Many people around the world use our sheet music scanner, and we frequently hear feedback that our lyric detection is too focused on English/Latin characters. Today we’re happy to announce a first step in fixing that: we now use custom OCR models for lyrics in 20 languages.

When you do a PDF/image scan, you can now specify which language the lyrics use:

Our system will use that to run appropriate OCR on the lyrics. This should result in higher-quality lyric conversion, using language-specific glyphs and diacritics.

At the moment, we support the following languages:

• Arabic
• Czech
• Chinese - Simplified
• Chinese - Traditional
• Dutch
• English
• French
• German
• Greek
• Hebrew
• Italian
• Japanese
• Korean
• Latin
• Polish
• Portuguese
• Russian
• Slovak
• Spanish
• Ukrainian