voice-lab / voicelab Goto Github PK

Hello, was interested to know if there is documentation available to understand what exactly is this doing?

Thanks

First off, great job. Appreciate all your hard work.

Second, before I get into the issue, thought I'd let ya know I've been working on project that uses VoiceLab's python bindings, and to make things easier on myself, I've been converting your settings dicts into settings classes, as well as unifying all the analysis and manipulation classes into a single class, so it's easier to run multiple tasks on a file, by using a single object. In case this is something you'd be interested in incorporating into the library, I thought I'd offer to send it your way when it's done. If not, definitely no worries!

Now on to the issue!

I downloaded the new version today, having used the previous version on a different machine, and loaded up a few sound files. Everything went fine until I tried saving the data, after which the program crashed. I went back in to test if it was just a fluke, and see if there was anything else that was bugging out, and turns out that the "Play Sound(s)" button and "Expand Spectrogram" button don't work either. Essentially, nothing is currently working for me other than loading the sounds, and changing the settings.

Now, since I am running it on a different machine than the previous version, I can't be sure that it's the new release, and not just some bug with this computer's interaction with the software (until I try one of the previous releases on this comp), but I thought I'd let you know, seeing as it was just updated.

The Help button in the Windows version calls up
http://voicelab/help/index.html
(in Internet Explorer, which is not my default browser!)
That link is invalid.

I presume it should go to:
https://voice-lab.github.io/VoiceLab/
Which is very helpful...

I am running the following code on the example wav file provided in the GitHub repository and encoutering the errors.
This is the example code that was taken from the website that I am trying to run:

from Voicelab.toolkits.Voicelab.MeasurePitchNode import MeasurePitchNode

measure_pitch_node = MeasurePitchNode()
measure_pitch_node.args['file_path'] = "tmp.wav"
results = measure_pitch_node.process()
print(results['Mean Pitch (F0) (Praat To Pitch (ac))'])

tmp.wav is the sample file in the Github repository.
The output is :
voice
If I print the contents of results dictionary above, I get the following output:

{'Pitch': "'voice'",
 'Pitch Values (Praat To Pitch (ac))': "'voice'",
 'Mean Pitch (F0) (Praat To Pitch (ac))': "'voice'",
 'Median Pitch (F0) (Praat To Pitch (ac))': "'voice'",
 'Standard Deviation Pitch (F0) (Praat To Pitch (ac)': "'voice'",
 'Pitch Min (F0) (Praat To Pitch (ac))': "'voice'",
 'Pitch Max (F0) (Praat To Pitch (ac))': "'voice'",
 'Pitch (Praat To Pitch (ac))': "'voice'",
 'Pitch Floor': "'voice'",
 'Pitch Ceiling': "'voice'"}

I am running the python code in the same virtual environment as the UI application, I am getting the correct results in the python voicelab.py application.

Is there a way I can get this working?

I am unable to change the default settings for the "Tilt line. qeufrency upper bound " I want to change ut to 0.05 but the "." is not responding

I wonder whether the program can be operated on Linux as the system can not display GUI and show the following error:

Hello,
Congrats on a beautiful piece of software. I am a delighted user.

I am working on a manuscript where I want to report that I used VoiceLab to analyse voice recordings. I would love to report the version I used, but the software does not state or report the version anywhere. And the Help link goes to http://voicelab/help/index.html.

IMHO, it would be helpful to see which version I am running or report the version used in the settings output file.

I am running the GUI version of the program on Mac OS 11.5.2, and when I try to start processing > ~100 audio files, the program consistently crashes. I am able to load them into the editor, just not start the processing. About a month ago I was not running into this problem. I cannot feasibly manually upload and run the processing in batches, as I have ~20k 1 second files I am trying to process.

Any suggestions? Thanks!

Hello,

I've tried both the python script and the I've just downloaded the GUI which is very easy to use and seems far more precise than what I'm getting from Praat. It worked once but it hasn't worked me since. Here's the issue.

Describe the bug
Every time I try to download the results I get an excel file with 0 Ko of data. The settings file is fine but the results file doesn't contain any data. Here's the message I see in my terminal: "All arrays must of the same length".

To Reproduce
Steps to reproduce the behavior:

1. I hit save results
2. I get a message saying the results have been saved successfully
3. Empty file in my folder

Desktop (please complete the following information):

• OS: MacOs
• Environment created with Python 3.9

So first off, awesome work, this is a really cool project.

Second, this is not my forte and I may have missed something vital, so please ignore me if I did something wrong. I think something's missing in the example code for the API. The documentation states that the args['voice'] is a parselmouth.Sound objected, but the nodes treat it as a tuple of the Sound object values and the sample rate (lines 45 and 46 in ManipulateLowerPitchandFormantsNode.py for example). To get it to work, I had to add the following lines to the example code in the API to create the sound object, then decompose to the values and sample rate (for the node to use those to recompile to a Sound object).

import parselmouth
from Voicelab.toolkits.Voicelab.ManipulateLowerPitchAndFormantsNode import ManipulateLowerPitchAndFormantsNode

lower_pitch_and_formants_node = ManipulateLowerPitchAndFormantsNode()
lower_pitch_and_formants_node.args['file_path'] = "./voice.wav"

# Explicitly adding the data
sound = parselmouth.Sound(lower_pitch_and_formants_node.args['file_path'])
lower_pitch_and_formants_node.args['voice'] = sound.values, sound.sampling_frequency

results = lower_pitch_and_formants_node.process()
results['voice'].save(file_path='voice-lower-pitch.wav', format='WAV')

Again, awesome project! Thanks!

When attempting to use a manipulate setting on Windows