PratiCanto
Voice analysis and vocal training
What is PratiCanto?
PratiCanto is a voice analysis and training tool developed for speech professionals. The software is split into easy to use specific modules and general modules that provide extensive tools for analysis and training.
PratiCanto helps amateurs and professionals to perfect and analyze spoken voice, singing voice and other aspects related to speech.
Note that PratiCanto does not perform clinical diagnosis; it is a tool designed to help professionals.
How can I download and use PratiCanto?
Download the latest release from our Github. If PratiCanto helps your research, please cite us.
Is it free?
Yes, it is free and open source.
As of 2021, in view of the Covid19 pandemic, Prof. Dr. Domingos Sávio de Ferreira de Oliveira and myself (Douglas Coimbra de Andrade) have decided to open PratiCanto's source under a permissive license to enable speech professionals to work remotely with their clients and foster speech research in general.
What is in PratiCanto?
General modules
Vocal Trainer: allows singing and spoken language practice. The available tools allow checking for correct singing (in tune), rythm and checking voice parameters using vocalization, MIDI files or audio/video playback.
Main features:
- Intensity and pitch training
- Sing with playback of audio or video
- Use MIDI files with lyrics to practice singing
Voice Analysis: aids professionals to perform detailed voice analysis using real time and post-processed spectrogram, side by side comparison, filtering and other advanced tools.
Main features:
- Accurate spectrogram, fundamental frequency and formants computation
- Analyze audio and video files
- Analyze audio files side by side
- Phonetic symbols for audio annotation to save/load with audio files
- Real-time spectrogram
- AI tools for automatic speech tasks
Melody Creator: easy to use tool to create exercises that can be used in the Vocal Trainer and Voice Extension modules.
Specific modules
Sustained Vowel: analyzes voice parameters using a simple test. The sustained vowel test enables the speech pathologist to quickly diagnose common aspects related to voice quality.
Voice Extension: identifies the individual comfort zone for singing by checking the vocal range are reachable for the singer.
Audio Annotation: enables the creation of audio instructions that can be sent through messaging apps. It is possible to intertwine audio (music and exercises, for example) with comments recorded using the microphone.
Speech to Text: automated speech transcription tool that parses spoken language to written text (speech-to-text). Requires Google Speech API key.
Intonation: analyzes spoken language for intonation training. Enables accent adjustment, increase/decrease in voice frequency and general speech practice.
Inspector: performs similarity search in long audios. Intended for correction of bad language habits, identification of speakers and audio summarization.
PratiCanto has an extensive list of features, ranging from drawing the spectrogram on top of the video for analysis, pitch training, real time recording and feedback, and much more. The video below showcases the analysis of a Brazilian actor speech while using the voice of his female character in contrast with his own spoken voice.
How can I help?
You can help in multiple ways:
Like us
Star our repo and add your Like to our Facebook page
Cite our papers
If you find PratiCanto useful in your research, please cite us:
@article{neuralspeechcmd2018freq,
author = {Andrade, Douglas and Trabasso, Luís Gonzaga and Oliveira, Domingos},
year = {2018},
month = {11},
pages = {},
title = {A ROBUST FREQUENCY-DOMAIN METHOD FOR ESTIMATION OF INTENDED FUNDAMENTAL FREQUENCY IN VOICE ANALYSIS}
}
@misc{neuralspeechcmd2018archit,
title={A neural attention model for speech command recognition},
author={Douglas Coimbra de Andrade and Sabato Leo and Martin Loesener Da Silva Viana and Christoph Bernkopf},
year={2018},
eprint={1808.08929},
archivePrefix={arXiv},
primaryClass={eess.AS}
}
Get involved
Use our Github issue tracker to ask for new features or report bugs.