In almost everything related to graphics, OS X apps tend to be more flexible and more powerful than anything you can find in Windows, with one major exception: Optical-character-reading (OCR) software for Windows has always been more powerful than Mac-based OCR apps. Now that Abbyy's FineReader Pro ($99.99) has arrived for the Mac, it's still true that Windows has better OCR software, but that's only because Abbyy's Windows-based OCR app, Abbyy FineReader 11 Professional Edition, is even more powerful than its Mac-based OCR app. Abbyy's apps are now are our Editors' Choice OCR products for both Windows and OS X, but the OS X version hasn't caught up with all the features in the Windows version.
It's All About the OCR
Like all Abbyy's products, FineReader Pro uses the best OCR engine on the market. Whether the app reads pages from a scanner or from pictures or PDF files on your disk, it does a spectacularly good job of extracting text, arranging tables, and preserving layout.
FineReader Pro outputs editable documents in Word, Excel, PowerPoint, HTML, plain text, and e-book formats—and these documents typically require only a minimum of editing to correct any mistakes the app made in reading the original text or layout. It also outputs PDF files that display either clean text instead of the original scanned image of the text, or with searchable invisible text hidden under the original picture so you get the combined benefits of accurate appearance and searchable text.
Getting Started
You can use FineReader Pro either in its automated mode or in a mode that lets you adjust its settings at each stage of its operation. In automated mode, you simply choose an operation from the opening menu and let the program do its work—for example. On the left-hand side of the opening menu, you select a source—either your scanner or a file on your disk. On the right-hand side, you choose an automated operation, such as "Convert to Excel Spreadsheet." A gear icon next to the name of the automated operation lets you specify some basic output options, such as image quality and whether to use CSS styles in creating HTML pages. After a few seconds, the program prompts you for an output location and file name, and your output document is written to disk.
If you perform a manual operation, you first import pictures or scanned pages into a FineReader Pro document which you can modify for best results. An image editor lets you deskew images, erase or crop out parts of an image that you don't want, adjust perspective, color, and brightness and much else. The app then analyzes each page image to detect text, pictures, and tables. If, as sometimes happen with scans or pictures of printed material, the app misidentifies a smudge as a picture, or plain text as a table, use an Inspector panel to remove the smudge from the output or mark the text properly as text, not table. Then you tell the app to read the pages, and then, when you export the final document, you can fine-tune output options so that you preserve the original page layout or simply export text.
As in the Windows version, FineReader Pro's OCR engine is exceptionally accurate, producing flawless results from clean copy, and excellent results even from smudgy xeroxes of old books and typescripts. Its table-recognition engine is extraordinarily effective, even though it sometimes failed to detect thin border lines and I had to reapply the border lines when opening the output in Excel or Apple's Numbers app.
Not Mistake-Proof
Despite these exceptional abilities, FineReader Pro still doesn't equal its Windows counterpart, because it lacks the Windows' version proofreading feature. In the Windows version, you can open a small editing window and check every doubtful detail in the recognized text, deleting specks that were misrecognized as letters or numbers, removing italicization, and correcting spelling. This feature is completely absent from the Mac version, so you have to perform all corrections in the target app—Word, Excel, or anything else.
One downside of this lack is that you can't correct the output before exporting it as a PDF. If you want to create a PDF with corrected output, you'll need to export the document to Word or Excel, make corrections there, and then export the corrected document to PDF. When you export a file, a checkbox lets you tell the app to highlight characters that it's doubtful about so that you can find them quickly in the output document, but this means you'll have to remove the highlighting after making corrections. In the Windows version, you can do all this without exiting the OCR app.
The menu structure in FineReader Pro could also use some fine-tuning. When I performed step-by-step OCR, after importing pages from my scanner, I went to the Page menu and clicked "Read page," at the top of the menu, in order to convert the image into text. But the app gave me an error message saying that the page had not been processed yet. My mistake was that I should have clicked "Analyze page" before clicking "Read page." But "Analyze page" is lower down on the Page menu than "Read page" so I didn't realize that I should start with it.
The Best for Mac
Despite these minor first-version problems, Abbyy's FineReader Pro is the only serious choice for OCR on a Mac. FineReader Pros' OCR engine outclasses anything else available, and its overall design is clean, efficient, and fast. It's our no-contest Editors' Choice for OCR under OS X.