.TH VOIRS-CLI-TEST 1 "2025-07-05" "voirs-cli" "VoiRS CLI Manual"
.SH NAME
voirs-cli-test \- Test VoiRS synthesis pipeline and validate installation
.SH SYNOPSIS
.B voirs-cli test
[\fIOPTIONS\fR]
[\fITEST_TEXT\fR]
.SH DESCRIPTION
The test command provides a quick way to validate that the VoiRS synthesis pipeline is working correctly. It performs end-to-end synthesis with a test phrase, verifying that all components (G2P, acoustic model, vocoder) are functioning properly.
This command is particularly useful after installation, model downloads, or configuration changes to ensure the system is operating correctly.
.SH OPTIONS
.TP
.I TEST_TEXT
Custom text to use for testing. If not provided, uses the default test phrase: "Hello, this is a test of VoiRS speech synthesis."
.TP
.B \-\-play
Play the generated audio through system speakers instead of saving to file.
.TP
.B \-v, \-\-voice \fIVOICE\fR
Specify voice to use for testing. Uses default voice if not provided.
.TP
.B \-q, \-\-quality \fILEVEL\fR
Set synthesis quality level for testing (low, medium, high, ultra).
.TP
.B \-f, \-\-format \fIFORMAT\fR
Output format for test audio file (wav, mp3, flac, opus, ogg).
.TP
.B \-o, \-\-output \fIFILE\fR
Specify output file for test audio. Default: "voirs_test_TIMESTAMP.wav"
.TP
.B \-\-benchmark
Include performance benchmarking in the test, measuring synthesis speed and quality metrics.
.TP
.B \-\-verbose
Enable detailed output showing each pipeline stage and timing information.
.TP
.B \-\-validate-models
Perform additional validation of loaded models including checksum verification.
.SH PIPELINE VALIDATION
The test command validates the following pipeline components:
.SS Text Processing
.TP
.B Text normalization
Validates unicode handling, number expansion, and abbreviation processing.
.TP
.B SSML parsing
Tests Speech Synthesis Markup Language parsing capabilities.
.SS Grapheme-to-Phoneme (G2P)
.TP
.B Phoneme conversion
Verifies text-to-phoneme conversion accuracy.
.TP
.B Language detection
Tests automatic language detection for multilingual text.
.SS Acoustic Modeling
.TP
.B Mel spectrogram generation
Validates acoustic model inference and mel spectrogram quality.
.TP
.B Speaker conditioning
Tests speaker embedding and voice characteristics.
.SS Vocoding
.TP
.B Audio generation
Verifies vocoder model inference and waveform generation.
.TP
.B Audio post-processing
Tests audio enhancement, normalization, and format conversion.
.SH EXAMPLES
.TP
.B Basic system test
voirs-cli test
.TP
.B Test with custom text
voirs-cli test "The quick brown fox jumps over the lazy dog"
.TP
.B Test with audio playback
voirs-cli test --play
.TP
.B Test specific voice
voirs-cli test --voice en-US-female-calm
.TP
.B Comprehensive test with benchmarking
voirs-cli test --benchmark --verbose --validate-models
.TP
.B Test different quality levels
voirs-cli test --quality ultra --format flac
.TP
.B Test SSML capabilities
voirs-cli test '<speak><prosody rate="slow">Testing SSML</prosody></speak>'
.SH OUTPUT
The test command provides detailed feedback:
.SS Success Output
- Pipeline validation results
- Synthesis timing information
- Audio quality metrics
- Generated file location
.SS Error Diagnostics
- Component failure identification
- Suggested troubleshooting steps
- Configuration recommendations
- Model availability status
.SH PERFORMANCE METRICS
When using \fB\-\-benchmark\fR, the following metrics are reported:
.TP
.B Real-Time Factor (RTF)
Ratio of synthesis time to audio duration. Values < 1.0 indicate real-time capability.
.TP
.B Audio Quality
Basic quality metrics including peak amplitude, RMS level, and THD+N.
.TP
.B Memory Usage
Peak memory consumption during synthesis.
.TP
.B Component Timing
Individual timing for G2P, acoustic model, and vocoder stages.
.SH TROUBLESHOOTING
Common issues and solutions:
.TP
.B Model not found
Run 'voirs-cli download-model MODEL_ID' to install required models.
.TP
.B Audio playback fails
Check system audio configuration or use file output instead.
.TP
.B Poor synthesis quality
Try different voice or quality settings, or update models.
.TP
.B Performance issues
Use 'voirs-cli optimize-model' or adjust thread/GPU settings.
.SH EXIT STATUS
.TP
.B 0
Test completed successfully, all components working.
.TP
.B 1
General test failure or invalid usage.
.TP
.B 2
Configuration error preventing test execution.
.TP
.B 3
Required models or voices not available.
.TP
.B 4
Audio processing or output error.
.SH FILES
.TP
.B voirs_test_TIMESTAMP.wav
Default output file for test audio.
.TP
.B ~/.config/voirs/config.toml
Configuration file affecting test behavior.
.TP
.B ~/.cache/voirs/
Cache directory for temporary test files.
.SH ENVIRONMENT VARIABLES
.TP
.B VOIRS_TEST_VOICE
Default voice for testing.
.TP
.B VOIRS_TEST_QUALITY
Default quality level for testing.
.TP
.B VOIRS_AUDIO_DEVICE
Audio output device for playback testing.
.SH SEE ALSO
.BR voirs-cli (1),
.BR voirs-cli-synthesize (1),
.BR voirs-cli-models (1),
.BR voirs-cli-voices (1),
.BR voirs-cli-config (1)
.SH AUTHOR
VoiRS Development Team
.SH REPORTING BUGS
Report bugs at: https://github.com/voirs-project/voirs/issues