounds open up a plane of experience - of immersive interpretation, if you will - in ways that written text can't do, and "data" or a computer, possibly not either.

On the other hand, it is important to note that most of the 40 poems - directions for the ear - in the box have never been published. And one reason was the challenge of how to retain the relation between the poems and the sounds in a published format.

The kind of database proposed by the Curating Data Cultures project could be a way to begin addressing this, besides the translation into the technical that you propose, which is interesting to experience in itself.