Dynamic Filter Coefficients for Acoustic Piano Tone Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing electronic musical instruments struggle to replicate the dynamic tone color changes of an acoustic piano, particularly in controlling the amplitude ratio of harmonic tones, due to limitations in filter characteristics and the PCM method, which fixes tone colors and fails to accurately simulate the attenuation of higher-order harmonics over time.
Innovation Solution
A filter device with a filter coefficient outputting mechanism that uses a parameter table to generate filter coefficients based on frequency and signal strength, allowing for dynamic filter processing that mimics the tone color changes of an acoustic piano by adjusting filter characteristics such as turnover frequency and gain levels, approximating ideal filter characteristics using a combination of IIR filters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a low-order IIR filter is used to minimize computational complexity, then computational requirements are reduced, but the ability to accurately simulate complex tone color changes of an acoustic piano is limited
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting filter coefficients based on multiple parameters including fundamental frequency, harmonic order, and time. This allows a simple low-order IIR filter to simulate complex tone color changes by continuously adapting its characteristics to match the frequency-dependent and time-dependent behavior of acoustic piano harmonics, thereby achieving high simulation accuracy with minimal computational complexity
Solution Approach 2:
The patent implements dynamics by making the filter characteristics time-variant and frequency-dependent. The filter coefficients are not fixed but are dynamically calculated for each harmonic component based on its frequency and the elapsed time since sound onset. This dynamic adaptation enables the filter to accurately track the evolving spectral envelope of the piano sound throughout its attack, sustain, and decay phases
2Ease of operation
If filter characteristics are changed using traditional cutoff frequency and Q parameters, then filter control is simplified, but the ability to replicate acoustic piano tone color changes is insufficient
Solution Approach 1:
The patent transforms the traditional filter control parameters (cutoff frequency and Q) into a new parameter system based on fundamental frequency, harmonic order, and time. By pre-calculating and storing optimal filter coefficients for various combinations of these parameters in lookup tables, the system achieves both ease of operation through simple parameter selection and high tone color replication accuracy through precise coefficient matching
Solution Approach 2:
The patent applies preliminary action by pre-calculating and storing optimal filter coefficients in lookup tables before runtime. These tables contain pre-computed coefficient sets for various fundamental frequencies, harmonic orders, and time points. During actual sound generation, the system only needs to perform simple table lookups and interpolation rather than complex real-time calculations, thereby achieving accurate tone color replication with minimal computational effort
3Illumination intensity
If the amplitude ratio of high-order harmonic tones is increased for strong performance, then tone color brightness is enhanced, but computational complexity increases when dynamically adjusting filter characteristics
Solution Approach 1:
The patent implements dynamics by making the filter characteristics adaptive to performance intensity. When strong performance is detected (higher velocity or amplitude), the system dynamically selects or interpolates filter coefficients that correspond to brighter tone colors with enhanced high-order harmonic content. This dynamic adaptation allows the filter to automatically adjust its frequency response to match the desired tone color for different performance intensities without requiring complex real-time calculations
Solution Approach 2:
The patent applies parameter changes by linking filter coefficient selection to performance parameters such as key velocity or signal amplitude. The system maintains multiple sets of filter coefficients corresponding to different performance intensities, and dynamically switches between or interpolates these sets based on the detected performance strength. This approach enables tone color brightness to be controlled through simple parameter selection rather than complex computational adjustments
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The solution enables electronic musical instruments to achieve tone color changes that closely resemble those of an acoustic piano, providing a more natural and dynamic sound by effectively controlling the attenuation and enhancement of harmonic tones, while reducing computational complexity and hardware requirements.
Implementation Method 1
a low-order IIR filter to minimize computational complexity
Data Source
AI summary
In a filter device, a filter coefficient calculation circuit has a parameter table. The parameter table stores a plurality of sets of filter coefficients associated with a first parameter based on a frequency and a second parameter based on respective plurality of levels representing a degree of attenuation or enhancement of a gain of a filter in filter characteristics. The filter coefficient calculation circuit extracts a set of filter coefficients from a parameter table with the use of the first parameter and the second parameter determined according to a frequency and a strength of a musical sound signal, and outputs the extracted set of filter coefficients to the filter. The filter circuit performs filter processing for the musical sound signal, based on the filter characteristics determined by the set of filter coefficients.