nitrox 0.1.1

High-performance Python library for slicing and converting audio/video files with zero bloat

🫧 Nitrox

Deep, fast, clean. Media processing at maximum depth.

Nitrox is a Python library for high performance audio and video processing. Built on ffmpeg with a clean, intuitive API.

Installation

pip install nitrox

With audio processing:

pip install nitrox[audio]

FFmpeg Requirement

Nitrox requires ffmpeg to be installed on your system.

macOS:

# Using Homebrew
brew install ffmpeg

Linux (Ubuntu/Debian):

sudo apt update
sudo apt install ffmpeg

Linux (RHEL/CentOS):

sudo dnf install ffmpeg
# or
sudo yum install ffmpeg

Other platforms: Download from ffmpeg.org

Quick Start

from nitrox import Media

# Basic usage
Media("input.mp4").slice(10, 30).to("output.mp4")

# Audio extraction
Media("video.mp4").extract_audio().to("audio.wav")

# Video processing
Media("input.mp4").resize(720, 480).to("resized.mp4")

Minimal Syntax

For power users, Nitrox offers an ultra-minimal X alias:

from nitrox import X

# Same functionality, minimal syntax
X("input.mp4")[10:30].to("output.mp4")
X("video.mov").extract_audio().to("audio.mp3")

Core Features

Intuitive Slicing

media = Media("video.mp4")

# Time-based slicing
media[0:10]              # First 10 seconds
media[30:60]             # From 30s to 60s
media[10:]               # From 10s to end
media[:30]               # First 30 seconds

Chainable Operations

# Video processing
Media("input.mp4") \
    .slice(10, 60) \
    .resize(1280, 720) \
    .to("processed.mp4")

# Audio processing
Media("audio.wav") \
    .normalize_audio() \
    .fade_in(1.0) \
    .fade_out(2.0) \
    .to("clean.wav")

Audio Processing

# Basic audio operations
Media("input.mp3").resample(44100).to("output.wav")
Media("stereo.wav").to_mono().to("mono.wav")

# With numpy integration
audio_data = Media("song.wav")[10:20].to_numpy()
Media.from_numpy(audio_data, 44100).to("processed.wav")

Examples

Extract Highlights

# Extract multiple clips
video = Media("long_video.mp4")
highlights = [
    video[0:30],      # Opening
    video[300:330],   # Middle segment  
    video[1200:1230]  # Ending
]

for i, clip in enumerate(highlights):
    clip.to(f"highlight_{i}.mp4")

Podcast Processing

# Clean up podcast audio
Media("raw_recording.wav") \
    .normalize_audio() \
    .fade_in(0.5) \
    .fade_out(1.0) \
    .to("podcast_ready.mp3")

Batch Processing

# Process multiple files
import os

for filename in os.listdir("raw_videos/"):
    if filename.endswith('.mp4'):
        Media(f"raw_videos/{filename}") \
            .resize(720, 480) \
            .to(f"processed/{filename}")

API Reference

Core Operations

Video:

• resize(width, height) - Resize video
• crop(x, y, width, height) - Crop video
• set_fps(fps) - Change frame rate

Audio:

• extract_audio() - Extract audio track
• normalize_audio() - Normalize audio levels
• resample(rate) - Change sample rate
• to_mono() - Convert to mono
• fade_in(duration) - Add fade in
• fade_out(duration) - Add fade out

Slicing:

• slice(start, end) - Extract time range
• [start:end] - Python slice notation

Output:

• to(path) - Save processed media
• to_numpy() - Extract as numpy array (requires numpy)
• from_numpy(data, rate) - Create from numpy array (requires numpy)

Utility:

• info() - Get media information
• preview_command() - Show ffmpeg command

Command Line

# Basic usage
nitrox slice input.mp4 output.mp4 --start 10 --end 30
nitrox resize video.mp4 small.mp4 --width 720 --height 480
nitrox convert video.mp4 audio.mp3 --extract-audio

# Get media info
nitrox info video.mp4

Requirements

• Python 3.8+
• ffmpeg (see installation instructions above)
• numpy (optional, for array processing)

License

MIT License - see LICENSE file.

---

**🫧 Nitrox **