What happens when you scan
A scanner photographs a page and sends the image to your computer. At that stage the file is just a picture of the page, even if it shows text. To search it, copy words from it or make it easier on the eyes, you need a second step called optical character recognition, or OCR, which reads the letters in the image and adds a hidden text layer. Many scanner programs offer this as an option.
Understanding that difference helps with most of the choices below.
Before you start
- Clean the glass with a soft, lint-free cloth, with the device off, following the maker’s guidance. Smudges and dust appear on every scan.
- Remove staples, clips and sticky notes. They can jam a feeder or scratch the glass.
- Smooth folds and curls. Creased pages scan with shadows.
- Choose the right input. The flat glass suits single sheets, books and fragile or odd-sized items. An automatic document feeder suits stacks of ordinary pages. Feeders are not gentle, so do not use one for old or valuable papers.
- Check that the computer sees the scanner. If it does not, installing official device drivers or the maker’s scanning app is usually the fix. The basics of making a printer visible are described in the guide to printer setup, and most all-in-one scanners behave similarly.
Choosing a resolution
Resolution is measured in dots per inch, or DPI. It controls how much detail is captured and how large the file becomes.
- For ordinary printed text that you will read on screen or print again, around 200 to 300 DPI is commonly enough.
- For scans you plan to run through OCR, 300 DPI is a widely used setting, because it tends to give reliable recognition without very large files.
- Small print, fine detail or documents you want to enlarge may justify a higher setting.
Doubling the DPI makes the image four times as many pixels, because it doubles in both width and height. That is why file sizes grow quickly. A stack of high-resolution color scans can take up a surprising amount of disk space, so choose a setting for the job rather than defaulting to the maximum.
Color or black and white
Use color for anything where color carries meaning, such as a chart, a stamp or a photograph. For plain black text on white paper, grayscale or black-and-white modes produce smaller files. Very light or faded originals often scan better in grayscale than in strict black and white, which can drop faint marks.
Choosing a file type
- PDF is the usual choice for documents. It can hold many pages in one file, and it can carry a text layer from OCR. A searchable PDF is generally the most useful format for paperwork.
- JPEG is a compressed image format, good for single pages or photos where small size matters more than perfect fidelity. Text can show slight blurring around the edges at low quality settings.
- PNG is lossless and handles sharp edges well, but is bigger than JPEG.
- TIFF is a lossless format sometimes used for archiving.
For most household paperwork, a multi-page, searchable PDF is a good starting point. If a form needs to be sent to an agency or an employer, check what format they accept before you scan, because some have limits on file size or type.
Making scans searchable
If your scanning software offers a searchable PDF or an OCR option, turn it on. Afterward, check that it worked by opening the file and using the find command to look for a word you know appears. OCR is not perfect: handwriting, small print, stamps, poor copies and unusual fonts may be misread, so treat the hidden text as a convenience and keep the visible image as the authoritative record.
If a document contains numbers that matter, such as amounts or account details, compare them against the image rather than trusting the recognized text.
Keeping scans organized
A pile of files called scan001 to scan400 is nearly as hard to search as the original paper pile. A little structure helps.
- Pick a naming pattern and stick with it. A common approach starts with the date in year-month-day order so files sort naturally, followed by the source and a short description.
- Use folders by topic or by year, such as household, medical, taxes or warranties. Don’t create more layers than you will really use.
- Keep one document per file, unless pages clearly belong together.
- Avoid special characters in names. Letters, numbers, hyphens and underscores are safe on most systems.
Privacy and security
Scanned paperwork often includes sensitive details such as identity numbers, bank information and medical records. Keep it in a protected location, avoid emailing it casually and think carefully before uploading it to a cloud service. If you use encryption or password protection, store your password safely. Delete scans from shared or public computers when you finish. Be sure that old scanned files are also removed properly when you sell or give away a machine, as described in the guide to wiping drives.
Back up and decide what to keep
A scan only protects you if it survives. Include your scan folders in your regular backups. The guide to backup basics outlines a general approach, and cloud backups offer an offsite copy, though sensitive files deserve extra thought.
Keep the paper originals of documents that may be legally required in original form, such as certain certificates and legal papers. Rules vary by location and by document, so confirm with the relevant authority before discarding anything important. If you do shred paper after scanning, make sure the scan is checked and backed up first, and open a few pages to confirm they are legible.