The HOTBED project has supplied the BA Scottish Music Course at the RSAMD with a unique and custom-built resource, designed to cater for specific pedagogic needs, focussed on the transmission of oral/aural rather than written knowledge. This new, web-based access to materials that were otherwise difficult or impossible to reach has alerted the staff and students across the entire institution to the possibilities afforded by new technologies in practice-based and multimedia-focussed areas.

While the culture change necessary to completely embed HOTBED in the curriculum is taking time, students and staff recognise the possibilities offered by the facility to gain access to a wealth of resources instantly from any internet-equipped PC and then store, manipulate and share these resources in a manner that is intuitive. This, in turn, has fuelled a wider enthusiasm for technology in learning across the Academy and HOTBED has become the focal point for VLE developments in the future.

The multimedia manipulation tools (MMTs) in HOTBED that allow marking and looping of sound items have sparked a great deal of enthusiasm amongst musicians as they provide precise playback control. In combination with the organisational, sharing and user profiling tools in HOTBED, the MMTs form part of a package that addresses both existing difficulties with materials (particularly access and format) and provides new tools that enhance the learning experience.

HOTBED has sparked interest beyond music and even outwith the arts. Presentations and demonstrations to academics from across the board have generated a great deal of discussion, resulting in plans to provide a similar user-centred, practice-based resource to the wider HE/FE community.

• A searchable database of over 1100 items with associated metadata
• A powerful search engine that provides targeted and free searching as well as a browse facility and a synonym search
• Delivery of text transcriptions and images as secondary materials
• User profiling tools to allow users to collect resources together and track their work
• Collaboration systems to allow users to share resources and readily communicate with one another
• Multimedia Manipulation Tools to provide the facility to markup a resource in realtime and loop user-defined sections of the resource
• Dual broad and narrow band delivery

Information on standards and guidelines on best practice were researched – specifically from the JISC – and adhered to wherever possible. In the design and planning phases it was agreed that the system itself should be upwardly scalable due to its modular approach, with each functional subsystem simply adding to a bedrock database management system. This modularity had positive ramifications for maintainability: the bedrock system (evolutionary prototype) took account of research already undertaken in this area, and utilised Open Source technologies in as many areas as possible, and has resulted in a robust system.

The technical research phase also focussed closely on interoperability, so that the evolutionary prototype took account of this issue from the outset. The creation of the HOTBED metadata schema was built on the Dublin Core guidelines, and, in addition, ongoing integration research throughout the project (e.g. into XML) as well as liaison with the AHDS have ensured scope for the porting of HOTBED materials to other systems. Furthermore, Open Source developments should ensure that Multimedia Manipulation Tools will be accessible to the HE/FE community at large.

HOTBED was developed on Linux and Mac OS X, as both provide ready tools for creating dynamic web content with Apache, PHP, and MySQL. The delivery server runs Mac OS X Server. This was chosen as it provides access to these Open Source technologies ‘out of the box’ as well as having proven reliability (as it is based on a Unix BSD core), and also because it includes Quicktime Streaming Server. Quicktime is recognised as having better audio quality than its competitors, and Apple does not charge for use per stream, so it was the most economic choice as well as the one that would ensure high audio quality. Use of Quicktime also allowed us to provide video easily and quickly when required to do so.

All the code for HOTBED was written by hand, using the text editors BBEdit and vi. In addition to the mainstay PHP and HTML code, Javascript was used to create the HOTBED looping and marking system. This was written from scratch also, utilising the Quicktime Java API with its Javascript support (which is browser dependent).

The original metadata schema was mapped onto a relational database structure. This proved to be a difficult process, as the Dublin Core is object oriented and MySQL is a relational database. However, each useful Dublin Core descriptor was given a corresponding entry in the HOTBED database, and dummy entities were created in order to make the database work. Therefore, although the structure of the HOTBED database could not be pure Dublin Core, an XML parser could be written that would relate the relational fields back to the original Dublin Core dataset.

Given the limited IT support in an institution the size of the RSAMD circumventing firewall issues and integrating a new server based system was problematic. Network problems were encountered but resolved over the life of the project.

As the materials in the HOTBED system were provided for accessibility and manipulation and not archival purposes, the digitisation processes were not concerned with preservation, but simply with the most effective manner of retrieving materials from a number of different sources, some in archaic formats. The initial bulk, and mainstay, of recorded materials came from the University of Edinburgh’s School of Scottish Studies sound archive. The highly specialised systems used for describing and storing the archived materials, while appropriate to the unique nature of its holdings and typical of sound/video archives, required a great deal of work in transferring the holding to a digital repository. Actually physically locating and playing a tape while transposing the information from a record on a paper register sheet into electronic format was a considerable challenge.

Many of the materials targeted for use in HOTBED were held on reel-to-reel tape. Some of these materials date back to the 1950s, and some work had to be done on pitch correction, as field machines used often had batteries dying towards the end of recordings, and were also not always properly calibrated. The chosen items were transferred from reel-to-reel to CD and brought back to the RSAMD, where they were edited if necessary, and converted to Quicktime files.

Other materials came from CD and 78 rpm record. As the CDs were native digital they were easily converted, but the materials on 78 were almost 100 years old, and also suffered pitch problems. Some editing work had to be undertaken on these to bring them in line with a musician who would want to play along with them at the correct pitch, but it was decided not to ‘clean’ them further.

Creation of metadata proved more resource-intensive than anticipated, partly due again to the specialised nature of the source archival records. For example, many piping items had quite detailed accompanying notes on the index sheets which had been filled in by the original fieldworker. There was no access to electronic records and the very faint photocopies were at times difficult to read. The material also proved very time consuming to put on the system. This was due to the fact that many items were lengthy interviews with informants. Each one had to be listened to in real time, divided into segments, at sensible points, which would then constitute individual HOTBED items, and have the metadata entered. It proved difficult to enter the metadata as many decisions had to be made about the nature of the item, a task which mainly fell to the Learning and Teaching Officer and tied up her 0.4 post for several months.

In order to facilitate the creation of a large body of materials – a ‘critical mass’ – in HOTBED, the methodology for item description had to change. Originally the goal had been to describe each item as thoroughly as possible, but it became evident that paring down the granularity of the description and working with a smaller subset of fields would allow for swifter transferring of materials from the archive to HOTBED.

Questionnaires were distributed to all involved, and were designed to provide information in the following areas:

• Usability of current music web resources
• Quality and appropriateness of content
• Which features were desirable from a performance study perspective
• Level of internet experience of the user group

HOTBED’s Learning and Teaching Officer undertook research to scope in detail the facilities required by the target user group, resulting in this ‘Wish List’ which reflects the specialised needs of this group.

Important factors were that any database had to be searchable by various different approaches e.g. tune type, title, key, language and that the ability to manipulate the sound recordings by dividing them into loopable phrases and slowing the loops down without altering the pitch would be incredibly useful. This was unanimously agreed by all staff members.

Many other factors were taken into account during the design and build process which were direct results of looking at the system from the users point of view. For example, the ability to create lists and send them to other users, the fact that the user profile would store recent searches and history lists and the emailing system were all to aid the users in their work. Other suggestions such as students being able to keep their practice logs or work diaires as part of their user profile or to submit essay work via the system did not in the end make it into the final design.

These training sessions also acted as a form of evaluation, inasmuch as users were asked to comment on the usability of the system and its appropriateness for learning music. While users highlighted some small usability issues, they were generally confident about finding their way around HOTBED and its facilities.

Interventions like HOTBED depend heavily on the co-operation of others for success and we were fortunate to have enthusiastic support and co-operation from members of staff. However, factors including long-term illness of key staff in a small department and the general staffing profile of that department (common for small specialist arts institutions) meant that full integration into the curriculum has still to be achieved.

The tutors then agreed to use these ballads in their teaching and to direct students to the HOTBED system rather than give them out taped copies. HOTBED provided easy access to a greater range of differing regional styles and different versions of the same song. However, it met with mixed reactions from the students. Initially it was felt that HOTBED was less accessible as they had to go to a computer to use the system rather than have a tape of their own which they could play whenever they wanted. Indeed as the students stated that part of the learning process was to listen to the tape over and over again, and that they did this while engaged in other more mundane activities such as washing the dishes or travelling by bus, the HOTBED team began to investigate the possibility of acquiring MP3 players which could be lent out to students who could then download the required items and treat them in the same way as they did the tapes. This raised various issues of costs, copyright and security and it was decided not to pursue it further.

Rather than use HOTBED for a broad span of time on a course, the fiddle tutor, in consultation with the Learning and Teaching Officer decided to focus on specific parts of the syllabus, namely the work of Niel Gow and J Scott Skinner. The tutor recorded himself playing 7 Niel Gow tunes in his own home studio and gave them to HOTBED on a CD. These were then uploaded onto to the system along with basic metadata (title, composer, tune type, key and time signatures, performer, date and place of recording).  The fiddle tutor owned a large collection of 78 rpm recordings made by J Scott Skinner in 1910 and these were included on HOTBED. It was also decided to digitise recordings made in the Shetland Islands collected by fieldworkers from the School of Scottish Studies. Several other versions of Gow tunes were later added the following year when an additional tutor joined the staff and gave HOTBED permission to use items from his commercial recordings.

Students were directed to these recordings on HOTBED and asked to learn material directly from them. However, while this worked well in individual trials, there were various technical problems that negated the success of the integration as a whole because of the tight scheduling of the relevant modules. Some of these were outside the control of the project as they related to the installation of a new firewall within the Academy, which had implications for computers accessing the HOTBED site at the period that the students were studying these modules.

The RSAMD does not have many teaching rooms with network access; this inhibits takeup and causes practical problems that eat into class time. Using the HOTBED clone on a standalone laptop was one solution to this, but there were also problems with currency of the clone. HOTBED proved much more successful as a resource to direct student toward for their own study. It has become commonplace for tutors to include HOTBED items on the ‘Listening Lists’ that are given out to the first year class. In this way all the students in that year can have equal and readily available access to the set pieces rather than all trying to check the same record or CD out of the library at the same time.

The evaluator highlighted some key points:

• HOTBED could be used both to learn a new style of playing a known tune and to learn a new tune
• HOTBED provides access to many tunes and allows students a large choice of materials
• Students said that they could not have learned a tune so easily from a CD as in HOTBED, as HOTBED provides a looping mechanism where CDs do not
• HOTBED can also improve transcription skills

The initial HOTBED plan was to implement networked sound resources and examine their use within the Scottish Music curriculum. However, it became clear that video would is also of great benefit in learning ‘aurally’, and the impact of video in HOTBED has been considerable, both in terms of learning use and demonstration of the system.

The HOTBED project was the first project of its kind to be undertaken by the RSAMD and had a major impact on attitudes to use of ICT in performance learning and teaching. The School of Drama at the RSAMD is keen to use HOTBED to develop a database of regional accents and dialects for use with students of acting. Course leaders of various jazz courses at an LTSN presentation also saw immediate possible applications for such a system in their own disciplines. Senior management at the RSAMD are looking at incorporating HOTBED in a larger VLE/MLE development and actively embracing the new technologies HOTBED offers, such as webcasting.

Although HOTBED itself is only available to staff and students at the RSAMD and School of Scottish studies, presentations and demonstrations have generated a great deal of wider interest. For example, at one presentation of HOTBED two academics, one involved in politics, the other in physiotherapy, asked if they could use HOTBED to teach their own disciplines.

The wider traditional music community has been very keen to gain access to HOTBED as it is at the moment. Every talk or presentation ends up with a large number of requests for access, and offers of payment for use or asking whether we have plans to produce it on any kind of CD rom package. This has come both from individual tutors and performers and from educational projects such as the Feisean movement.

• a ‘critical mass’ of many more recorded items is needed
• richer metadata would make the system more valuable to a wider group of users, and in more circumstances
• incorporating more non-sound items such as biographical articles, teacher ‘hand-outs’, links to useful websites and images would enhance student usage of the system and move further down the road towards HOTBED as a Virtual Learning Environment

RSAMD staff have also expressed a strong desire to incorporate students’ own work into the system, for example their fieldwork projects involving recording directly from informants. There is also the possibility of adding recordings of student performances and even recitals to the system.

As noted above, there is huge interest from the wider traditional music community in HOTBED, but the main obstacle is one of copyright. The project only has clearance for use by staff and students of either the RSAMD or the School of Scottish Studies. However with current developments in ‘Open Content’ this is an issue that may be resolved in the medium - if not short – term.

• Multimedia Manipulation Tools (MMTs): The looping and marking system in HOTBED has proven itself to be a useful and attractive tool across the entire sphere of practice-based, multimedia-focussed learning and teaching. Further developments in this area would include facilities to slow down the playback of resources, alter pitch, and extend markup facilities
• Video: The demand for video materials continues, and the provision of more of these materials is a priority
• Mobile Devices: Users have highlighted the problems associated with being tied to a PC and have expressed interest in the possibilities afforded by mobile devices – especially MP3 players – in this area. Given the current state of the art and the ubiquity of this technology, developments to extend HOTBED in this area would be entirely achievable

HOTBED DnA is a six month extension to HOTBED that will disseminate HOTBED to a wider audience and seek feedback from the wider HE/FE community. In addition it will focus on developing the MMTs further, for wider use across academic disciplines.

RSAMD March 2004