What Is a Pixel?

A pixel (short for "picture element") is the smallest unit of a digital image. Every digital photo or graphic is made up of a grid of pixels, each one assigned a specific color. Zoom into any digital image far enough and you'll see the individual colored squares.

The total number of pixels in an image is described by its pixel dimensions — width × height. An image that is 1920 pixels wide and 1080 pixels tall has a total of 2,073,600 pixels, which is approximately 2 megapixels (MP).

More pixels means more detail. A 12-megapixel photo captures more visual information than a 2-megapixel photo of the same scene, which means you can print it larger or crop into it more deeply without the image looking pixelated.

Key Terms Explained

Abbreviation
PPI

Pixels Per Inch. Describes how densely pixels are packed on a screen. Higher PPI = sharper display. Most modern smartphone screens are 300–460 PPI.

Abbreviation
DPI

Dots Per Inch. A print term — how many ink dots a printer lays down per inch. Higher DPI = sharper print. Standard for quality print is 300 DPI.

Abbreviation
MP

Megapixels. One megapixel = one million pixels. A 12MP camera produces images roughly 4000×3000px. More MP = more detail and larger possible print sizes.

Term
Resolution

Often used loosely to mean either pixel dimensions or PPI/DPI. Context matters: "high resolution" on screen means many pixels; in print, it means high DPI.

PPI vs DPI: What's the Difference?

PPI and DPI are often used interchangeably, but they describe different things. PPI is a screen measurement; it describes how many pixels fit in one inch of screen space. DPI is a print measurement; it describes how many ink dots a printer places in one inch.

When you're working with digital images for web or screen use, DPI is essentially irrelevant — what matters is pixel dimensions. A 1920×1080px image displays at 1920×1080px on screen regardless of what DPI is embedded in the file's metadata. You can change the DPI metadata number without changing any pixels, and the image will look identical on screen.

DPI matters when you print. When a printer receives an image, it uses the DPI value to determine how large to print it. An image with 300 DPI tells the printer to pack 300 dots per inch — producing a sharp, detailed result. An image at 72 DPI tells the printer to pack only 72 dots per inch, producing a much larger but noticeably softer print.

Simple rule: For screens and web, think in pixels. For print, think in DPI. The DPI embedded in your image file has no effect on how it looks on screen.

What Resolution Do You Need for Print?

Print resolution requirements depend on the print size and viewing distance. The standard recommendation is 300 DPI at the intended print size. Here's what that means in practice:

Print SizePixels Needed at 300 DPIApprox. Megapixels
4 × 6 inches1200 × 1800 px~2 MP
5 × 7 inches1500 × 2100 px~3 MP
8 × 10 inches2400 × 3000 px~7 MP
11 × 14 inches3300 × 4200 px~14 MP
16 × 20 inches4800 × 6000 px~29 MP
24 × 36 inches (poster)7200 × 10800 px~78 MP

For large-format prints (posters, banners) that are viewed from a distance, you can use lower DPI — 150 DPI is often sufficient for posters viewed from 2+ meters away. Billboards and large outdoor prints may only need 10–30 DPI because the viewing distance is so great.

What Resolution Do You Need for Screens?

For screen use, resolution requirements are simpler: match your pixel dimensions to the display size where the image will appear. A 1920×1080px image fills a 1080p monitor perfectly. A 1280×720px image is sufficient for a YouTube thumbnail.

The complication is high-DPI ("Retina") screens. Modern smartphones, MacBooks, and 4K monitors have screens where each CSS/logical pixel actually contains 2×, 3×, or even 4× as many physical screen pixels. An image displayed at 400×400 CSS pixels on a 2× Retina display is actually being rendered at 800×800 physical pixels. For sharp display on these screens, your images need to be 2× the size of their display dimensions.

What Happens When You Resize an Image?

When you make an image smaller (downscale), pixels are removed and the image loses detail — but because you're viewing fewer pixels at the same display size, this typically isn't visible. Downscaling is safe and common.

When you make an image larger (upscale), the software has to invent pixels that weren't there. It does this through interpolation — estimating what the new pixels should look like based on their neighbors. Basic interpolation (nearest-neighbor) produces blocky results. Better algorithms (bicubic, Lanczos) produce softer but smoother results. AI upscaling tools can do a better job by using learned patterns to reconstruct plausible detail, but they can't fully recover detail that wasn't captured in the original.

The practical takeaway: always work from the highest-resolution source image you have, and resize down to your target size. Starting from a small image and scaling up will always produce inferior results.

↔ Resize your images to exact pixel dimensions — free, no sign-up, no watermarks

Open Image Resizer →

Common Resolution Questions

Is 72 DPI enough for web images? DPI is irrelevant for web images — only pixel dimensions matter. A 72 DPI image and a 300 DPI image that have the same pixel dimensions will look identical on screen. The "72 DPI for web" convention is a legacy from early Mac displays and doesn't carry practical meaning today.

Why does my photo look great on screen but blurry when printed? Your image probably has enough pixels to look sharp on screen, but not enough to print at the size you chose. A 1920×1080px image looks great on a monitor but would only print sharply at about 6.4×3.6 inches at 300 DPI. Printed at 8×10 inches, it would be stretched well below 300 DPI and look noticeably soft.

How many megapixels do I need? For web use, social media, and screen display, 2–5MP is more than enough. For print up to 8×10 inches, 8–12MP covers most needs. For large-format print, you need 20MP or more. Most modern smartphones capture 12–50MP, which is sufficient for almost any use case.