wav2svp: Waveform to Synthesizer V Project

Description

wav2svp is a project that converts a waveform to a Synthesizer V Project (SVP) file. It is based on the SOME and RMVPE. In addition to automatically extracting MIDI, this project can also extract pitch lines simultaneously. But unfortunately, at present, it's unable to simultaneously extract lyrics.

Usage

You can download the One click startup package from releases, unzip and double click go-webui.bat to start the WebUI.

Run from Code

1. Clone this repository and install the dependencies. We recommand to use python 3.10.

   pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118
   pip install -r requirements.txt

2. Download pre-trained models:

   • 0119_continuous128_5spk and unzip it to weights.
   • rmvpe and unzip it to weights and rename it to rmvpe.pt.
   • Order the weights folder as follows:

   weights
     ├-config.yaml
     ├-model_steps_64000_simplified.ckpt
     └-rmvpe.pt

3. Run the following command to start WebUI:

   python webui.py

4. You can download the inference results from WebUI interface or from the results folder.

Command Line Usage

Use infer.py:

usage: infer.py [-h] [--model_path MODEL_PATH] [--tempo TEMPO] audio_path

Inference for wav2svp

positional arguments:
  audio_path            Path to the input audio file

options:
  -h, --help            show this help message and exit
  --model_path MODEL_PATH
                        Path to the model file, default: weights/model_steps_64000_simplified.ckpt
  --tempo TEMPO         Tempo value for the midi file, default: 120

You can find the results in the results folder.

Thanks

• [openvpi/SOME] openvpi/SOME
• [Dream-High/RMVPE] Dream-High/RMVPE
• [yxlllc/RMVPE] yxlllc/RMVPE