Text-to-Speech Synthesis Techniques for MIDI-to-Audio Synthesis

TL;DR

This paper explores adapting text-to-speech synthesis techniques, specifically Tacotron and neural source-filter models, for piano MIDI-to-audio synthesis, highlighting challenges and potential for cross-domain application.

Contribution

It demonstrates that TTS components can be adapted for MIDI-to-audio synthesis, revealing performance bottlenecks and encouraging further research in this cross-domain application.

Findings

TTS components can be applied to MIDI-to-audio synthesis with minor modifications.

Conversion from MIDI to acoustic features remains a key challenge.

Current systems are still inferior to traditional sound modeling methods.

Abstract

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.