🎙️ Continuous Audio Recorder Setup Guide

This guide explains how to set up and run the continuous_audio_recorder.py script on macOS. The script continuously records audio until stopped (via Ctrl+C), saving files as 128kbps stereo MP3s in daily folders. Each filename includes the date, start time, stop time, timezone offset, and recording number.

---

🧰 Prerequisites

• macOS: Tested on macOS Ventura, Sonoma.
• Homebrew: macOS package manager. Install if not already present.
• Admin Access: Required for installations.
• Audio Hardware: Microphone and system audio for aggregate device setup.

---

⚙️ Setup Instructions

1. Install Homebrew

/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"

Add Homebrew to your shell path:

echo 'eval "$(/opt/homebrew/bin/brew shellenv)"' >> ~/.zshrc
source ~/.zshrc

Verify:

brew --version

---

2. Install Python 3.12

brew install python@3.12
python3.12 --version  # Should output Python 3.12.x

Note: Python path is typically /opt/homebrew/bin/python3.12.

---

3. Install FFmpeg

brew install ffmpeg
ffmpeg -version  # Verify installation

---

4. Create a Virtual Environment

/opt/homebrew/bin/python3.12 -m venv ~/codexaudio/venv
source ~/codexaudio/venv/bin/activate  # Activate the environment

---

5. Install PortAudio and Python Libraries

Install PortAudio (required for PyAudio):

brew install portaudio

6. Install Python Libraries

pip install pyaudio pydub

Verify:

pip show pyaudio
pip show pydub

• pyaudio: For audio recording
• pydub: For converting WAV to MP3

---

7. Set Up Aggregate Audio Device

1. Open Audio MIDI Setup via Spotlight.
2. Click ➕ and choose Create Aggregate Device.
3. Rename it to System + Mic.
4. Add:
   • Your microphone
   • System output (e.g., Built-in Output, Soundflower, or BlackHole)
5. Ensure:
   • At least 3 input channels
   • Sample rate: 44100 Hz

Note:
To capture system audio, install BlackHole:

brew install blackhole-2ch

Add BlackHole to your aggregate device via Audio MIDI Setup.

---

8. Configure the Script

1. Save continuous_audio_recorder.py to your project directory, e.g.:

mkdir -p ~/codexaudio
mv continuous_audio_recorder.py ~/codexaudio/

2. Edit the script:

base_dir = "/Users/viszya/audio_recordings"  # Update with your preferred path

3. Ensure the directory exists and is writable:

mkdir -p /Users/viszya/audio_recordings
chmod -R u+w /Users/viszya/audio_recordings

---

9. Run the Script

cd ~/codexaudio
source venv/bin/activate
python continuous_audio_recorder.py

The script will display:
Recording started. Press Ctrl+C to stop.

Audio files are saved in:
/Users/viszya/audio_recordings/YYYY-MM-DD/
e.g., 20250707_150000-0700_to_151000-0700_1.mp3

---

🧪 Troubleshooting

🔒 Read-Only File System Error

• Ensure base_dir is writable.
• Run:

  chmod -R u+w /Users/viszya/audio_recordings

🎧 Aggregate Device Not Found

• Recheck the "System + Mic" setup in Audio MIDI Setup.
• Confirm name and channel count in script.

❌ FFmpeg Errors

• Check FFmpeg accessibility:

  which ffmpeg

• If not found, reinstall or update your PATH.

🔇 No Audio Recorded

• Confirm mic/system audio are active and included in the aggregate device.
• Test input via Audio MIDI Setup.

💾 Memory Issues

• For long recordings, consider limiting session time to avoid RAM issues.

---

📁 File Output

• Location:
  base_dir/YYYY-MM-DD/
  e.g., /Users/viszya/audio_recordings/2025-07-07/

• Filename Format:
  YYYYMMDD_HHMMSS±HHMM_to_HHMMSS±HHMM_N.mp3
  Example: 20250707_150000-0700_to_151000-0700_1.mp3

• Format:
  128kbps stereo MP3, converted from a temporary WAV file.

---

✅ Requirements Summary

Category	Requirement
Software	Homebrew, Python 3.12, FFmpeg, BlackHole (optional)
Python Packages	pyaudio, pydub
Hardware	Microphone, optional loopback for system audio
macOS Config	Aggregate device: "System + Mic", ≥3 input channels

---

For help, contact the script maintainer or consult Python and FFmpeg documentation.

MP3 Transcription Script

This script automatically transcribes .mp3 files in a recordings folder using OpenAI Whisper and outputs the results as JSON files. It only processes files that do not already have a corresponding transcript.

Features

• Processes all .mp3 files in the recordings directory that do not have a corresponding _transcript.json file.
• Outputs only the transcript as a JSON file per MP3 (e.g., yourfile_transcript.json).
• Skips files that have already been transcribed.
• Shows a progress bar and live elapsed time for each file.
• Estimates processing time for new files based on previous runs (if .transcribe_stats.json is present).
• Optionally displays MP3 duration if mutagen is installed.
• Exits automatically when done.

Setup Instructions

1. Install Python 3.8+

• Download and install Python from python.org if you don't already have it.
• On macOS, the official installer is recommended.

2. Clone or Download This Repository

git clone <your-repo-url>
cd <your-repo-directory>

3. Create the recordings Folder

mkdir recordings

Place your .mp3 files inside this folder.

4. Install Required Python Packages

It is recommended to use a virtual environment:

python3 -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

Then install dependencies:

pip install openai-whisper tqdm

(Optional) For MP3 duration display:

pip install mutagen

5. Run the Script

python3 Transcribe.py

• The script will process all .mp3 files in the recordings folder that do not already have a _transcript.json file.
• Transcripts will be saved as <originalfilename>_transcript.json in the same folder.
• The script will estimate processing time for new files based on previous runs if .transcribe_stats.json is present (created automatically after processing qualifying files).

Notes

• The first run will download the Whisper model (can be several GBs, requires internet connection).
• For best results, use short, clear audio files.
• If you add new .mp3 files later, just re-run the script.
• The .transcribe_stats.json file is used for time estimation and is created/updated automatically.

Troubleshooting

• If you see SSL certificate errors on macOS, run the Install Certificates.command script in your /Applications/Python 3.x/ folder.
• If you see missing package errors, ensure you are using the correct Python environment and have installed all dependencies.
• If you want MP3 duration display, install mutagen as shown above.

License

MIT