Printto TH Dataset
Thai DiffSinger dataset for multi-speaker training.
About This Dataset
The Printto TH dataset is specifically curated for training DiffSinger to sing in the Thai language. It is sourced from Printto Magicbeat and comprises 59 minutes and 6 seconds of Thai singing data and 6 minutes and 18 seconds of additional foreign language singing data. The dataset is fully labeled using PRINTmov's Thai phoneme system, ensuring precise and accurate phonetic representation for effective training.
Phonemes
Below is the PRINTmov's Thai phoneme system used in this dataset.
The spreadsheet can also be accessed from https://www.printmov.com/thai-diffsinger.html
Download
This section provides access to the sample dataset, primarily designed for multi-speaker training purposes.
Please strictly follow the usage guidelines and rules provided in readme.txt and usage_guidelines.txt within the dataset. It is strictly prohibited from using this dataset with any other person's voice
without obtaining their explicit
consent and permission.
The Printto TH dataset contains fully labeled Thai singing data in NNSVS format separated into separate speakers and languages.
Please note that the dataset may contain some copyrighted lyrics. The primary intended use of this dataset is for educational purposes.
Dataset Usage Terms
• This dataset is intended primarily for educational purposes. Please refrain from using this dataset for any commercial projects.• The current version of the dataset is distributed privately by the owner, PRINTmov.
• You must give appropriate credit using one of the following names: "Printto TH", "Printto", or "PRINTmov".
• You are strictly prohibited from using this dataset with another person's voice clone without obtaining their explicit consent and permission.
• Please review and strictly adhere to the usage guidelines.
• If you have any questions, feel free to contact the author at the addresses below:
Email: printtopipatkasira@gmail.com
Website: http://www.printmov.com
DiffSinger Training Terms
✅ The DiffSinger voicebanks produced through parallel training with this dataset may be used for commercial purposes, such as creating and selling music.❌ Selling or commercially distributing these voicebanks is NOT permitted without explicit permission from PRINTmov, which will be considered on a case-by-case basis.
❌ For any other usage beyond DiffSinger training, please contact PRINTmov directly.
❌ Do NOT produce voice cloning SVC models, such as Diff-SVC or RVC, using the vocals in this dataset.
• If you have any questions, feel free to contact the author at the addresses below:
Email: printtopipatkasira@gmail.com
Website: http://www.printmov.com
- Dataset Name : Printto TH
- Dataset Version : v25.05.08
- Author : Pappim P.
- License : CC BY-NC-SA 4.0
- Additional License : View Additional License
To make it easier to manage the policy, the download link will be sent to your email provided below.
Training Results
Dataset Result
The demo below is the result of DiffSinger training with only samples from Printto TH dataset (without parallel/multi-speaker training).
Multi-speaker Training Result
Dictionary
Thai DiffSinger Dictionary
Thai dsdict-th.yaml is required in order to use the Thai DiffSinger Phonemizer in OpenUTAU. It uses the same phonemes for both long vowels and short vowels, so the timing adjustment should be made manually in OpenUTAU by the user.
Download dsdict-th.yaml File
You can download dsdict-th.yaml taken from Printto Magicbeat's voice library here.
Sample USTx
USTx OpenUTAU Project Files
A USTx file is a musical sequence created by OpenUTAU, an open-source vocal synthesizer designed as a modern successor to UTAU. These files store musical notes, dynamics, vibrato, pitch, and lyrics in the OpenUTAU Sequence Text (USTx) format.
Sample Thai USTx
You can find Thai USTx files on printmov.com in the Assets section. The USTx files marked with a Thai flag are specifically designed for use in OpenUTAU with the Thai phonemizer. Older Thai UST files are also available, but they were created for UTAU's older Thai VCCV system, meaning the phonemes are not the same.