Topaz was also used for long tail books, as there was a pipeline that could convert the high resolution search inside the book scans with minimal amount of huma

Topaz was also used for long tail books, as there was a pipeline that could convert the high resolution search inside the book scans with minimal amount of human work. At the time OCR wasn't an option, so it used glyph binning similar to JBIG. The OCR'd text was included for searching. The biggest downside was no font changes