Matt Chiu
Research Subjects
Matt Chiu has a total of 32 co-authors. Here are the top co-authors by number of shared publications:
Amy E. Schrlau
Department of Biomedical Engineering, University of Rochester, Rochester, NY, USA
1 shared publication- Andrew J. Oxenham
Department of Psychology, University of Minnesota, Minneapolis, MN 55455, USA
1 shared publication Anna Hagedorn
Department of Speech, Language, and Hearing Sciences, Purdue University, West Lafayette, IN, USA
1 shared publicationAnne E. Luebke
Department of Biomedical Engineering, University of Rochester, Rochester, NY, USA
1 shared publicationAshfaquzzaman Khan
1 shared publicationAudra Irvine
Neuroscience Institute, Carnegie Mellon University, Pittsburg, PA, USA
1 shared publication
Latest Publications
- 14 June 2026
- Published by Springer Nature
- p. 505-514
Towards a Formalization of Fretboards and Gestural Effort
- 14 June 2026
- Published by Springer Nature
- p. 505-514
Towards a Formalization of Fretboards and Gestural Effort
- 14 June 2026
- Published by Springer Nature
- p. 487-501
The Predictive Benefit of Order Increases in Melodic N-Gram Models
- 14 June 2026
- Published by Springer Nature
- p. 487-501
The Predictive Benefit of Order Increases in Melodic N-Gram Models
- Top Cited PapersOpen Access
- 8 August 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 7152
Large-scale multi-site study shows no association between musical training and early auditory neural sound encoding
Musical training has been reported to be associated with enhanced neural processing of sounds, as measured via the frequency following response (FFR), implying the potential for human subcortical neural plasticity. We conducted a large-scale multi-site preregistered study (n > 260) to replicate and extend the findings underpinning this important relationship. We failed to replicate any of the major findings selected for replication that were published previously in smaller studies. Musical training was not associated with enhanced neural encoding strength of a speech stimulus (/da/) in babble, whether measured via the spectral or temporal representations of the FFR. Similarly, the strength of neural tracking of a speech sound with a dynamic pitch trajectory was not related to either years of musical training or age of onset of musical training. Our findings provide no evidence for associations between early auditory neural responses and either musical training or musical ability.
- 8 August 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 7152
Top Cited PapersOpen AccessLarge-scale multi-site study shows no association between musical training and early auditory neural sound encoding
Musical training has been reported to be associated with enhanced neural processing of sounds, as measured via the frequency following response (FFR), implying the potential for human subcortical neural plasticity. We conducted a large-scale multi-site preregistered study (n > 260) to replicate and extend the findings underpinning this important relationship. We failed to replicate any of the major findings selected for replication that were published previously in smaller studies. Musical training was not associated with enhanced neural encoding strength of a speech stimulus (/da/) in babble, whether measured via the spectral or temporal representations of the FFR. Similarly, the strength of neural tracking of a speech sound with a dynamic pitch trajectory was not related to either years of musical training or age of onset of musical training. Our findings provide no evidence for associations between early auditory neural responses and either musical training or musical ability.
- Open Access
- 30 January 2024
- Published by SAGE Publications in Music & Science
- Vol. 7
Melodic Differences Between Styles: Modeling Music With Step Inertia
A well-known phenomenon in melodic structure is “step inertia”: the tendency for a step to be followed by another step in the same direction. There is strong evidence of step inertia in three corpora of Western common-practice melodies: European folk songs, classical instrumental themes, and English hymn tunes. Surprisingly, modern Western popular music does not reflect step inertia. In Billboard’s Hot 100, and Rolling Stone magazine's list of “greatest songs,” inertial (same-direction) steps are less likely than non-inertial ones. To further explore the role of step inertia in different corpora, we created a generative model that assigns probabilities to melodies, considering just four factors: range, pitch proximity, scale-degree frequency within a key, and step inertia. We optimized the weights of these factors for the Essen Folksong Collection and the Billboard corpus, and compared them with n-gram models. The optimal (normalized) weight of the inertia factor is large and positive for the Essen collection (.51) and small for the Billboard corpus (.02). This is further evidence that step inertia plays a much smaller role in popular melodies than common-practice ones, and that non-inertial steps are slightly favored.
- 30 January 2024
- Published by SAGE Publications in Music & Science
- Vol. 7
Open AccessMelodic Differences Between Styles: Modeling Music With Step Inertia
A well-known phenomenon in melodic structure is “step inertia”: the tendency for a step to be followed by another step in the same direction. There is strong evidence of step inertia in three corpora of Western common-practice melodies: European folk songs, classical instrumental themes, and English hymn tunes. Surprisingly, modern Western popular music does not reflect step inertia. In Billboard’s Hot 100, and Rolling Stone magazine's list of “greatest songs,” inertial (same-direction) steps are less likely than non-inertial ones. To further explore the role of step inertia in different corpora, we created a generative model that assigns probabilities to melodies, considering just four factors: range, pitch proximity, scale-degree frequency within a key, and step inertia. We optimized the weights of these factors for the Essen Folksong Collection and the Billboard corpus, and compared them with n-gram models. The optimal (normalized) weight of the inertia factor is large and positive for the Essen collection (.51) and small for the Billboard corpus (.02). This is further evidence that step inertia plays a much smaller role in popular melodies than common-practice ones, and that non-inertial steps are slightly favored.
- 4 June 2022
- Published by Springer Nature
- p. 405-410
Investigating Style with Scale Embeddings
- 4 June 2022
- Published by Springer Nature
- p. 343-348
Identifying Metric Types with Optimized DFT and Autocorrelation Models
- 4 June 2022
- Published by Springer Nature
- p. 343-348
Identifying Metric Types with Optimized DFT and Autocorrelation Models
- 1 September 2021
- Published by Society for Music Theory in Music Theory Online
- Vol. 27 (3)
Macroharmonic Progressions through the Discrete Fourier Transform
This article examines macroharmony through the lens of thediscrete Fourier transform(DFT) using computational analysis. It first introduces the DFT, giving an interpretive framework to understand the theory of chord quality first introduced by Ian Quinn (2007) before extending the theory to macroharmonies. Subsequently, the paper discusses different approaches—including different weighting and windowing procedures—to retrieving pitch data for computational analysis. An analysis of macroharmony inDomine Jesufrom Maurice Duruflé’s Requiem, Op. 9 follows. I show that the DFT reflects intuition, reveals form-functional macroharmonies in the movement, and provides us with a perspective to find novel hearings.
- 1 September 2021
- Published by Society for Music Theory in Music Theory Online
- Vol. 27 (3)
Macroharmonic Progressions through the Discrete Fourier Transform
This article examines macroharmony through the lens of thediscrete Fourier transform(DFT) using computational analysis. It first introduces the DFT, giving an interpretive framework to understand the theory of chord quality first introduced by Ian Quinn (2007) before extending the theory to macroharmonies. Subsequently, the paper discusses different approaches—including different weighting and windowing procedures—to retrieving pitch data for computational analysis. An analysis of macroharmony inDomine Jesufrom Maurice Duruflé’s Requiem, Op. 9 follows. I show that the DFT reflects intuition, reveals form-functional macroharmonies in the movement, and provides us with a perspective to find novel hearings.
- 1 May 2011
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 73-76
Efficient Calculation of Pairwise Nonbonded Forces
A major bottleneck in molecular dynamics (MD) simulations is the calculation of the pair wise nonbonded interactions. Previous work on FPGAs has shown that these calculations can be implemented with a number of force computation pipelines operating in parallel (4 and 8 for the Stratix-III and Stratix-V, respectively). Optimization has received some attention previously in CPU, GPU, FPGA, and ASIC implementations, with direct computation of the equations of interaction being replaced with table lookup with interpolation, and the order and granularity of those interpolations being optimized. FPGAs lend themselves to a particularly rich design space both of opportunities and constraints. We explore and evaluate this space with respect to both resource requirements and simulation quality. We find that FPGAs' BRAM architecture makes them well suited to support unusually fine-grained intervals. This leads to a reduction in other logic and a proportional increase in performance. We demonstrate these designs with prototype implementations supporting full electrostatics and integrated into NAMD-lite. Throughput is improved by 50% over the previous best FPGA implementation while simulation quality is maintained.
- 1 May 2011
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 73-76
Efficient Calculation of Pairwise Nonbonded Forces
A major bottleneck in molecular dynamics (MD) simulations is the calculation of the pair wise nonbonded interactions. Previous work on FPGAs has shown that these calculations can be implemented with a number of force computation pipelines operating in parallel (4 and 8 for the Stratix-III and Stratix-V, respectively). Optimization has received some attention previously in CPU, GPU, FPGA, and ASIC implementations, with direct computation of the equations of interaction being replaced with table lookup with interpolation, and the order and granularity of those interpolations being optimized. FPGAs lend themselves to a particularly rich design space both of opportunities and constraints. We explore and evaluate this space with respect to both resource requirements and simulation quality. We find that FPGAs' BRAM architecture makes them well suited to support unusually fine-grained intervals. This leads to a reduction in other logic and a proportional increase in performance. We demonstrate these designs with prototype implementations supporting full electrostatics and integrated into NAMD-lite. Throughput is improved by 50% over the previous best FPGA implementation while simulation quality is maintained.