Comment on “Novel Web-Based Music Re-Engineering Software for Enhancement of Music Enjoyment Among Cochlear Implantees” by Hwa Et Al. (Otol Neurotol 2021;42(9))

To the Editor With great interest, we have read the recent article “Novel web-based music re-engineering software for enhancement of music enjoyment among cochlear implantees” by Hwa et al. (Otol Neurotol 2021;42(9), 1347–1354), which reports on a music preprocessing method for cochlear implants (CIs) and presents an evaluation with CI users. Although we acknowledge the novel contributions of this study, we disagree with claims made by the authors regarding the novelty of “music re-engineering for CIs” per se and the first use of harmonic/percussive source separation (HPSS) methods in this field, and like to point toward a number of relevant publications from our groups, which have not been considered in the mentioned article. In fact, music preprocessing for CIs has been investigated for almost a decade now. In their pioneering work, Buyens et al. (1) studied the effect of music remixing on music preferences in CI users and found that remixes with amplified vocals, bass, and drums were favored over the original mix. This finding led to the development of several music preprocessing algorithms. HPSS (e.g., Ono et al. (2), Driedger et al. (3)), and remixing was first proposed in the context of music processing for CIs by Buyens et al. (4–6). As an alternative to the HPSS approach, deep-learning–based algorithms for blind source separation and remixing were also derived and evaluated (Pons et al. (7), Gajȩcki and Nogueira (8), Tahmasebi et al. (9)). A different strategy based on dimensionality reduction and sparse subspace tracking methods in the time-frequency domain, which is particularly suited for classical music, was proposed by Nagathil et al. (10,11) and Gauer et al. (12–14). Lentz et al. (15) combined the dimensionality reduction approach with HPSS, facilitating its application to a wider range of music genres. An overview on many of these approaches was presented by Nogueira et al. (16). It seems that both authors and reviewers of the article mentioned previously have completely overlooked all of these publications. We hope that, with this letter, we may contribute to bridging apparently disparate scientific communities. Rainer Martin, Ph.D. Institute of Communication Acoustics Ruhr-Universität Bochum Germany [email protected]Wim Buyens, Ph.D. SoundTalks, Leuven, BelgiumAnil Nagathil, Ph.D. Institute of Communication Acoustics Ruhr-Universität Bochum GermanyWaldo Nogueira, Ph.D. Hanover Medical School and Cluster of Excellence “Hearing4all,” Hanover GermanyBas van Dijk, Ph.D. Cochlear Technology Centre Cochlear Ltd., Mechelen, BelgiumJan Wouters, Ph.D.