IronOCR on Windows: OCR Processing Hangs on Large PDFs When Using EnglishFast as a Secondary Language
Overview
IronOCR can silently stall mid-processing on large PDFs when OcrLanguage.EnglishFast is added as a secondary language. No exception is thrown and execution does not resume. Removing EnglishFast from the secondary language list resolves the hang.
Environment
- OS: Windows
Application type: Web Application (.NET 8 MVC)
Language: C# / .NET 8
Version Metadata
- Version Found: 2026.3.3
- Version Resolved: N/A
Steps
1. Locate the secondary language configuration
Find any call to AddSecondaryLanguage that passes OcrLanguage.EnglishFast in your OCR setup.
// This line causes the hang on large PDFs — remove it ocr.AddSecondaryLanguage(OcrLanguage.EnglishFast);
2. Remove EnglishFast from the secondary language list
Comment out or delete the offending call. Keep OcrLanguage.English as the primary language if English coverage is required. Other secondary languages are unaffected.
//ocr.AddSecondaryLanguage(OcrLanguage.EnglishFast);
var ocr = new IronTesseract
{
Language = OcrLanguage.English
};
ocr.AddSecondaryLanguage(OcrLanguage.Hindi);
ocr.AddSecondaryLanguage(OcrLanguage.Marathi);
3. Re-run processing
After removing EnglishFast, rerun the affected PDF. Processing should complete through all pages without stalling.
The hang has been reproduced in IronOCR versions 2025.12.3, 2026.1.2, 2026.2.1, and 2026.3.3 through 2026.7.2.
Until a confirmed fixed version is identified, we should assume that all versions within this range are affected.
Notes and Limitations
- No exception is thrown and no error is logged when the hang occurs — execution simply stops at the
ocr.Read()call. - Reordering
EnglishFastrelative to other secondary languages does not resolve the hang. - Removing
EnglishFastrestores processing but removes the fast English model. If your documents contain significant English text, useOcrLanguage.Englishas the primary language (as shown in Step 2) to maintain English recognition quality.