v2.3 Multilingual Demo Videos with Background Music

v2.3 Multilingual Demo Videos with Background Music

All videos now include background music: "Happy Upbeat Ukulele" by MaxKoMusic

Music Attribution

Happy Upbeat Ukulele by MaxKoMusic | https://maxkomusic.com/
Music promoted by https://www.chosic.com/free-music/all/
Creative Commons Attribution-ShareAlike 3.0 Unported (CC BY-SA 3.0)

Videos

• 🇺🇸 English: demo-video-en.mp4 (4.7 MB, 40s)
• 🇹🇼 繁體中文: zh-TW-demo.mp4 (4.5 MB, 40s)
• 🇨🇳 简体中文: zh-CN-demo.mp4 (4.5 MB, 40s)
• 🇯🇵 日本語: ja-demo.mp4 (4.7 MB, 40s)
• 🇪🇸 Español: es-demo.mp4 (4.7 MB, 40s)
• 🇫🇷 Français: fr-demo.mp4 (4.9 MB, 40s)

Watch Online

https://laihenyi.github.io/NBLM2PPTX/demo.html

v2.3 Multilingual Demo Videos

NBLM2PPTX v2.3 Multilingual Demo Videos

All 6 language versions of the 40-second product demonstration.

Videos

Language	File	Size
🇺🇸 English	demo-video-en.mp4	4.5 MB
🇹🇼 繁體中文	zh-TW-demo.mp4	4.3 MB
🇨🇳 简体中文	zh-CN-demo.mp4	4.3 MB
🇯🇵 日本語	ja-demo.mp4	4.5 MB
🇪🇸 Español	es-demo.mp4	4.6 MB
🇫🇷 Français	fr-demo.mp4	4.7 MB

Watch Online

Demo Video Player

Features Shown

• Dual Mode OCR System (Lite vs Standard)
• Parallel Processing Technology
• Multi-Language Support (6 languages)
• 4-Step Workflow

v2.3 Demo Video

NBLM2PPTX v2.3 Demo Video

40-second product demonstration showcasing:

• Dual Mode OCR System (Lite vs Standard)
• Parallel Processing Technology
• Multi-Language Support
• 4-Step Workflow

Language: English
Duration: 40 seconds
Size: 4.7 MB

---

Download

Download demo-video-en.mp4

v2.3: Dual-Mode OCR - Time Saver Edition

🚀 v2.3: Dual-Mode OCR System - Perfect for Time Saver Fans

⚡ Why Lite Mode is a Game Changer

For users who value speed and efficiency, the new Lite model delivers:

• ⏱️ 40-50% faster processing - from 3-4s to 2-3s per page
• 💰 50% API quota savings - process twice as many documents
• 🎯 Perfect for text-focused work - meeting notes, study materials, content drafts

🆕 What's New

Dual-Mode OCR System

• Lite Model (Recommended): Uses gemini-2.5-flash-lite - blazing fast, quota-efficient
• Standard Model: Uses gemini-2.5-flash - preserves font sizes and styling
• User choice: Switch modes in the page selection phase

Parallel Processing Architecture

• Text removal and OCR run simultaneously instead of sequentially
• Smart error handling ensures one API failure doesn't break the entire process
• Overall speed improvement: 40-50% faster than v2.2

Visual Comparison

See the difference between Lite and Standard models in our updated README with side-by-side examples.

📊 Model Comparison

Feature	Lite Model	Standard Model
Speed	2-3s per page	3-4s per page
API Quota	50% less	Standard consumption
Font Size Detection	❌ Uniform styling	✅ Preserved hierarchy
Font Weight Detection	❌ Not available	✅ Bold/normal detected
Color Detection	❌ Not available	✅ Original colors
Best For	Text content, speed priority	Visual presentations, design work

💡 Our Recommendation

Start with Lite Model - it's the default for a reason! Most users find uniform text styling perfectly acceptable for:

• 📝 Study notes and research materials
• 💼 Meeting minutes and work documents
• ✍️ Content drafts and brainstorming
• 📚 Quick documentation extraction

Switch to Standard only when you need:

• 🎨 Brand presentations with specific typography
• 📊 Educational slides with visual hierarchy
• 🏢 Client-facing materials with design precision

🌍 Multi-Language Support

All features available in:

• 🇹🇼 繁體中文 (Traditional Chinese)
• 🇺🇸 English
• 🇪🇸 Español (Spanish)
• 🇯🇵 日本語 (Japanese)
• 🇫🇷 Français (French)
• 🇨🇳 简体中文 (Simplified Chinese)

🔗 Quick Start

1. Open your preferred language version (index.html, index-en.html, etc.)
2. Upload your NotebookLM PDF
3. Select pages to process
4. Choose Lite Model (default) for speed, or Standard for style preservation
5. Export your editable PPTX in seconds!

---

Time is precious. Choose Lite. Get more done. ⚡

Full changelog and technical details in the commit history.

v2.2.1 - Complete i18n Overhaul

🌍 Complete i18n Overhaul

This release brings a comprehensive redesign of all language versions with professional light theme, improved UX, and standardized documentation.

Major Changes

Professional Design Across All Languages

• Completely redesigned all language versions (English, Spanish, Japanese, French, Simplified Chinese)
• Migrated from dark theme to modern light theme (#F8FAFC background)
• Implemented Poppins (headings) + Open Sans (body) font system with language-specific fallbacks
• Consistent professional blue color scheme (#3B82F6) across all versions

Enhanced User Experience

• Browser-Based API Key Storage: New modal with localStorage integration - no more code editing required
• Collapsible UI Elements: Alert banner and tools section can now be collapsed for cleaner interface
• Professional SVG Icons: Replaced emoji icons with proper SVG graphics
• Responsive Typography: Optimized font sizes and spacing for all screen sizes

Standardized Documentation

• All language README files now follow comprehensive 204-line structure
• Added Quick Start guide (3 simple steps) for better onboarding
• Clear documentation of Google Gemini's free tier (15 RPM, 1500 RPD, no credit card)
• Complete FAQ section with 5 Q&A pairs covering common questions

Design System

• Light Theme: #F8FAFC background for better readability
• Modern Card Layout: Clean borders (#E2E8F0) and subtle shadows
• Color System: Professional blue (#3B82F6) primary color with consistent accent colors
• Font System: Poppins + Open Sans with Noto Sans fallbacks for CJK languages

Language Versions

All following versions have been updated to v2.2.1:

• index.html - Traditional Chinese (繁體中文)
• index-en.html - English
• index-es.html - Spanish (Español)
• index-ja.html - Japanese (日本語)
• index-fr.html - French (Français)
• index-zh-CN.html - Simplified Chinese (简体中文)

Documentation

All README files have been updated:

• README.md - English
• README-zh-TW.md - Traditional Chinese
• README-zh-CN.md - Simplified Chinese
• README-ja.md - Japanese
• README-es.md - Spanish
• README-fr.md - French

---

Note: For information about v2.2 features (Soft Reset, Speed Optimization, IMAGE_RECITATION fixes), please see the release notes in the README files.

v2.2 - Soft Reset, Speed Optimization & IMAGE_RECITATION Fix

🚀 NBLM2PPTX v2.2 Release

Major update with user experience improvements, speed optimization, and critical bug fixes!

✨ What's New

🎯 Soft Reset with API Key Persistence

• No More Re-entering: API Key is now preserved in memory when you click "Restart"
• Unlimited Restarts: Process multiple batches without re-entering your credentials
• Smart State Management: Resets all processing state while keeping your API Key safe

Technical Implementation:

• Replaced location.reload() with softReset() function
• Prevents accidental fileInput change event triggers
• Restores clean initial UI instead of loading spinner

⚡ 70% Speed Improvement

• Faster Processing: Reduced inter-page delay from 3.5s to 1.0s
• Removed API Warm-up Delay: No more 2-second wait for 5+ page batches
• Parallel Processing: Fully optimized for maximum efficiency

Performance Gains:

• 10 pages: 35s → 10s (71% faster)
• 20 pages: 70s → 20s (71% faster)

🔧 IMAGE_RECITATION Error Fix

Google Gemini API was rejecting images due to copyright detection. We've fixed this with:

Enhanced AI Prompt:

• Before: "Remove all text and numbers. Reconstruct background. Return only image."
• After: "Edit this image by removing all visible text and numbers while preserving the original background patterns, colors, and design elements. Use content-aware fill to seamlessly reconstruct the areas where text was removed. Return the edited image."

API Optimizations:

• responseModalities: ['IMAGE'] (removed 'TEXT')
• temperature: 0.4 (more consistent results)
• Explicit IMAGE_RECITATION error detection with user-friendly messages

📝 UI/UX Improvements

• Clearer Instructions: Updated API Key setup guide to match actual workflow
• Removed Outdated References: Deleted mentions of removed "Auto-paste" feature
• Better Error Messages: More helpful feedback when things go wrong

🌍 Multi-Language Support

All 6 language versions updated:

• 🇹🇼 繁體中文
• 🇨🇳 简体中文
• 🇬🇧 English
• 🇯🇵 日本語
• 🇪🇸 Español
• 🇫🇷 Français

📊 Technical Details

Feature	v2.1	v2.2	Improvement
Inter-page delay	3500ms	1000ms	↓ 71%
Reset method	location.reload()	softReset()	Preserves API Key
IMAGE_RECITATION handling	❌	✅	Dedicated error handling
Prompt length	9 words	32 words	↑ 256%
Temperature	default (1.0)	0.4	↓ 60%
API warm-up delay	2000ms (5+ pages)	0ms	↓ 100%

🐛 Bug Fixes

• Fixed soft reset triggering unnecessary file processing
• Fixed UI stuck in "Reading and creating thumbnails..." after reset
• Fixed IMAGE_RECITATION errors causing JSON parsing failures
• Fixed misleading API Key setup instructions

📦 Files Changed

• 6 HTML files (all language versions)
• 6 README files (all language versions)
• 1,346 insertions, 100 deletions

🙏 Acknowledgments

Special thanks to the community for reporting issues and providing valuable feedback!

📥 How to Use

1. Download index.html (or your preferred language version)
2. Open in Chrome/Edge browser
3. Follow the 3-step API Key setup guide
4. Start processing your NotebookLM PDFs!

---

Full Changelog: v2.1...v2.2