Features Overview

Complete guide to Ozen-web capabilities

Introduction

Ozen-web is a browser-based tool for acoustic analysis and annotation. This page provides an overview of all major features with links to detailed documentation.

Core Features

Audio Loading & Recording

Load audio from multiple sources:

Drag & drop audio files (WAV, FLAC, MP3, OGG)
File picker dialog
Microphone recording with real-time visualization
Remote URL loading (CORS-enabled)
Data URL embedding for self-contained iframes

See: Getting Started Guide

Visualization

Waveform Display:

Amplitude visualization
Synchronized with spectrogram
Responsive to zoom and pan
Downsampling for long file processing (until zoom)

Spectrogram Display:

Praat-style grayscale spectrogram
Adjustable frequency range (5/7.5/10 kHz)
Dynamic resolution enhancement when zoomed
Cached rendering for smooth performance

See: Spectrogram Documentation | Waveform Documentation

Acoustic Analysis

Ozen-web provides acoustic measurements using Praat-compatible algorithms:

Available overlays:

Feature	Description
Pitch (F0)	Fundamental frequency tracking
Formants (F1-F4)	Resonance frequencies
Intensity	Sound pressure level
HNR	Harmonics-to-Noise Ratio
CoG	Spectral Center of Gravity
Spectral Tilt	High/low frequency balance
A1-P0	Harmonic amplitude measure

All measurements are computed using praatfan WebAssembly.

See: Acoustic Overlays Documentation | Tutorial: Acoustic Analysis

Annotations

Multi-tier TextGrid-compatible annotation system:

Create unlimited annotation tiers
Add, move, and remove boundaries
Edit interval labels with keyboard shortcuts
Boundary snapping between tiers
Full Praat TextGrid import/export (both short and long formats)
Interval and point tier support

Efficient workflow:

Double-click to add boundaries
Drag to adjust timing
Right-click context menus
Keyboard shortcuts for tier selection (1-5 keys)
Unified undo/redo (Ctrl+Z / Ctrl+Y)

See: Annotations Documentation | Tutorial: Annotations

Data Points

Point-and-click acoustic measurement collection:

Data points allow collecting comprehensive acoustic measurements at specific time/frequency locations:

Features:

Double-click on spectrogram to add points
Automatic capture of all acoustic values (F0, F1-F4, intensity, HNR, etc.)
Inherits labels from ALL annotation tiers at that time
Drag to adjust position
Export to TSV for statistical analysis
Import TSV to restore data points

Export format (TSV):

time    freq    pitch   f1  f2  f3  f4  intensity   phones  words
0.234   1500    156 678 1234    2890    3654    72  i   see
0.567   1200    234 543 987 2567    3234    68  u   two

Perfect for building datasets for R, Python, SPSS, or Excel analysis.

See: Data Points Documentation | Tutorial: Data Collection

Audio Playback

Flexible playback controls:

Play selected region (Space key)
Play visible window (Tab key)
Play from cursor to end
Real-time cursor tracking during playback
Pause and resume
Stop and deselect (Escape key)

See: Audio Playback Documentation

Mobile Viewer

Touch-optimized interface at /viewer route:

Touch gestures:
- Tap → Place cursor
- Drag → Select region
- Two-finger drag → Pan view
- Pinch → Zoom in/out
- Long-press → Copy selected position visible properties
Compact layout for small screens
Landscape mode optimization
Safe area support for notched phones
Settings drawer for overlay toggles
Floating play button
URL parameters for pre-configuration

Embedding support:

Iframe-compatible
Pre-load audio via ?audio= parameter
Configure overlays via ?overlays= parameter
CORS-enabled remote loading
Data URL support for self-contained embeds

See: Mobile Viewer Documentation | Embedding Guide

Advanced Features

Long Audio Handling

For files >60 seconds:

Ozen-web uses an intelligent on-demand analysis strategy to prevent UI freezing:

On load: Waveform displays immediately, spectrogram shows “Zoom in for spectrogram”
When zoomed: Analysis runs for visible window only (≤60s)
On-demand: Results update automatically as you pan/zoom
Debounced: 300ms delay prevents excessive recomputation

This allows working with arbitrarily long recordings (hours of audio) while maintaining responsive UI.

Why 60 seconds?

Browser memory constraints
Real-time user experience (analysis must complete in <1s)
Balance between convenience and performance

See: Getting Started: Troubleshooting

Multiple WASM Backends

Choose from multiple acoustic analysis backends:

Backend	Source	License	Offline
`praatfan-local`	Bundled (`static/wasm/`)	MIT/Apache-2.0	✅ Yes
`praatfan`	GitHub Pages CDN	MIT/Apache-2.0	❌ No
`praatfan-gpl`	GitHub Pages CDN	GPL	❌ No

Praatfan-local and praatfan are identical, and approximate praatfan-gpl, which was built to be very similar to Praat. Select via dropdown in the app interface.

See: WASM Backends Reference

Configuration

Customize via config.yaml:

# Example configuration
backend: "praatfan-local"
maxFrequency: 5000
showPitch: true
showFormants: true
theme: "dark"

# Pitch settings
pitchFloor: 75
pitchCeiling: 600

# Formant settings
maxFormants: 5
maxFormantFrequency: 5500

# Colors
pitchColor: "#0000FF"
formantColor: "#FF0000"

See: Configuration Reference

Unified Undo/Redo

Single undo stack for all editable operations:

Adding/removing/moving boundaries
Editing interval labels
Adding/removing/moving data points
Chronological order across all operation types

Keyboard shortcuts:

Ctrl+Z (Cmd+Z on Mac) — Undo
Ctrl+Y or Ctrl+Shift+Z — Redo

Non-undoable operations (by design):

Adding/removing tiers
Loading audio/TextGrid files

See: Tutorial: Annotations

Platform Features

Offline & PWA Support

Works completely offline:

All processing happens locally in browser
No data uploaded to servers
After initial load, works without internet (with local backend)
PWA-ready with app icons for home screen installation

Browser Compatibility

Should work as well on all modern browsers

Requirements:

WebAssembly support
Web Audio API
Canvas API
ES6 modules

Privacy

100% local processing:

Audio never leaves your browser
No analytics or tracking
No server-side processing

File Format Support

Import Formats

Format	Extension	Notes
Audio	`.wav`, `.flac`, `.mp3`, `.ogg`	Could be browser-dependent
TextGrid	`.TextGrid`	Both short and long formats
Data Points	`.tsv`	Tab-separated values
Configuration	`.yaml`	Optional app configuration

Export Formats

Format	Content	Use Case
TextGrid	Annotations (all tiers)	Praat, R, Python
TSV	Data points + acoustic values	Statistical analysis
WAV	Audio (16-bit PCM)	Save recordings, share clips

See: File Formats Reference

Use Cases

Research:

Vowel formant analysis (F1/F2 plots)
Pitch contour studies (intonation, tone)
Voice quality assessment (HNR, spectral tilt)
Consonant acoustics (VOT, CoG, duration)

Teaching:

Embed interactive spectrograms in lecture slides
Quarto/R Markdown integration
No software installation required for students
Interactive demonstrations

Next Steps

New users:

Getting Started Guide — Installation and first use
Complete Tutorial — 30-minute guided walkthrough

Specific features:

Spectrogram — Visualization details
Annotations — TextGrid workflow
Data Points — Measurement collection
Mobile Viewer — Touch interface

Reference:

Keyboard Shortcuts — All shortcuts
Configuration — Customization options
WASM Backends — Backend comparison

Development:

Architecture — Technical design
Contributing — How to contribute

--- title: "Features Overview" subtitle: "Complete guide to Ozen-web capabilities" --- ## Introduction Ozen-web is a browser-based tool for acoustic analysis and annotation. This page provides an overview of all major features with links to detailed documentation. ## Core Features ### Audio Loading & Recording **Load audio from multiple sources:** - Drag & drop audio files (WAV, FLAC, MP3, OGG) - File picker dialog - Microphone recording with real-time visualization - Remote URL loading (CORS-enabled) - Data URL embedding for self-contained iframes **See:** [Getting Started Guide](../getting-started.html#loading-your-first-audio-file) ### Visualization **Waveform Display:** - Amplitude visualization - Synchronized with spectrogram - Responsive to zoom and pan - Downsampling for long file processing (until zoom) **Spectrogram Display:** - Praat-style grayscale spectrogram - Adjustable frequency range (5/7.5/10 kHz) - Dynamic resolution enhancement when zoomed - Cached rendering for smooth performance **See:** [Spectrogram Documentation](spectrogram.html) | [Waveform Documentation](waveform.html) ![Main interface overview](../screenshots/main-interface-overview.png){.screenshot} ### Acoustic Analysis Ozen-web provides acoustic measurements using Praat-compatible algorithms: **Available overlays:** | Feature | Description | |---------|-------------| | **Pitch (F0)** | Fundamental frequency tracking | | **Formants (F1-F4)** | Resonance frequencies | | **Intensity** | Sound pressure level | | **HNR** | Harmonics-to-Noise Ratio | | **CoG** | Spectral Center of Gravity | | **Spectral Tilt** | High/low frequency balance | | **A1-P0** | Harmonic amplitude measure | All measurements are computed using **praatfan** WebAssembly. **See:** [Acoustic Overlays Documentation](acoustic-overlays.html) | [Tutorial: Acoustic Analysis](../tutorial/03-acoustic-analysis.html) ![Spectrogram with all acoustic overlays](../screenshots/acoustic-overlays-all.png){.screenshot} ### Annotations **Multi-tier TextGrid-compatible annotation system:** - Create unlimited annotation tiers - Add, move, and remove boundaries - Edit interval labels with keyboard shortcuts - Boundary snapping between tiers - Full Praat TextGrid import/export (both short and long formats) - Interval and point tier support **Efficient workflow:** - Double-click to add boundaries - Drag to adjust timing - Right-click context menus - Keyboard shortcuts for tier selection (1-5 keys) - Unified undo/redo (Ctrl+Z / Ctrl+Y) **See:** [Annotations Documentation](annotations.html) | [Tutorial: Annotations](../tutorial/04-annotations.html) ![Multi-tier annotation example](../screenshots/annotations-multitiered.png){.screenshot} ### Data Points **Point-and-click acoustic measurement collection:** Data points allow collecting comprehensive acoustic measurements at specific time/frequency locations: **Features:** - Double-click on spectrogram to add points - Automatic capture of all acoustic values (F0, F1-F4, intensity, HNR, etc.) - Inherits labels from ALL annotation tiers at that time - Drag to adjust position - Export to TSV for statistical analysis - Import TSV to restore data points **Export format (TSV):** ``` time freq pitch f1 f2 f3 f4 intensity phones words 0.234 1500 156 678 1234 2890 3654 72 i see 0.567 1200 234 543 987 2567 3234 68 u two ``` Perfect for building datasets for R, Python, SPSS, or Excel analysis. **See:** [Data Points Documentation](data-points.html) | [Tutorial: Data Collection](../tutorial/05-data-collection.html) ![Data points on spectrogram](../screenshots/data-points-collection.png){.screenshot} ### Audio Playback **Flexible playback controls:** - Play selected region (Space key) - Play visible window (Tab key) - Play from cursor to end - Real-time cursor tracking during playback - Pause and resume - Stop and deselect (Escape key) **See:** [Audio Playback Documentation](audio-playback.html) ### Mobile Viewer **Touch-optimized interface at `/viewer` route:** - **Touch gestures:** - Tap → Place cursor - Drag → Select region - Two-finger drag → Pan view - Pinch → Zoom in/out - Long-press → Copy selected position visible properties - **Compact layout** for small screens - **Landscape mode optimization** - **Safe area support** for notched phones - **Settings drawer** for overlay toggles - **Floating play button** - **URL parameters** for pre-configuration **Embedding support:** - Iframe-compatible - Pre-load audio via `?audio=` parameter - Configure overlays via `?overlays=` parameter - CORS-enabled remote loading - Data URL support for self-contained embeds **See:** [Mobile Viewer Documentation](mobile-viewer.html) | [Embedding Guide](../embedding/overview.html) ![Mobile viewer in landscape mode](../screenshots/mobile-viewer-landscape.png){.screenshot} ## Advanced Features ### Long Audio Handling {#long-audio-handling} **For files >60 seconds:** Ozen-web uses an intelligent on-demand analysis strategy to prevent UI freezing: 1. **On load**: Waveform displays immediately, spectrogram shows "Zoom in for spectrogram" 2. **When zoomed**: Analysis runs for visible window only (≤60s) 3. **On-demand**: Results update automatically as you pan/zoom 4. **Debounced**: 300ms delay prevents excessive recomputation This allows working with arbitrarily long recordings (hours of audio) while maintaining responsive UI. **Why 60 seconds?** - Browser memory constraints - Real-time user experience (analysis must complete in <1s) - Balance between convenience and performance **See:** [Getting Started: Troubleshooting](../getting-started.html#long-audio-files-show-zoom-in-for-spectrogram) ### Multiple WASM Backends Choose from multiple acoustic analysis backends: | Backend | Source | License | Offline | |---------|--------|---------|---------| | `praatfan-local` | Bundled (`static/wasm/`) | MIT/Apache-2.0 | ✅ Yes | | `praatfan` | GitHub Pages CDN | MIT/Apache-2.0 | ❌ No | | `praatfan-gpl` | GitHub Pages CDN | GPL | ❌ No | Praatfan-local and praatfan are identical, and approximate praatfan-gpl, which was built to be very similar to Praat. Select via dropdown in the app interface. **See:** [WASM Backends Reference](../reference/backends.html) ### Configuration **Customize via `config.yaml`:** ```yaml # Example configuration backend: "praatfan-local" maxFrequency: 5000 showPitch: true showFormants: true theme: "dark" # Pitch settings pitchFloor: 75 pitchCeiling: 600 # Formant settings maxFormants: 5 maxFormantFrequency: 5500 # Colors pitchColor: "#0000FF" formantColor: "#FF0000" ``` **See:** [Configuration Reference](../reference/configuration.html) ### Unified Undo/Redo **Single undo stack for all editable operations:** - Adding/removing/moving boundaries - Editing interval labels - Adding/removing/moving data points - Chronological order across all operation types **Keyboard shortcuts:** - <kbd>Ctrl+Z</kbd> (Cmd+Z on Mac) — Undo - <kbd>Ctrl+Y</kbd> or <kbd>Ctrl+Shift+Z</kbd> — Redo **Non-undoable operations** (by design): - Adding/removing tiers - Loading audio/TextGrid files **See:** [Tutorial: Annotations](../tutorial/04-annotations.html#undo-and-redo) ## Platform Features ### Offline & PWA Support **Works completely offline:** - All processing happens locally in browser - No data uploaded to servers - After initial load, works without internet (with local backend) - PWA-ready with app icons for home screen installation ### Browser Compatibility Should work as well on all modern browsers **Requirements:** - WebAssembly support - Web Audio API - Canvas API - ES6 modules ### Privacy **100% local processing:** - Audio never leaves your browser - No analytics or tracking - No server-side processing ## File Format Support ### Import Formats | Format | Extension | Notes | |--------|-----------|-------| | Audio | `.wav`, `.flac`, `.mp3`, `.ogg` | Could be browser-dependent | | TextGrid | `.TextGrid` | Both short and long formats | | Data Points | `.tsv` | Tab-separated values | | Configuration | `.yaml` | Optional app configuration | ### Export Formats | Format | Content | Use Case | |--------|---------|----------| | TextGrid | Annotations (all tiers) | Praat, R, Python | | TSV | Data points + acoustic values | Statistical analysis | | WAV | Audio (16-bit PCM) | Save recordings, share clips | **See:** [File Formats Reference](../reference/file-formats.html) ## Use Cases **Research:** - Vowel formant analysis (F1/F2 plots) - Pitch contour studies (intonation, tone) - Voice quality assessment (HNR, spectral tilt) - Consonant acoustics (VOT, CoG, duration) **Teaching:** - Embed interactive spectrograms in lecture slides - Quarto/R Markdown integration - No software installation required for students - Interactive demonstrations ## Next Steps **New users:** - [Getting Started Guide](../getting-started.html) — Installation and first use - [Complete Tutorial](../tutorial/index.html) — 30-minute guided walkthrough **Specific features:** - [Spectrogram](spectrogram.html) — Visualization details - [Annotations](annotations.html) — TextGrid workflow - [Data Points](data-points.html) — Measurement collection - [Mobile Viewer](mobile-viewer.html) — Touch interface **Reference:** - [Keyboard Shortcuts](../reference/keyboard-shortcuts.html) — All shortcuts - [Configuration](../reference/configuration.html) — Customization options - [WASM Backends](../reference/backends.html) — Backend comparison **Development:** - [Architecture](../development/architecture.html) — Technical design - [Contributing](../development/contributing.html) — How to contribute