A Model of the Shamisen based on Finite Difference Schemes
Author
Lasickas, Titas
Term
4. Term
Education
Publication year
2020
Submitted on
2020-12-18
Pages
56
Abstract
Many musicians and producers want to use digital versions of instruments without owning or mastering the real thing. Software synthesizers increasingly replace the time-consuming process of recording and sampling real instruments. While common, simpler instruments are already well emulated, rarer and more complex instruments still lack convincing models. This thesis explores a physics-based computer model of the shamisen using finite difference schemes (also known as finite difference time domain, FDTD), a numerical approach that simulates vibrations by updating them step by step over time. The model’s output was assessed with listening-based tests of perceived quality and authenticity. Although the current model does not yet produce a realistic shamisen sound, the results indicate genuine potential for a future digital shamisen built with FDTD methods.
Mange musikere og producere vil gerne bruge digitale versioner af instrumenter uden at eje eller beherske det fysiske instrument. Software-synthesizere erstatter i stigende grad den tidskrævende proces med at optage og samplere rigtige instrumenter. Hvor almindelige og enklere instrumenter allerede er godt efterlignet, mangler sjældnere og mere komplekse instrumenter stadig overbevisende modeller. Dette speciale undersøger en fysikbaseret computermodel af shamisen ved hjælp af finitte differensmetoder (også kendt som finite difference time domain, FDTD) – en numerisk tilgang, der simulerer vibrationer ved at opdatere dem trin for trin over tid. Modellens output blev vurderet med lyttebaserede tests af oplevet kvalitet og autenticitet. Selvom den nuværende model endnu ikke frembringer en realistisk shamisen-lyd, peger resultaterne på et reelt potentiale for en fremtidig digital shamisen baseret på FDTD-metoder.
[This apstract has been rewritten with the help of AI based on the project's original abstract]
Keywords