How To Compress İmages
📖 Bu rehber ToolPazar ekibi tarafından hazırlanmıştır. Tüm araçlarımız ücretsiz ve reklamsızdır.
Lossy vs lossless
A 4MB hero image on your landing page is the fastest way to lose mobile visitors. Google’s own data shows bounce rate climbing sharply once page weight crosses 2MB on 3G — and an unoptimized iPhone photo alone often weighs more than that. The fix is image compression, which sounds simple but hides real tradeoffs: lossy versus lossless, JPEG quality sliders, WebP adoption, PNG optimization, EXIF stripping, and the difference between “looks fine” and “ships fast.” This guide covers how each compression mode actually works, the target file sizes for different page contexts, why mobile and desktop need different settings, and the small details that separate professional compression from the default export.
JPEG quality: what the slider actually does
JPEG quality is a number from 0 to 100 that controls how aggressively the DCT coefficients get quantized. It’s not a linear scale. The sweet spot for web photos is 75–85. Below 60 you start seeing visible block artifacts in flat areas like sky gradients. Above 90 you’re adding file size with almost no perceptible quality gain.
WebP: the modern default
Save once from the source at quality 85. Never re-save a JPEG at high quality expecting to recover detail — once it’s gone, it’s gone, and each re-encode introduces additional loss (“generation loss”).
PNG optimization
WebP beats JPEG at the same visual quality by roughly 25–35% file size, and adds transparency support. It’s been supported by all major browsers since 2020, so there’s no practical reason not to ship it. A 430KB JPEG at quality 85 is typically a 290KB WebP at the same visual fidelity.
Target sizes by context
Tools like pngquant and oxipng can reduce a typical PNG screenshot by 40–70% with no visible change. For photographic content, always prefer JPEG or WebP over PNG — PNG-on-photos is 5–10x the file size of JPEG for the same perceived quality.
Resolution before compression
These are photo-realistic targets. Flat illustrations in SVG ship at 2–10KB and should almost always be used over raster for logos, icons, and diagrams.
Mobile vs desktop
Compressing a 6000×4000 image to 200KB damages it more than resizing first to 1920×1280 and compressing. The compressor has fewer pixels to work with, so the per-pixel bitrate is higher. Always resize to the display dimension (or 2x for Retina) before compressing. A 400px-wide card thumbnail doesn’t need a 4000px source.
EXIF stripping
Photos from phones and cameras carry EXIF metadata: GPS coordinates, camera model, lens, exposure settings, sometimes the photographer’s name. This adds 5–50KB per image and, more importantly, can leak location data you didn’t intend to publish — a photo of your kid’s birthday party can geotag your home address. Strip EXIF on any image going to the public web unless you have a specific reason to keep it (photojournalism, stock photography attribution).
Progressive JPEG
A progressive JPEG loads in layered passes — blurry preview, then sharper, then final — instead of top-to-bottom. On slow connections the user sees something meaningful fast. File size is nearly identical to baseline JPEG (often slightly smaller). Most encoders have a “progressive” or “interlaced” checkbox; turn it on.