基于MIDI控制器的软音源插件自动控制系统研究

Research on the automatic control system of soft audio source connector based on MIDI controller

针对传统电子音乐系统存在软音源插件控制效果不佳,设计一个基于MIDI控制器的软音源插件自动控制系统,并实现音符的识别。其中,系统硬件部分主要由AT89C51单片机的MIDI控制器、USB通信和音源和上位机组成;系统软件部分则采用基于RBF人工神经网络和MFCC特征提取相结合的算法进行音乐音符自动识别;最后通过识别结果进行软音源插件自动控制。结果表明,在低音符和高音符的检测结果中,基于RBF神经元网络的正确识别率均保持在80%及以上,识别精度较高。噪声环境下,基于RBF神经元网络的别正确率明显高于现有的基音检测方法,识别效果和稳定性更强,基于此方法可提升软音源插件的自动控制效果,具备一定的可行性。

In view of the low accuracy of soft sound source plug-in in traditional electronic music system,a system based on MIDI controller is proposed.The hardware part consists of the MIDI controller,USB communication,audio source and upper computer;the software part adopts the algorithm based on RBF artificial neural network and MFCC feature extraction,and automatically controls the soft sound source plug-in through the recognition result.The simulation results show that the correct recognition rate based on the RBF neuron network remains at 80%and above,and the recognition accuracy is high.In the noise environment,the accuracy rate of RBF neurons network is significantly higher than the existing base sound detection method,and the recognition effect and stability are stronger.Based on this method,the automatic control effect of soft sound source plug-in can be improved,which has certain feasibility.