How to Change the DPI of a Picture: What It Actually Does and How to Do It

DPI is one of those settings that looks simple on the surface but quietly controls whether your printed photos look crisp or blurry, whether your files are accepted by a print shop, and how large your images can scale. Here's what's really happening when you change it — and why the same number can mean very different things depending on your situation.

What DPI Actually Means

DPI stands for dots per inch. In the context of digital images, the closely related term PPI (pixels per inch) is technically more accurate, but the two are used interchangeably in most software.

The number tells output devices — primarily printers — how densely to pack pixels when reproducing an image. A setting of 72 DPI is the web standard; 150 DPI is acceptable for casual prints; 300 DPI is the professional print standard for sharp, detailed output.

Here's the part most people miss: changing DPI in metadata alone does not change the actual pixel data in your image. It only changes a label embedded in the file. A 3000 × 2000 pixel image tagged at 72 DPI and the same image tagged at 300 DPI contain exactly the same number of pixels. What changes is how large a printer thinks it should render that image.

DPI ValueTypical Use CasePrint Size at 3000px Wide
72 DPIWeb, screen display~41 inches
150 DPICasual home printing~20 inches
300 DPIProfessional print quality~10 inches
600 DPIFine art, archival printing~5 inches

Two Different Things Called "Changing DPI"

This is where confusion usually starts. There are actually two separate operations most people call "changing DPI":

1. Changing the DPI Metadata Tag Only

This resets what the file claims its density is, without touching pixel dimensions. The image looks identical on screen. It will print at a different size, but quality is unchanged because no data was added or removed. This is what most basic tools do.

2. Resampling the Image

This actually adds or removes pixels to hit a target DPI at a specific physical size. If you want a 5-inch-wide print at 300 DPI, the image needs exactly 1500 pixels across. If it only has 800, software must interpolate — essentially guess at new pixel values. This is called upsampling, and it can introduce softness or artifacts. Shrinking pixel count (downsampling) generally causes less quality loss.

🖨️ When a print shop specifies "300 DPI minimum," they usually mean actual pixel density at your intended print size — not just the metadata tag.

How to Change DPI in Common Tools

Adobe Photoshop

Go to Image → Image Size. You'll see Width, Height, and Resolution fields. To change only the metadata tag, uncheck "Resample" before adjusting Resolution. To resample (add or remove pixels), keep Resample checked and choose an interpolation method — Preserve Details 2.0 or Bicubic Smoother work well for upscaling.

GIMP (Free)

Go to Image → Scale Image for resampling, or Image → Print Size to adjust DPI metadata without resampling. GIMP separates these functions clearly, which helps avoid accidental quality loss.

Preview (macOS)

Open the image, go to Tools → Adjust Size. Check or uncheck "Resample image" depending on whether you want to change pixel count. This tool is convenient for quick adjustments but offers fewer interpolation options than dedicated editors.

Microsoft Paint / Paint 3D

Basic and limited. Neither tool gives you direct DPI metadata control in the way Photoshop or GIMP do. For serious DPI work, these aren't the right tools.

Online Tools (Img2Go, ResizePixel, etc.)

These handle simple DPI retagging and basic resampling without software installs. Suitable for occasional use, but you're uploading potentially sensitive images to third-party servers, and interpolation quality varies.

Variables That Change What Approach You Should Take

Several factors determine which method makes sense:

  • Your starting pixel count — A high-resolution original (from a modern smartphone or DSLR) often has enough pixels to hit 300 DPI at reasonable print sizes without any resampling at all. A heavily compressed or small web image likely doesn't.

  • Intended output size — Printing a wallet photo versus a poster changes everything. The math is straightforward: divide pixel width by target DPI to get print width in inches.

  • Output destination — Screen display doesn't care about DPI. Email attachments, web uploads, and social platforms render based on pixel dimensions only. DPI only becomes relevant at the print stage.

  • Software access and skill level — Photoshop gives the most control but requires familiarity. GIMP is free and capable but has a steeper learning curve. Preview covers basic needs on Mac without any setup.

  • File formatJPEG files store DPI metadata. PNG does too. WebP and some other modern formats handle this differently, and some online tools strip or reset metadata during conversion.

🔍 The Quality Ceiling No Tool Can Bypass

AI-powered upscaling tools (like Topaz Gigapixel, Adobe Super Resolution, or free options like Upscayl) use machine learning to make intelligent guesses when adding pixels. They often produce better results than traditional interpolation — especially on faces and textures. But even these tools work best when your original image has reasonable quality to begin with. Heavy compression artifacts, motion blur, or very small originals limit what any upscaler can recover.

The relationship between your original file's pixel count, your target print size, and your target DPI is fixed math. Software can influence how gracefully you approach the limits, but it can't rewrite them.

Whether a simple metadata change solves your problem or you need actual resampling — and how much quality loss is acceptable — depends entirely on what your source image contains, what size you need to hit, and where that image is going.