<img src="voicelab/src/Voicelab/favicon.ico">VoiceLab

Automated Reproducible Acoustical Analysis
Voice Lab is an automated voice analysis software. What this software does is allow you to measure, manipulate, and visualize many voices at once, without messing with analysis parameters. You can also save all of your data, analysis parameters, manipulated voices, and full colour spectrograms and power spectra, with the press of one button.

Version 2.0.0

License

VoiceLab is licensed under the MIT license. See the LICENSE file for more information.

Cite Voicelab

If you use VoiceLab in your research, please cite it:

• Feinberg, D. (2022). VoiceLab: Software for Fully Reproducible Automated Voice Analysis. Proc. Interspeech 2022, 351-355.
• Feinberg, D. R., & Cook, O. (2020). VoiceLab: Automated Reproducible Acoustic Analysis. PsyArxiv

Installation instructions:

• Install from pip using Python 3.9-3.11
  • pip install voicelab
• To install on Windows, download the .exe file from the releases page.
  • run the voicelab.exe file
• To install on OSX, download the .zip file from the releases page.
  • Unzip the file, and run the VoiceLab app
• Install on Ubuntu (standalone)
  • Download voicelab
  • change the permissions to executable: chmod +x voicelab
  • try: voicelab or ./voicelab
  • You might need to run this to install a dependency: sudo apt-get install libxcb-xinerama0

Changes from 1.3.1 to 2.0

New Features

• MeasureAlphahNode measures the alpha ratio

  • It is a measure of the ratio of low frequency energy to high frequency energy
  • It's written by me using NumPy

• Pitch-corrected RMS Energy (Voice Sauce) --See bug fixes.

• Pitch-corrected Equivalent Continuous Sound Level (Leq)

• New viewport window for LPC spectra

Bug fixes

• Major bugfix affecting all users of Energy in VoiceLab and Voice Sauce

In Voice Sauce source code, they report to calculate RMS in the documentation, but instead calculated Total energy in each pitch-dependent frame. This means the Energy value in Voice Sauce that was translated to VoiceLab was not scaled for wavelength, and therefore not pitch-independent. To fix this I have implemented and RMS calculation at every frame. You can see the values are much closer to those given by Praat now, but are different, and that is because of the pitch-dependent frame length. I've removed the old calculation of mean energy, and if you are using RMS energy as a single value, that is the RMS of all of the frames. If for some reason you want the old calculation, it is in all of the older versions of VoiceLab.

If you are concerned, I recommend anyone who has published using this algorithm, or plans to in Voice Sauce or older versions of VoiceLab, re-run their Energy measurements and use the new values. The values will be similar, but they won't be the same.

• Fixed spectrograms and spectra
  • You can now see them in the boxes and you can expand them

API is no longer supported until further notice

If you clone the GitHub repo and look in the tests, you can see how to use the API. However, it is not supported at this time.

Contact

David Feinberg: feinberg@mcmaster.ca

Software Development Team

Maintainer

David Feinberg, PhD

Past Contributors

• Oliver Cook
• El Khanoussi Yassine

Documentation

https://voice-lab.github.io/VoiceLab

Voice Lab

Voice Lab is written in Python and relies heavily on a package called parselmouth-praat. parselmouth-praat is a Python package that essentially turns Praat's source code written in C and C++ into a Pythonic interface. What that means is that any praat measurement in this software is using actual Praat source code, so you can trust the underlying algorithms. Voice Lab figures out all analysis parameters for you, but you can always use your own, and these are the same parameters as in Praat, and they do the exact same thing because it is Praat's source code powering everything. That means if you are a beginner an expert, or anything in-between, you can use this software to automate your acoustical analyses.

All of the code is open source and available here on our GitHub repository, so if this manual isn't in-depth enough, and you want to see exactly what's going on, go for it. It is under the MIT license, so you are free to do what you like with the software as long as you give us credit. For more info on that license, see here.

Load Voices Tab

Load Sound File

Press this button to load sound files. You can load as many files as you like. At the moment, Voice Lab processes the following file types:

• wav
• mp3
• aiff
• ogg
• aifc
• au
• nist
• flac

Remove Sound File

Use this button to remove the selected sound file(s) from the list.

Start

Pressing this begins analysis. If you want to run the default analysis, press this button. If you want to select different analyses or adjust analysis parameters, go to the 'Settings' tab and press the 'Advanced Settings' button. Only the files selected (in blue) will be analyzed. By default we will select all files.

Settings Tab

To choose different analyses, select the Use Advanced Settings checkbox. From here, you'll be given the option to select different analyses. You can also change any analysis parameters. If you do change analysis parameters, make sure you know what you are doing, and remember that those same analysis parameters will be used on all voice files that are selected. If you don't alter these parameters, we determine analysis parameters automatically for you, so they are tailored for each voice to give the best measurements.

Save Results

Save Results saves two xlsx files. One is the results.xlsx file and one is the settings.xlsx file. Here you can choose the directory you want to save the files into. You don't have to click on a file, just go to the directory and press the button.

results.xlsx

The results file saves all of the voice measurements that you made. Each measurement gets a separate tab in the xlsx file.

settings.xlsx

This file saves all of the parameters used in each measurement. Each measurement gets a separate tab in the xlsx file. This is great if you want to know what happened. It can also accompany a manuscript or paper to help others replicate analyses.

Measure Duration

This measures the full duration of the sound file. There are no parameters to adjust.

Measure Pitch

This measures voice pitch or fundamental frequency. This uses Praat's Sound: To Pitch (ac)..., by default. You can also use the cross-correlation algorithm: Sound: To Pitch (cc)...{.python .sourceCode}. For full details on these algorithms, see the praat manual pitch page. Measure Pitch returns the following measurements:

• Minimum Pitch
• Maximum Pitch
• Mean Pitch
• Standard Deviation of Pitch
• Pitch Floor
• Pitch Ceiling

We use the automated pitch floor and ceiling parameters described here.<floor-ceiling>

Measure Pitch Yin

This is the Yin implementation from Librosa.

Measure Subharmonics

This measures subharmonic pitch and subharmonic to harmonic ratio. Subharmonic to harmonic ratio and Subharmonic pitch are measures from Open Sauce (Yu et al., 2019), a Python port of Voice Sauce (Shue et al., 2011). These measurements do not use any Praat or Parselmouth code. As in (Shue et al., 2011) and (Yu et al., 2019), subharmonic raw values are padded with NaN values to 201 data points.

• Shue, Y.-L., Keating, P., Vicenik, C., & Yu, K. (2011). VoiceSauce: A program for voice analysis. Proceedings of the ICPhS XVII, 1846–1849.
• Yu, T. M., Murray, R. D., Silverstein, K., & Yu, K. M. (2019). OpenSauce: Open source software for voice analysis, v0.1. https://doi.org/10.5281/zenodo.2638411

Automated pitch floor and ceiling parameters

Praat suggests adjusting pitch settings based on gender . It's not gender per se that is important, but the pitch of voice. To mitigate this, VoiceLab first casts a wide net in floor and ceiling settings to learn the range of probable fundamental frequencies is a voice. Then it remeasures the voice pitch using different settings for higher and lower pitched voices. VoiceLab by default uses employs very accurate. VoiceLab returns minimum pitch, maximum pitch{.python .sourceCode}, mean pitch, and standard deviation of pitch. By default VoiceLab uses autocorrelation for Measuring Pitch<Pitch>{.interpreted-text role="ref"}, and cross-correlation for harmonicity<HNR>, Jitter<Jitter>, and Shimmer<Shimmer>,

Measure Harmonicity

This measures mean harmonics-to-noise-ratio using automatic floor and ceiling settings described here.<floor-ceiling>{.interpreted-text role="ref"} Full details of the algorithm can be found in the Praat Manual Harmonicity Page. By default Voice Lab use To Harmonicity (cc)... You can select To Harmonicity (ac) or change any other Praat parameters if you wish.

Measure Jitter

This measures and returns values of all of Praat's jitter algorithms. This can be a bit overwhelming or difficult to understand which measure to use and why, or ca