The work here aims to evaluate the usability of software applications and define their quality for stakeholders in the music industry. Initial work focuses on standardised tools and procedures and sets benchmarks for performance times and completion rates across software packages, before aiming to make some suggestions about how improvements could be made in the design of said interfaces. Further work goes on to explore industry tools in the context of the real world, live performance tools, categorising them according to purpose and evaluating their success. Finally, a series of workshops and discussion groups aim to identify problems and solutions, suggesting a novel way of evaluating music information systems from a usability perspective. The work here explores usability issues in terms of efficiency, effectiveness and user satisfaction, showing that systems can fail in all three categories. While typical software tools such as Cubase are found to be somewhat usable, the changing requirements of users mean that software systems are no longer effective in performing day to day tasks required of them. There is further exploration into how software tools are used incorrectly or inefficiently, where learning curves are too steep to overcome and where systems inevitably fail. The thesis culminates in a suggested set of heuristics which can be used to evaluate current systems and used as a guideline in developing human-centred systems within the context of music performance and production. The work highlights the strengths of existing systems in terms of enabling creativity and providing an efficient platform for content creation, while making suggestions about future directions of such systems including a discussion in social web integration and pervasive interfaces.