Selecting any audio input results in an error

[pioneerAlone]

Hi ，

I select any audio input will result in an error, it seems to be related to the system language(My laptop uses Chinese), is there a way to correctly identify it , Or we can let the tool configure the language selection

Regards,
Bake

[Sharrnah]

Hi. thanks for reporting.
I don't think this is because of the special characters since this is still completely handled by the native UI application which should have no issues with UTF-8 characters. 🤔

Do other Audio devices work or is the same error shown for every device?
Did you try to run the Profile despite the error? Since at Profile loading the device ID is handed over to the Python part, that can behave very differently (Though i tried to make them behave as similar as possible)

Did you try to change to MME Audio API? That one is older and does not support Loopback devices, but is often more forgiving about errors.

[pioneerAlone]

Hi , Thanks for reply.

I tried MME API , but it still shows a error as follow:

And also there are garbled codes for audio devices under MME API

Although I have changed the PC system language to English, some parts of the content are still explained in Chinese, including the application UI，So these audio devices still cause errors.

[pioneerAlone]

Hi ，

I‘ve set system language to English，but there's the same issue as follow and I cannot test speaker.

[pioneerAlone]

Since you mentioned that configuration parameter parsing is handled by the native Python program, is there any requirement for the Python version? Since I have installed different Python versions, Python3 is 3.9.1.

[pioneerAlone]

Hi ，

Is there a document that can guide me to build this application from source? I want to try to debug this problem.

[Sharrnah]

Hi. Sorry for the late reply. And sorry for this very long and technical reply now following 👀.

Python is run once you save and start a profile. Python is build using PyInstaller and as such includes everything needed to run already. That includes Python 3.11 currently. So the installed python version on your system should not matter at all.

Since you receive the error already in the Profile setting window, that is not a python problem but from the UI.

The error afterwards is from Python itself.

Can you maybe post the log from the python error you got when starting the profile despite the error in the profile window? Maybe that gives another clue. To get the log, close the error popup and go to Advanced -> Logs where you can press the Copy Button.

I would love to debug this, but i don't have any audio device with special characters in its name, so currently i can only try to research what could be the problem here.

What could be interesting in this case:
You can run the Python backend without the UI with arguments.

"--device_index", c.DeviceIndex,
"--device_out_index", c.DeviceOutIndex,
"--config", c.SettingsFile,

For this you need to first delete the audio_input_device and audio_output_device from the settings yaml file, because otherwise it tries to determine the device id from the name (which could be the culprit here).

To get the Device IDs yourself, you can run the python part with the --devices true argument. (Or use the get-device-list.bat file)

Building the UI should be pretty straight forward. So i haven't written any documentation for source building. I will try to write up what you need here. If you encounter any problems, let me know and in the end i can put all together to write some documentation for it.

UI Build:

• Get newest Golang version https://go.dev/dl/
• Install fyne dependencies. under windows this is MSYS2 MinGW. https://docs.fyne.io/started/
• run go install to download all dependencies of this project.
• build and run using go run (any IDE should be able to do this with a debugger session connected already)

Python Build:

(https://github.com/Sharrnah/whispering)

• Install Python 3.11
• Run install.bat to install all dependencies (For windows + CUDA support). (depending on if some of them have prebuild wheels or not, they might need additional tools. They should tell this)
• You can copy the build UI executable into the root python sourcecode directory to have the UI use the locally installed python version to run instead of the pyInstaller build version.