Text-to-Speech Synthesis Techniques for MIDI-to-Audio Synthesis
This is an incremental exploration for audio synthesis researchers, applying existing TTS methods to a new domain of piano music generation.
The study tackled piano MIDI-to-audio synthesis by adapting text-to-speech techniques like Tacotron and neural source-filter models, finding they could be applied with minor modifications but the full system underperformed sample-based or physical-modeling methods, with a bottleneck in converting MIDI to acoustic features.
Speech synthesis and music audio generation from symbolic input differ in many aspects but share some similarities. In this study, we investigate how text-to-speech synthesis techniques can be used for piano MIDI-to-audio synthesis tasks. Our investigation includes Tacotron and neural source-filter waveform models as the basic components, with which we build MIDI-to-audio synthesis systems in similar ways to TTS frameworks. We also include reference systems using conventional sound modeling techniques such as sample-based and physical-modeling-based methods. The subjective experimental results demonstrate that the investigated TTS components can be applied to piano MIDI-to-audio synthesis with minor modifications. The results also reveal the performance bottleneck -- while the waveform model can synthesize high quality piano sound given natural acoustic features, the conversion from MIDI to acoustic features is challenging. The full MIDI-to-audio synthesis system is still inferior to the sample-based or physical-modeling-based approaches, but we encourage TTS researchers to test their TTS models for this new task and improve the performance.