Imagine you have a picture of a document or a scanned paper. OCR (Optical Character Recognition) is like a special computer program that "reads" that picture and turns it into text you can edit, search, and copy. It's a bit like teaching a computer to read every letter on every page to understand what it says. Because it's doing such detailed work, it takes some time. OCR is running in the background so you can continue working on your project. Once the OCR has finished you will be notified by a dialog. You can easily see the OCR status of each of your documents by clicking Menu (•••) → Show columns → OCR Status. Note that OCR recognizes horizontal text. If the scanned text is somewhat screwed it is not a problem, but vertical text in e.g., a technical drawing might not be recognized. Limits on file size and processing time? To make sure our service works smoothly for everyone, we have a few limits: How many files at once: You can process up to 500 files in a single group (we call this a "batch"). How big can each file be: Each individual file can be no larger than 200 Megabytes (MB). How big can all files be together: The total size of all files in one batch can't be more than 500 MB. How long does it take: The entire batch of files needs to finish processing within 60 minutes (1 hour). Why the File Size Limit? (200 MB per file) What is the limit? Really big files need a lot of computer memory and power to process. A 200 MB PDF, for example, could have hundreds of high-quality scanned pages that the computer needs to analyze. What if my file is too big? If your file is over 200 MB, our system won't be able to process it. This helps prevent crashes and ensures that everyone gets reliable service. Typical file sizes: Most PDF documents are smaller than 20 MB. A 200 MB file usually contains about 200 to 400 scanned pages at a high quality. Why the Batch Size Limit? (500 MB total per batch) Why this limit? Trying to process more than 500 MB of PDFs at once significantly increases the chance of the process taking too long and timing out. What if I submit too many files? If your selected files go over the 500 MB total, our OCR system will automatically reduce the number of documents chosen to stay within the limit. It will pick files to get as close to the 500 MB limit as possible. You can always change which files are selected, but you can't choose a total that's larger than the limit. Why the Processing Time Limit? (60 minutes per batch) Why this limit? Our cloud services have built-in time limits. This prevents any single process from running indefinitely and ensures that there are enough resources for all users. What makes processing take longer? Number of pages: More pages mean more time to process. Image quality: Higher quality images can take longer to process, but they usually give better results. Language complexity: Documents with multiple languages can take more time. File size: Larger files naturally take longer to download and process. What happens if processing takes too long? Should a batch approach its 60-minute time limit, the process will gracefully conclude, saving all documents processed up to that point. Any remaining documents will not be processed within that specific batch, but a new OCR job can be easily initiated for them. For large files, the process may stop quite some time before the 60-minute mark if it's determined that the next large file cannot be completed within the remaining time. Tips for smooth OCR processing: Prepare your PDFs before uploading: Lower image resolution: You don't always need super high resolution. Try to keep it at 200 DPI or less. If your original documents are good quality, you can go down to 150 DPI. Remove blank pages: Get rid of any empty pages in your PDF. Compress images: Use a tool to reduce the size of images within your PDF. Break up large documents: If you have a very large document, split it into smaller parts using Visiodocs' "Split document" function. Once the OCR is done, you can merge them back together using the "Merge" function. Consider processing different chapters or sections separately. Choosing One vs. mMany Languages when doing OCR In Visiodocs, you can specify the language(s) for text extraction from scanned documents. Selecting the correct language(s) is crucial; for example, if "English" is chosen for a German document, the OCR may misinterpret German words as English. This selection also impacts the speed of the OCR process; the more languages you select, the slower the OCR processing will be. Using One Language Advantages: Faster processing, since the OCR process only needs to compare letters and words from one language. Higher accuracy if the document truly contains only that language, because there’s less chance of confusion between similar-looking characters. Disadvantages: If the document has text in another language (for example, a few English words inside a Danish document), those parts may be recognized incorrectly or not at all. Using Many Languages Advantages: Better recognition if your document mixes languages (e.g., English and French in the same text). Reduces the risk of losing important words that don’t belong to the main language Disadvantages: Slower processing, since more possibilities need to be checked. Sometimes less accurate: when languages share similar letters, the OCR processor can get confused and pick the wrong one e.g. mixing up a, å and ä. In Short If you know the document is in a single language → choose that one language for best speed and accuracy. If the document clearly mixes languages → select multiple but expect it to be a bit slower and sometimes less precise.