Overview

Singer Voice Tester (SVT) is a software application for analyzing singing voices. The program allows users to determine their singing voice type, analyze vocal range, and evaluate practical singing accuracy based on pitch and range.

SVT analyzes a sufficiently long fragment of singing (about 10 seconds) and builds a histogram of the fundamental frequency (F0) distribution across a musical scale spanning four octaves.

Developed as part of the Inton@Trainer project:

https://intontrainer.by/

SVT operates in two main modes:

  • MODE 1: Voice type identification
  • MODE 2: Singing Skills Training

Available as standalone for Windows 7/8/10 and Android platform.

SVT Analysis UI

Who is it for

Vocal Training

Vocal students and teachers

For tracking vocal range development, pitch accuracy, and intonation stability during singing lessons and structured training sessions.

Research & Experiments

Researchers and developers

For experimental studies of voice, intonation patterns, and pitch distribution, as well as for testing and validating vocal analysis algorithms.

Self-Assessment

Singers and voice enthusiasts

For independent analysis of vocal characteristics and comparison with reference performances in a controlled and reproducible way.

Core Features

Voice Type Identification

SVT identifies the user's singing voice type by analyzing pitch distribution and vocal range.

  • Supported Classifications: Full analysis for Male (Bass, Baritone, Tenor) and Female (Contralto, Mezzo-Soprano, Soprano) voices
  • Analytical Output: Provides the degree of match for each voice type and a detailed pitch frequency histogram
  • Visual Tools: Features an interactive 3-octave virtual piano keyboard
  • Precision Range: Specialized octave tracking (Great to 1st for men; Small to 2nd for women) to ensure accuracy across overlapping ranges

Singing Skills Training

This interactive mode evaluates singing accuracy by comparing the user's voice against a reference performance.

  • Dual-Parameter Analysis: Real-time evaluation based on voice pitch and vocal range
  • Visual Feedback: Results are displayed on a "Voice Target" grid, showing deviations in range (X-axis) and pitch (Y-axis)
  • Scoring System: Performance is rated on a precise 10-point scale, where 10 represents a perfect match
  • Users can manually adjust the permissible deviation thresholds for pitch and range in musical notes

Product Gallery

01
Record

Record a sufficiently long fragment of singing (approximately 10 seconds) using a microphone or load a prepared vocal audio file.

02
Analyze

Analyze the acoustic signal by calculating the fundamental frequency (F0) distribution across a multi-octave musical scale.

03
Visualize

Visualize pitch distribution as a histogram aligned with a virtual piano keyboard representing the vocal range.

04
Evaluate

Evaluate singing voice characteristics: determine voice type or assess deviations from a reference performance in pitch and range.

Build from source & Contributing

Use QT Creator or CMake to build Singer Voice Meter. You need to build Inton Core library first.

# Clone the repositories
git clone https://github.com/zhitko/singer-voice-tester.git
cd singer-voice-tester
git clone https://github.com/zhitko/inton-core.git

# Build the project
cmake --build . --target all

Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.
Please make sure to update tests as appropriate.

Specs

  • Format: WAV (8 kHz)
  • Analysis: F0 Frequency
  • Scale: 4-octave
  • OS: Win, Android

Science

  • Pitch Modeling based on F0 distribution validated by experts.
  • Advanced Male/Female range overlapping for vocal monitoring.
  • Professional tool for objective assessment.

Collaboration with RonikaLAB

Created with the participation of RonikaLAB music studio. The application is a core part of their vocal training programs for tracking progress in pitch accuracy.