The Collection and Management of Assets in the Production of a Multimedia Program on Yeats
Adrian Mallon, 1995
Table of contents
The paper describes a database tool for the management of multimedia assets in a title-production environment. The tool was developed in the context of creating an interactive multimedia program on the life and work of the poet W.B. Yeats. The approaches and lessons described have general application in multimedia title-development environments and will be of interest to any involved in electronic publishing.
Developing multimedia titles necessarily involves the collection of information in a variety of formats: text, still and animated graphics, photographs, video and sound. If the scale of a project is large, managing such resources, or 'assets', can be problematic. This paper describes one approach to the cataloguing and management of assets which has general applicability to the multimedia title development process. As such, it should be of interest to any involved in electronic publishing.
The work referred to in this article was carried out as part of a larger multimedia guidelines research programme, as aspect of which is the production of a teaching resource on the life and works of the Irish poet, W.B. Yeats. The resource is being designed as an introductory program, suitable for use in second- and third-level English Literature courses. The production of the tool for cataloguing and managing resources was a by-product of the initial stages of this work. The research catalogue will be published at a later date in both paper and interactive versions and will be useful in the production of future multimedia, radio or television programmes relating to Yeats.
Because of the need to manage large numbers of multimedia files and the relevance of this topic to title production generally, a multimedia asset research and management database tool was developed.
Brief Description of the Multimedia Title-development Process
Because multimedia publishing is a young field and commercially sensitive, the published literature on multimedia production processes is not yet substantial. Some key published references are Bunzel (1994), Hoffos (1992), Howard (1992), Mallon (1994), Philips (1992). And, as in film and television, most multimedia title production operates in a 'seat-of-the-pants' fashion,—developers often responding to each project's requirements in an ad hoc manner, according to the development team's composition and experience. It is possible, however, to break down the production process into a sequenced structure of key tasks, even if many counter-examples will be found in practise.
As in film and television, the multimedia title-development process can be divided into three phases: Planning, Production and Post-production.
- Framing the brief
- Producing a treatment
- Researching assets
- Copyright clearance: approaches to archives or stock houses (specialists in the sale of audio, photo, video clips for general use)
- Scripting and Storyboarding
- Product specification
- Commissioning talent
- Collection and translating assets
- Originating assets (graphic, sound, video)
- Designing product packaging
- Writing support materials
- Program editing
- Prototype testing
- Prototype evaluation
- Mastering and reproduction
- Support materials reproduction
Phase 1: Asset Research and Program Design
In planning and designing the database, a user-needs analysis was conducted. The type of question addressed together with answers to the questions follow.
Who will use the database? The project research assistant, program designer, graphic designer, audio-visual designers/production engineers.
Who designs the database? The project manager/interactive designer and research assistant.
Who is responsible for maintaining the database? The research assistant.
How will the database be used?
- The research assistant will use it to record and catalogue assets identified as relevant to the program design/subject area.
- The research assistant will need to respond to requests from the interactive designer, graphic designer and other production designers for particular assets or asset types, or assets relevant to a particular human subject or topic or year or place.
- The research needs to be able to log information relating to the acquisition of an asset from a third-party archive or agent. This should include names and addresses and contact lines of an artefact holder and copyright owner, incl. cross refs. to separately-filed communications with that party.
- The database should be easy to use and contain an on-line help facility.
- The facility to generate printable, editable lists of assets grouped under user-selectable categories should be included.
- The catalogue should be readily translatable into print form.
- The catalogue should link assets which have been digitised, so that assets referenced are easily consultable.
- The catalogue should be publishable in printed and interactive versions, to facilitate future media background research on the subject, W.B. Yeats. The interactive versions should be consultable on a standalone desktop computer, i.e., distributable on disc, and accessible over the internet as a HTML WWW document.
- The catalogue should be sufficiently generic to speed the creation of similar multimedia asset catalogues in future projects.
- The cataloque should reference the source of an asset.
- The catalogue should employ a reference system which allows each item to be identified by a unique number and/or letter code, and by the type of asset (whether textual, graphical, photo, animation, sound or video).
Navigational and other features
The following navigational features were judged desirable and built into the finished product.
To enable the user to
- Browse items sequentially.
- Access graphic, motion video, and sound files
- View files in 16-bit colour
- Randomly access items via category headings.
- Randomly access items via category contents.
- Find text strings input by the user.
- Create new index screens.
- Delete existing index screens (confirm to delete)
- Group and sort items variously
Design for Managing Assets within the Production Environment
The program was designed to serve also in the production process, to record and keep track of the type of information that is of use to the interactive designer, content writer and programmer. This meant considering the nature of the problems encountered within production environments that the database tool may go someway to alleviating.
The Final Design
Hypercard was used as the authoring program to build the database. Hypercard was chosen because of familiarity and speed of working with it, and because colour was not to be an important feature of the interface. The current version of Hypercard, 2.2, has poor colour-handling capabilities but provides a rich, flexible and fast programming/authoring environment for the integration and manipulation of multimedia assets, for linking to external files and applications, and for seeking, sorting and otherwise handling text.
As a result of considering the needs of the development team, in both asset research and asset management, the following buttons and text fields were included in the resource. (See Figure 1.)
A field omitted from the original design but ultimately considered desirable was one for file resolution. For example, sound files were stored as 16-bit 22kHz AIFF files, many graphics as 16-bit, 240 pixels high x 480 pixels wide PICT images. This information was listed in the product specification.
The Database in Practice
The finished database tool performed well in practice. A number of fields were not utilised as we had anticipated they might be in planning the database structure, particularly the "contacts", "list", "files" and "archived" options. The fact that they were not used extensively in this project, however, does not preclude the need for them in some future project.
The "find" facility proved to be sufficent for most of the team's needs but could not handle looking for multiple entries. This would need to be corrected in any future version.
The electronic publishing industry is expanding exponentially, and interactive multimedia title production grows according, pushing publishers towards the production of new styles of publication with unprecendented capacities for lower-cost, higher-volume data and for diverse media types. How multimedia assets are collected, collated, recorded and managed can significantly influence the title production process, making a difference between projects coming in on time and ones running over because of lost information, duplication of effort, or data-transfer bottlenecks between production team members.
This article has concentrated on highlighting these problems and suggesting how best to avoid them. The multimedia database management system described has been designed to serve the needs of a particular project, the life and works of Yeats; its approaches and lessons, however, apply to the production of any multimedia title involving large numbers of assets.