Christian Frisson and Edu Meneses, two members of the Metalab, are interested in the replicability of digital musical interfaces.
In a paper published at the NIME conference, they have studied the impact of documentation of the replicability of this kind of interface.
We asked them a few questions on their work and thoughts on the topic. Interview.
What is a DMI?
Edu: DMI stands for Digital Musical Instrument, or instrument de musique numérique in French. While this is a general idea, we use this term in the music technology research area to define musical instruments that are completely digital, with no acoustic sound generation, and that are normally made of a movement controller and a synthesizer.
Do you have any example of a DMI?
Edu: Yes, we can cite The Hands, created by Michel Waisvisz; Radio Baton, created by Max Mathews; and the Reactable, developed by a team from the Pompeu Fabra University, in Barcelona, as a few examples of famous DMIs. Other instruments used in the last few years are of growing importance, such as the T-Stick and the magnetic resonance piano.
Christian: We work with the IDMIL lab whose main advisor is Marcelo Wanderley at McGill University, who created a great number of DMIs. Here are a few of my favorite DMIs from IDMIL, mostly on the basis of them being longer term projects: the T-Stick as mentioned by Edu - he is one of the main contributors; the augmented nylon guitars called GuitarAMI that were created by Edu, the haptics feedback module TorqueTuner to which I have contributed, and the modular toolbox for DMIs Probatio.
In your own words, how would you define something that can replicated?
Edu : From a practical point of view, in order to make a DMI reproducible, we generally need instructions for the software and hardware components. In digital lutherie, or in instrument making, we usually work from a bill of materials, a firmware and sound generation algorithms, as well as instructions for building dedicated hardware if needed. These information obviously needs to be accessible and available at the building stage.
What are the main challenges that might arise when we try to reproduce a DMI?
Edu : Usually, any missing information is a major challenge. If any of the previously mentionned components are missing, the builder might have to improvize or to abandon the project. Other factors that frequently arises are materials availability and cost. Materials often become obsolete and might stop being available. In countries where the funding for research is limited, it can be difficult to get the materials needed.
Have you ever tried to reproduce a DMI?
Edu : Yes. One interesting case is the T-Stick. Theoretically, this DMI is only a movement-based controller of another DMI, which needs the audio part to be complete. The material documentation was adequate and the software only needed a few changes. It was interesting to witness how, after having played the instrument a few time, the instrument maker created a relationship with the instrument, giving unique caracter to the builded DMIs, similar to a luthier and guitars that they have built. We are quite close to being able to tell who built an instrument given how it is used.
Christian: Albert-Ngabo Niyonsenga is currently at IDMIL as a graduate student. His research is on the maintainability and reliability of DMIs. Together, we have worked on the maintenance of TorqueTuner, the haptics feedback module, in order to verify its replicability. Some of the work involved changing the material drivers for the actuator and for the rotation sensor since the previous one is not produced anymore, the new one featuring an engine technology that is more transparent for the user; adding a LCD Display to access the collection of embedded haptics effects; and finally, get a new home for the module by using a new 3D printed box made to be used as a device controller. For more information, access the project at: Sustainable Haptic Development.
According to you, what are the main tools and processes that helps with improving DMIs replicability?
Edu: A complete documentation is a first important step to make it possible for others to reproduce a DMI. Clear texts and pictures are really helpful when you play an instrument. A second important piece is a video version of the instructions. While the video is not as useful as the text/pictures combo to recreate the instrument step-wise, it allows the builder to get ready and to review the process as many times as needed in order to become familiar with the instrument before starting the construction process.
Christian: To add to Edu’s answer, I would say that it is essential to choose from open source software and hardware for the dependencies of the project in order to improve its longevity and to stay open on new perspectives for future developments.
What are the questions you would like to explore in future research?
Edu: The impact and effects of documentation on replicability have not been studied that much up to this day in the music technology field, even though replicability by itself has been well discussed in the DMI literature. An interesting research topic would be to discuss the effects of a diverse range of documentation media on the DMI replicability. Another interesting option could be to to study the effect of replicability on an instrument longevity, meaning if the extent of the documentation and the level of difficulty in reproducing the instrument helps make it useful and relevant to musical research and interpretation.
Christian: For further readings, we suggest that you access our publication: Calegario, F., Tragtenberg, J., Frisson, C., Meneses, E., Malloch, J., Cusson, V., Wanderley, M. M. (2021). Documentation and Replicability in the NIME Community. Proceedings of the 2021 International Conference on New Interfaces for Musical Expression (NIME 2021)., Shanghai, China.