chidiwilliams/buzz

Buzz transcribes and translates audio offline on your personal computer. Powered by OpenAI's Whisper.

Buzz

Documentation | Buzz Captions on the App Store

Transcribe and translate audio offline on your personal computer. Powered by OpenAI's Whisper.

Buzz is better on the App Store. Get a Mac-native version of Buzz with a cleaner look, audio playback, drag-and-drop import, transcript editing, search, and much more.

Installation

PyPI:

pip install buzz-captions
python -m buzz

macOS:

brew install --cask buzz

Windows:

Download and run the .exe file in the releases page.

Linux:

sudo apt-get install libportaudio2
sudo snap install buzz

buzz's Issues

UX suggestions

1. Set application window title (The app window is shown as "Unknown" in my Ubuntu 22.04.1)
2. Sort the list of languages in alphabetical order

Unable to open Buzz app on Mac once installed

Have not been able to successfully open the Buzz app once it is in the Applications folder. Seems that this is likely an OS binary verification issue, but not familiar enough with the topic to be sure

Set detect language as default language

Default microphone to system default

Run openai whisper using process like whisper.cpp

Won't run on mac M1

I get warning saying application is damaged.

MacBook M1 Pro MacOS Monterey

Where does the model get stored?

Hi,

I downloaded the large model, but can't seem to locate where it is stored in my Mac.
Also, there should be a cancel button for the "Downloading Resources" dialog.

Cheers!

Second window randomly opens after recording starts

• I think this has something to do with the multi-threaded recording

Export as .SRT

Linux responds with Segmentation fault on startup

With the latest version (0.4.2 ) gives this :

$ ./Buzz
[39642] PyInstaller Bootloader 5.x
[39642] LOADER: executable is /home/luis/Downloads/dist/Buzz/Buzz
[39642] LOADER: homepath is /home/luis/Downloads/dist/Buzz
[39642] LOADER: _MEIPASS2 is NULL
[39642] LOADER: archivename is /home/luis/Downloads/dist/Buzz/Buzz
[39642] LOADER: Cookie found at offset 0xCD1B58
[39642] LOADER: No need to extract files to run; setting up environment and restarting bootloader...
[39642] LOADER: LD_LIBRARY_PATH=/home/luis/Downloads/dist/Buzz
[39642] PyInstaller Bootloader 5.x
[39642] LOADER: executable is /home/luis/Downloads/dist/Buzz/Buzz
[39642] LOADER: homepath is /home/luis/Downloads/dist/Buzz
[39642] LOADER: _MEIPASS2 is /home/luis/Downloads/dist/Buzz
[39642] LOADER: archivename is /home/luis/Downloads/dist/Buzz/Buzz
[39642] LOADER: Cookie found at offset 0xCD1B58
[39642] LOADER: Already in the child - running user's code.
[39642] LOADER: Python library: /home/luis/Downloads/dist/Buzz/libpython3.9.so.1.0
[39642] LOADER: Loaded functions from Python library.
[39642] LOADER: Manipulating environment (sys.path, sys.prefix)
[39642] LOADER: sys.prefix is /home/luis/Downloads/dist/Buzz
[39642] LOADER: Pre-init sys.path is /home/luis/Downloads/dist/Buzz/base_library.zip:/home/luis/Downloads/dist/Buzz/lib-dynload:/home/luis/Downloads/dist/Buzz
[39642] LOADER: Setting runtime options
[39642] LOADER: Initializing python
[39642] LOADER: Overriding Python's sys.path
[39642] LOADER: Post-init sys.path is /home/luis/Downloads/dist/Buzz/base_library.zip:/home/luis/Downloads/dist/Buzz/lib-dynload:/home/luis/Downloads/dist/Buzz
[39642] LOADER: Setting sys.argv
[39642] LOADER: setting sys._MEIPASS
[39642] LOADER: importing modules from CArchive
[39642] LOADER: extracted struct
[39642] LOADER: running unmarshalled code object for struct...
[39642] LOADER: extracted pyimod01_archive
[39642] LOADER: running unmarshalled code object for pyimod01_archive...
[39642] LOADER: extracted pyimod02_importers
[39642] LOADER: running unmarshalled code object for pyimod02_importers...
[39642] LOADER: extracted pyimod03_ctypes
[39642] LOADER: running unmarshalled code object for pyimod03_ctypes...
[39642] LOADER: Installing PYZ archive with Python modules.
[39642] LOADER: PYZ archive: PYZ-00.pyz
[39642] LOADER: Running pyiboot01_bootstrap.py
[39642] LOADER: Running pyi_rth_pkgutil.py
[39642] LOADER: Running pyi_rth_inspect.py
[39642] LOADER: Running pyi_rth_subprocess.py
[39642] LOADER: Running pyi_rth_setuptools.py
[39642] LOADER: Running pyi_rth_pkgres.py
[39642] LOADER: Running pyi_rth__tkinter.py
[39642] LOADER: Running pyi_rth_pyqt5.py
[39642] LOADER: Running pyi_rth_multiprocessing.py
[39642] LOADER: Running main.py
fish: Job 1, './Buzz' terminated by signal SIGSEGV (Address boundary error)

My SO it's Kubuntu 22.04.01

Add custom icon

[documentation] Minimum MacOS version

Hi,

I tried to run Buzz on my Macbook pro 13" 2018 with 10.14.6 and it failed (immediately close on opening)

My guess is that my MacOS version is too old. Maybe it could be useful to mention the minimum supported version on MacOS on the readme page?

Update when system mics change (get removed/added)

Allow uploading audio files

Add support for GPU inference using CUDA

Keystroke audio recording

Hey man, firstly wanna say thanks for this cool app! I'm just wondering about one feature:

audio recording will not last a specific time until you press stop, but as long as you press one of the keypad buttons

You could think of it as a Discord / any video game microphone button that only records you as long as you press the key. I'm just also wondering about building an opensource "Developer Assistant" using Whisper, and I thought your app with such a feature would be pretty handy 🙌

Release crashes on Windows

Hello, thanks for sharing.
I am getting this error.
Also when opening as admin

Cheers

Add stable-ts support for more accurate timing

https://github.com/jianfch/stable-ts

Download models off main thread

Access violation exception when running whisper.cpp shared library on Windows

Compile whisper.cpp with:

gcc -O3 -std=c11   -pthread -mavx -mavx2 -mfma -mf16c -fPIC -c whisper.cpp/ggml.c -o whisper.cpp/ggml.o
g++ -O3 -std=c++11 -pthread --shared -fPIC -static-libstdc++ -DWHISPER_SHARED -DWHISPER_BUILD whisper.cpp/whisper.cpp whisper.cpp/ggml.o -o libwhisper.so

Then run:

whisper_cpp = ctypes.CDLL("libwhisper.so")

# Calling any one of the functions errors
whisper_cpp.whisper_init('path/to/model.bin'.encode('utf-8'))
whisper_cpp.whisper_lang_id(b'en`)

Stack trace:

Windows fatal exception: access violation

Current thread 0x00002b30 (most recent call first):
  File "C:\Users\willi\Documents\src\buzz\whispercpp_test.py", line 17 in <module>
Windows fatal exception: access violation

Current thread 0x00002b30 (most recent call first):
  File "C:\Users\willi\Documents\src\buzz\whispercpp_test.py", line 17 in <module>
Windows fatal exception: access violation

Current thread 0x00002b30 (most recent call first):
  File "C:\Users\willi\Documents\src\buzz\whispercpp_test.py", line 17 in <module>
Windows fatal exception: access violation

...

Current thread 0x00002b30 (most recent call first):
  File "C:\Users\willi\Documents\src\buzz\whispercpp_test.py", line 17 in <module>
Windows fatal exception: stack overflow

Current thread 0x00002b30 (most recent call first):
  File "C:\Users\willi\Documents\src\buzz\whispercpp_test.py", line 17 in <module>
Windows fatal exception: access violation

Current thread 0x00002b30 (most recent call first):
  File "C:\Users\willi\Documents\src\buzz\whispercpp_test.py", line 17 in <module>
Windows fatal exception: access violation

Current thread 0x00002b30 (most recent call first):
  File "C:\Users\willi\Documents\src\buzz\whispercpp_test.py", line 17 in <module>

App seems to crash after closing on Mac

Linux build doesn't work

Linux binary archive contains libraries and executables for macOS, not Linux

Add indication that recording/processing is in progress

Add own ggml models

Clean up properly when a window is closed while recording is in session

Windows Binary?

Are you planning for Windows binary?

Ubuntu install problem

Something isn't right with the source or instructions:

user@hp-laptop:~/Downloads/buzz$ pip install .
Defaulting to user installation because normal site-packages is not writeable
Processing /home/user/Downloads/buzz
  Installing build dependencies ... done
  Getting requirements to build wheel ... done
  Preparing metadata (pyproject.toml) ... error
  error: subprocess-exited-with-error
  
  × Preparing metadata (pyproject.toml) did not run successfully.
  │ exit code: 1
  ╰─> [16 lines of output]
      Traceback (most recent call last):
        File "/usr/lib/python3/dist-packages/pip/_vendor/pep517/in_process/_in_process.py", line 363, in <module>
          main()
        File "/usr/lib/python3/dist-packages/pip/_vendor/pep517/in_process/_in_process.py", line 345, in main
          json_out['return_val'] = hook(**hook_input['kwargs'])
        File "/usr/lib/python3/dist-packages/pip/_vendor/pep517/in_process/_in_process.py", line 164, in prepare_metadata_for_build_wheel
          return hook(metadata_directory, config_settings)
        File "/tmp/pip-build-env-s16k_t71/overlay/local/lib/python3.10/dist-packages/poetry/core/masonry/api.py", line 41, in prepare_metadata_for_build_wheel
          builder = WheelBuilder(poetry)
        File "/tmp/pip-build-env-s16k_t71/overlay/local/lib/python3.10/dist-packages/poetry/core/masonry/builders/wheel.py", line 56, in __init__
          super().__init__(poetry, executable=executable)
        File "/tmp/pip-build-env-s16k_t71/overlay/local/lib/python3.10/dist-packages/poetry/core/masonry/builders/builder.py", line 83, in __init__
          self._module = Module(
        File "/tmp/pip-build-env-s16k_t71/overlay/local/lib/python3.10/dist-packages/poetry/core/masonry/utils/module.py", line 69, in __init__
          raise ModuleOrPackageNotFound(
      poetry.core.masonry.utils.module.ModuleOrPackageNotFound: No file/folder found for package buzz
      [end of output]
  
  note: This error originates from a subprocess, and is likely not a problem with pip.
error: metadata-generation-failed

× Encountered error while generating package metadata.
╰─> See above for output.

note: This is an issue with the package mentioned above, not pip.
hint: See above for details.

App crashes when both windows are closed on Mac

Remove option to choose delay

Would be useful to choose default delay for user based on maybe speed of completing previous chunk. Padding the chunks would also be quite useful here.

Change from QWidgets to QML

Seems like it might fix the issue with fonts on Mac?

Change model selection to quality

Quality:

• very low
• low
• medium
• high
• very high

Calculate recommended quality based on system specs

Add modal to show when model is downloading

Add button to check for updates from about dialog

Add opt-in support for whisper.cpp

• Whisper.cpp implements high-perf inference of Whisper's models.

Sounddevice is an unmet dependency

Traceback (most recent call last):
File "/buzz/main.py", line 3, in
from gui import Application
File "/buzz/gui.py", line 5, in
import sounddevice
ModuleNotFoundError: No module named 'sounddevice'

Add large model selection in the "import audio file" option, please

Can you please add the large model selection in the import audio file option?
There are many errors when transcribing in high quality. It's not enough for an accurate transcription.
I could see that the first versions of Buzz had the large model as an option, but they didn't allow to import audios.

Suggestion to export as .srt subtitle file

Hello, I would like to make a suggestion for your app, it would be interesting to add an option to export as an .srt file, this other repository offers an option to export srt https://github.com/ahmetoner/whisper-asr-webservice another thing would be nice to integrate cuda, for gpu acceleration for a faster inference, and as a last suggestion add an option of "dark theme" is everything, thank you.

Error opening InputStream: Invalid sample rate [PaErrorCode -9997]

Running Ubuntu 22.04.1. Trying to run with default HDA intel Mic results in the following error. (Options "pulse" and "default" work on the surface but appear silent).

user@hp-laptop:~/Downloads/buzz$ poetry run python main.py
[2022-10-01 15:28:39,667] transcriber.start_recording:42 DEBUG -> Recording... language: "en", model: "tiny", task: "Task.TRANSCRIBE", device: "5", block duration: "10"
Expression 'paInvalidSampleRate' failed in 'src/hostapi/alsa/pa_linux_alsa.c', line: 2048
Expression 'PaAlsaStreamComponent_InitialConfigure( &self->capture, inParams, self->primeBuffers, hwParamsCapture, &realSr )' failed in 'src/hostapi/alsa/pa_linux_alsa.c', line: 2718
Expression 'PaAlsaStream_Configure( stream, inputParameters, outputParameters, sampleRate, framesPerBuffer, &inputLatency, &outputLatency, &hostBufferSizeMode )' failed in 'src/hostapi/alsa/pa_linux_alsa.c', line: 2842
Traceback (most recent call last):
  File "/home/user/Downloads/buzz/gui.py", line 266, in on_status_changed
    self.start_recording()
  File "/home/user/Downloads/buzz/gui.py", line 292, in start_recording
    self.transcriber.start_recording(
  File "/home/user/Downloads/buzz/gui.py", line 182, in start_recording
    self.transcriber.start_recording(
  File "/home/user/Downloads/buzz/transcriber.py", line 44, in start_recording
    self.current_stream = sounddevice.InputStream(
  File "/home/user/.cache/pypoetry/virtualenvs/buzz-DqnAh-gc-py3.10/lib/python3.10/site-packages/sounddevice.py", line 1421, in __init__
    _StreamBase.__init__(self, kind='input', wrap_callback='array',
  File "/home/user/.cache/pypoetry/virtualenvs/buzz-DqnAh-gc-py3.10/lib/python3.10/site-packages/sounddevice.py", line 898, in __init__
    _check(_lib.Pa_OpenStream(self._ptr, iparameters, oparameters,
  File "/home/user/.cache/pypoetry/virtualenvs/buzz-DqnAh-gc-py3.10/lib/python3.10/site-packages/sounddevice.py", line 2747, in _check
    raise PortAudioError(errormsg, err)
sounddevice.PortAudioError: Error opening InputStream: Invalid sample rate [PaErrorCode -9997]
Aborted (core dumped)

Save last settings for next time

Keeps closing when I click Record on Windows 10

Doesn't really give any Error message or anything. Just sort of stalls for 10 seconds and then the Buzz window closes.
Any idea what the issue might be?
I have whisper already installed in a virtual environment. Is that the issue?

Add Windows installer

Inno Setup on GitHub Actions: https://jrsoftware.org/ishelp/index.php?topic=compilercmdline

https://github.com/actions/runner-images/blob/main/images/win/Windows2022-Readme.md#tools

Microphone recording no sound when opened from Finder

When the app is opened from Finder on Mac (via double-clicking the app from Finder or opening from Spotlight Search), the recording stream returns arrays of all zeroes. But it works fine when the program is opened from the terminal (via python main.py or ./dist/Buzz.app/Contents/MacOS/Buzz).

This seems related: https://stackoverflow.com/questions/69193993/pyinstaller-app-not-getting-microphone-permission-works-fine-with-terminal

Sort list of languages in alphabetical order

#11

Update docs with import audio feature

Fix dark/light mode styling for Mac

Non-english transcription seems to be not working

Trying to transcribe Russian sound results in English characters in the output. Am I doing anything wrong here?

Check sizing in smaller screens

Add speaker diarization

Add ffmpeg binary to build

Set application window title

#11