Utility

Image Resizer

Resize images to specific dimensions or by percentage. Maintain aspect ratio option available.

Click to upload or drag and drop

JPEG, PNG, WebP, or GIF

Pixel dimensions vs. DPI: why it matters for print vs. web

Image size has two meanings: pixel dimensions (width × height in pixels) and physical size (inches or centimeters, determined by DPI or PPI—dots/pixels per inch). A 1920×1080 image at 72 DPI (web standard) is 26.7 × 15 inches. The same 1920×1080 image at 300 DPI (print standard) is 6.4 × 3.6 inches. The pixels haven't changed—only the intended output size. Web pages ignore DPI and just use pixel dimensions. Print requires 300 DPI to look sharp on paper.

Resizing by percentage is simple: scale down to 50% and you halve both dimensions. Downscaling (making smaller) always works well—you're throwing away redundant detail. Upscaling (making larger) is problematic: you need to invent pixels, and the result looks blurry unless you use specialized upscaling algorithms (bicubic, Lanczos, or AI-based upscaling).

Image resampling methods

  • Bilinear: Interpolates between the four nearest pixels. Fast, decent quality for downscaling, noticeable blurriness on upscaling.
  • Bicubic: Uses 16 neighboring pixels. Better than bilinear for both up and down scaling, still fast. Most image editors default to this.
  • Lanczos: Uses many pixels, producing sharper results, especially for downscaling. Slower; sometimes produces artifacts on upscaling. Professional standard.
  • Nearest neighbor: Copies the nearest pixel. Useful for pixel art and retro graphics where you want sharp blocky edges. Awful for photographs.

Recommended dimensions for major platforms

PlatformDimension
Instagram post (square)1080×1080px
Instagram story1080×1920px (9:16)
Facebook post1200×628px (16:9 recommended)
Twitter/X header1500×500px
LinkedIn post image1200×627px
Website hero image1920×600px to 1920×1080px
Print (4×6 at 300 DPI)1200×1800px

Aspect ratio preservation and why it matters

Locking aspect ratio means resizing one dimension automatically adjusts the other to maintain proportion. Unlocked, you can stretch or squash the image. Most use cases lock aspect ratio: if you crop to 16:9 for a video thumbnail, unlocking ratio would distort faces. This tool lets you toggle, but the default (locked) prevents accidents.

When uploading to social media, platforms apply their own crop/fit rules. Instagram squares are 1:1, but if you upload a 16:9 image, Instagram crops it down. Always resize to the platform's recommended dimension before uploading—let Instagram do no extra work.

Frequently asked questions

How much smaller can I make an image before it looks bad?

Downscaling to 50% looks great. 25% is still fine. Below 25%, you're losing meaningful detail. For photos, never downscale more than 75% at once—do it in steps if needed, allowing compression to smooth artifacts between steps.

Can I make a small image larger without it looking blurry?

Not really. Upscaling invents pixels, and bicubic/Lanczos do their best, but you're fighting physics. Upscaling 300% (tripling dimensions) often produces visible blur. If you need a larger image, start with a higher-res original.

What's the right DPI for printing?

300 DPI is the standard for sharp prints. 150 DPI is acceptable for large format (posters, billboards). Below 150 DPI, print quality degrades. Screens are 72–96 DPI, so you can never print a screen image without upscaling (which doesn't help quality).

Should I resize before or after compression?

Resize first, compress second. Resizing reduces pixel count, so compression has less work to do, resulting in smaller files. If you compress then resize, you might lose quality during compression, then resize the already-degraded image.