Windows ships with two scanning applications, neither of which is obviously the right one, and most scanners add a third from the manufacturer. Which you use changes the output more than you would expect — particularly whether you get a searchable PDF, a multi-page file, and a sensible file size. This guide covers the options in order of usefulness and the settings that actually matter.
Windows Scan (the modern app)
The Scan app from the Microsoft Store is the current first-party option. It is minimal in a good way.
- Install it from the Store if it is not present (it is not preinstalled on all builds).
- Connect the scanner and open the app. It detects WIA-compatible devices automatically.
- Set Source — Flatbed, Feeder, or Auto.
- Set File type to PDF.
- Under Show more, set colour mode and resolution.
- Scan.
What it does well: multi-page PDFs from a document feeder, straightforwardly, with no nagging. What it does not do: OCR, blank-page removal, deskewing, or any post-processing at all. The output is a PDF containing images and nothing else — searchable only after you run OCR separately, as in how to make a PDF searchable with OCR.
Files land in Pictures\Scans by default, which is an odd place for documents; change it in the app's settings before you accumulate two hundred of them.
Windows Fax and Scan (the legacy app)
Still present in every Windows install, still functional, and in one respect better: it gives you a Preview and a Custom settings dialog with brightness and contrast controls that the modern app lacks.
Its significant limitation: it saves scans as TIFF or JPEG, not PDF, unless the scanner driver intervenes. You then convert — see how to batch convert images to PDF. For a single page this is fine. For a fifty-page feeder job it is a poor workflow.
Use it when you need the finer exposure controls, or when the modern app will not see your scanner.
The manufacturer's software
Every serious document scanner ships software that is better than both Windows options for document work: Epson ScanSmart, Canon IJ Scan Utility, Brother iPrint&Scan, HP Smart, Fujitsu/Ricoh ScanSnap Home. They are frequently ugly and frequently try to sell you cloud storage, and they are still usually the right choice, because they include the processing that makes scans usable:
- Bundled OCR, producing a searchable PDF in one pass.
- Automatic deskew and crop, which matters more than people think — a crooked scan OCRs worse and looks unprofessional.
- Blank page removal for duplex batches (see how to remove blank pages from a PDF).
- Multi-feed detection, catching two sheets pulled through together — the failure that silently loses a page.
- Profiles, so "receipts" and "contracts" scan with different settings at one click.
If you scan more than occasionally, spend twenty minutes setting up profiles in the manufacturer's tool. It is the single biggest quality-of-life improvement available here.
Settings that actually matter
Regardless of which application you use, five settings determine whether the output is good.
Resolution. 300 dpi for ordinary printed text. 400–600 dpi for small type, poor originals, or anything you will OCR seriously. 150 dpi only for throwaway copies. Higher is not automatically better — 1200 dpi quadruples the file size of 600 dpi for no readability gain on ordinary text. The reasoning is laid out in scanning resolution and DPI for documents.
Colour mode. Greyscale is the sensible default for text documents: it handles faint pencil, coloured stamps and highlighter far better than black-and-white, at a fraction of colour's size. Use colour when colour carries meaning — signatures in blue ink, coloured charts, photographs. Use bilevel (black and white) only for crisp, high-contrast printed text where you want the smallest possible file.
File type: PDF, not JPEG. Multi-page and universally readable. If the app offers "PDF (searchable)", take it.
Duplex. Turn it on for double-sided originals, off for single-sided. Leaving it on for single-sided originals is the main source of blank pages, and if the pages come out in the wrong order see fixing page order from duplex scanning.
Compression / quality. Most tools offer a quality slider or a small/medium/large choice. Medium is almost always right. The lowest setting produces the mushy text described in why PDF text looks blurry.
Scanning without a scanner
Two routes that produce genuinely good results:
Your phone. Modern phone scanning apps — including the built-in ones in Windows' Phone Link, Microsoft Lens, and the iOS Notes and Files apps — do edge detection, perspective correction and contrast enhancement that a flatbed does not. For a handful of pages this beats walking to the office scanner. See scanning documents with your phone.
A multifunction printer's scan-to-email or scan-to-folder. Most office MFPs will deposit a PDF straight into a network share or your inbox without any software on your PC at all. This is often the most reliable path in a corporate environment where you cannot install anything.
Adding OCR afterwards
If your scanning app produced an image-only PDF, add the text layer separately. On Windows the practical options:
- Adobe Acrobat: Scan & OCR → Recognise Text. Best accuracy, costs money.
- OCRmyPDF (free, open source, runs under Python or WSL):
ocrmypdf --deskew --rotate-pages in.pdf out.pdf— excellent results and scriptable across a folder. - Microsoft OneNote: import the image, right-click → Copy Text from Picture. Crude but free and occasionally exactly enough.
- PowerToys Text Extractor: for grabbing text off the screen rather than out of a file.
Background on what OCR does and does not achieve: PDF OCR explained and OCR language support and accuracy.
After the scan: the file size problem
Scanned PDFs are large, because they are collections of images. A 20-page colour scan at 600 dpi can easily exceed 100 MB, which will not survive email. The compression path is well-trodden: how to compress a scanned PDF covers the specifics, and the general reasoning is in why is my PDF so large.
The best fix is upstream: scan at a sensible resolution in greyscale, and you will rarely need to compress at all.
A workable default setup
For an office where people scan contracts, invoices and correspondence:
- Manufacturer software, with three profiles: Document (300 dpi greyscale, duplex, OCR on, blank removal on), Photo/Colour (400 dpi colour, simplex, no OCR), and Archive (600 dpi colour, OCR on, PDF/A output).
- Output to a single watched folder on a network share, not to each person's Pictures directory.
- A consistent naming convention applied at scan time — see naming and versioning shared files.
- PDF/A for anything with a retention requirement: PDF/A archival format explained.
That takes an afternoon to set up and removes essentially all of the recurring friction.
Summary
Use the Store's Scan app for quick multi-page PDFs, the manufacturer's software for anything regular because it brings OCR and deskew, and Windows Fax and Scan only when you need its exposure controls. Scan at 300 dpi in greyscale by default, produce searchable PDFs rather than image-only ones, and set up profiles once instead of choosing settings every time.
