What DPI Should You Scan At? A Practical Guide

Scanner software almost always defaults to a single resolution regardless of what you’re actually scanning, which means most people either scan everything too low to be useful later, or everything too high and end up with enormous files clogging their storage. Matching the resolution to the document actually solves both problems, and the right number is usually easy to work out once you know which of two questions you’re actually answering: what output size do you need, or what kind of document is this.

What DPI actually measures

DPI, or dots per inch, describes how many pixels the scanner captures per inch of the original. A higher DPI captures more detail but produces a larger file; a lower DPI produces a smaller, faster-to-scan file that may look blurry once you zoom in or print it larger than the original.

Work backward from what you need

The most reliable way to pick a DPI is to start from your intended output. If you want a scanned document to be exactly as wide as 2,550 pixels and the original is a standard 8.5-inch-wide letter page, the scan DPI selector divides the two — 2,550 ÷ 8.5 — and returns exactly 300 DPI, with no rounding needed since 300 is already a standard scanner setting. That’s the setting that gets you precisely the pixel dimensions you asked for, instead of guessing and checking after the fact.

When the math lands between two settings

Output targets rarely line up as neatly as the letter-page example. Scan a 4-inch-wide photo print and ask for 1,800 pixels of output width, and the raw math comes out to 450 DPI — a number no scanner actually offers. The tool rounds up to the nearest setting that meets or exceeds it, which in this case is 600 DPI, not 400; rounding down would leave you short of the 1,800 pixels you actually asked for. Always round up when the math falls between two standard settings, even though it means capturing a bit more resolution than the bare minimum.

Baseline settings by document type

When you don’t have a specific pixel target in mind, a few widely used baselines cover most home-office scanning needs. Ordinary text documents — letters, receipts, forms — scan well at 300 DPI, which keeps them sharp and readable by optical character recognition software. Documents with small or fine print benefit from stepping up to 400 DPI. Photographs and printed images are usually scanned at 600 DPI to preserve fine detail. Large-format originals, like posters or blueprints where the source is already big and doesn’t need much magnification, can often drop to around 200 DPI without losing anything visible at their working size. Film negatives and slides sit at the opposite extreme, often around 2400 DPI, because the physical original is so small that any usable output size represents heavy magnification. Choosing a Scan DPI for the Job goes deeper on that archival end of the range, including why a library or museum digitizing the same negative might scan it differently than a home office would.

Round up to what your scanner actually offers

Most scanners only provide a fixed menu of resolutions to choose from — commonly increments like 75, 100, 150, 200, 300, 400, 600, 1200, and 2400 DPI — rather than an arbitrary number. If your calculation comes out to something in between, round up to the next available setting so the result meets or exceeds what you actually need, the same way the 450-to-600 example above does.

Resist the urge to scan everything at maximum

It’s tempting to just scan everything at the highest setting to be safe, but doing that for a large batch of ordinary text documents multiplies file sizes and scanning time for no real benefit. A standard letter page scanned at 300 DPI in pure black-and-white comes out to 2,550 by 3,300 pixels — about 8.4 million pixels, or roughly 1MB before any compression at one bit per pixel. Scan the same page at 2400 DPI instead and the pixel count (and, roughly, the file size) jumps by a factor of 64, for a page that doesn’t become any more readable than the 300 DPI version once either one is printed or viewed on screen. That storage and scanning-time cost multiplies across an entire filing cabinet’s worth of paper.

Color mode matters as much as DPI

DPI and color mode are two separate settings that both affect file size, and it’s easy to leave color mode on a default that doesn’t match the document. Before any compression, a full-color scan stores 24 bits of color information per pixel versus 1 bit per pixel for a black-and-white (bitonal) scan — 24 times the raw data for the same page at the same DPI, for a document that has no color information worth keeping in the first place. Compression narrows that gap somewhat in the saved file, but a bitonal or grayscale scan of a plain text page is still reliably the smaller, more appropriate choice. Reserve color scanning for documents where color is actually part of the content — photographs, certificates with colored seals or ink, anything where a black-and-white copy would lose real information — and use grayscale or bitonal modes for plain text, which keeps both the DPI and the color setting matched to what the page actually needs.

OCR accuracy depends on more than DPI alone

Optical character recognition — the software that turns a scanned image of text into searchable, selectable text — generally does better with more resolution, but only up to a point, and DPI isn’t the only factor. A crisp, well-lit scan of clean printed text at 300 DPI usually recognizes better than a blurry or skewed scan at 600 DPI; scanner glass that needs cleaning, a page that wasn’t flat against it, or heavy compression applied after scanning can all undermine OCR accuracy regardless of the DPI setting. If OCR results look poor at your usual setting, check the scan quality itself — focus, skew, lighting — before assuming the fix is simply a higher DPI. It’s also worth spot-checking the OCR text against the original page after scanning a new document type for the first time, rather than assuming it caught everything correctly; a missed digit in a scanned invoice total is the kind of error that’s easy to overlook until it matters.

Duplex and multi-page documents

A document scanner that can scan both sides of a page in one pass (duplex scanning) is worth using for anything printed on both sides, since a single-sided scan of a two-sided document silently loses whatever was on the back. It’s also worth confirming your scanning software combines multi-page documents into a single file rather than saving each page separately — a 40-page contract saved as 40 separate image files is technically digitized, but far less usable than the same pages combined into one searchable PDF, regardless of what DPI each page was captured at. It’s a small setting to check once and forget, rather than something worth re-verifying every scanning session.

Save the settings you land on

Once you’ve worked out the right resolution for the types of documents you scan regularly, most scanner software lets you save named presets — “Text 300,” “Photo 600,” and so on — so you’re not re-deriving the number every time you sit down to digitize a new stack of paper.

File format matters as much as resolution

A carefully chosen DPI can still be undermined by the wrong file format. Text documents scanned for archiving or OCR generally do best saved as PDF, which keeps the pages together as one searchable file rather than a folder full of loose images, and which typically compresses a mostly-white text page far below the roughly 1MB uncompressed estimate above. Photographs are usually better saved as high-quality JPEG or, if you want to preserve every bit of original detail for future editing, an uncompressed format like TIFF — JPEG’s compression is efficient for viewing and sharing, but it does throw away some information every time the file is re-saved. Either way, the DPI you scan at and the file size you end up with are related but not identical questions: DPI decides the pixel dimensions going in, and the format and compression setting decide how much of that data actually survives to disk.

Batch scanning changes the calculation slightly

When you’re digitizing a large backlog rather than a single document, it’s worth scanning a small test batch at your chosen settings first and checking the resulting file sizes before committing to hundreds of pages. A resolution that seemed reasonable in the abstract can produce surprisingly large files once multiplied across an entire filing cabinet, and it’s far better to catch that on ten pages than after an entire evening of scanning. If the batch you’re digitizing is destined for a physical filing cleanup rather than just a hard drive, sizing the boxes and labels for whatever paper doesn’t get scanned is worth doing at the same time.