Skip to content
← Articles

How to Reduce Image File Size Without Making It Look Bad

Reduce image file size by choosing the right dimensions, format and compression settings instead of blindly lowering quality.

ReFormatic ·

Quick answer

The most effective way to reduce image file size is to remove data you do not need: oversized pixel dimensions, inefficient format choice, unnecessary metadata and excessive quality settings. Resize to the actual use case first, then choose a suitable format and compression level. Do not start by crushing a full-resolution image with extreme compression.

File size has several causes

An image file can be large because of:

  • high pixel dimensions;
  • high bit depth;
  • lossless encoding;
  • complex image detail;
  • animation;
  • alpha transparency;
  • embedded metadata;
  • thumbnails or profiles;
  • conservative compression settings.

That means “compress harder” is not always the best first move.

Step 1: fix dimensions before quality

Suppose a website displays an image around 1200 pixels wide, but the uploaded file is 6000 pixels wide.

You are storing and transferring far more pixel data than the displayed version needs.

Downscaling the image can reduce file size dramatically before touching aggressive compression.

For many web workflows, correct dimensions provide the biggest easy win.

Step 2: choose a format that fits the content

Photographs

JPG, WebP and AVIF are common candidates.

Screenshots and line art

PNG or lossless WebP can preserve sharp edges better.

Transparency

PNG, WebP and AVIF can support alpha transparency.

Do not convert a transparent asset to ordinary JPG unless you are happy to flatten the transparent background.

Step 3: use moderate quality settings

Lossy compression is a trade-off.

At high quality, file size may remain unnecessarily large.

At very low quality, you may see:

  • blockiness;
  • mosquito noise;
  • banding;
  • smeared texture;
  • halos around edges.

The useful setting is the lowest file size that still looks acceptable for the actual use case.

There is no universal “85 is perfect” rule.

Different encoders and image types behave differently.

Step 4: judge the final viewing size

Do not inspect only at 800% zoom.

If the image will appear as a 600-pixel-wide product card, inspect it near that display size as well.

Compression defects that are obvious under extreme magnification may be irrelevant in real use.

Conversely, text and line art can show visible artifacts even at normal size.

Step 5: avoid recompressing an already compressed copy

If possible, return to the original.

Bad workflow:

original → JPG → edited JPG → smaller JPG → social-media JPG

Better:

original → final social-media export

Each lossy generation can discard more information.

Should you convert PNG photos to JPG?

Often, but not always.

A photographic PNG may be much larger because PNG is lossless.

If the image has no transparency and does not need a lossless workflow, a well-encoded JPG, WebP or AVIF can be much more practical.

But do not apply that rule to:

  • transparent logos;
  • UI screenshots;
  • graphics with fine text;
  • images where exact pixel preservation matters.

Should you remove metadata?

Metadata can contribute to file size, although for a large photograph the pixel data usually dominates.

Metadata removal is more important for privacy and cleanliness than for dramatic compression in many ordinary cases.

Still, if you do not need:

  • GPS;
  • camera serial data;
  • editing fields;
  • embedded thumbnails;

removing them may be sensible.

WebP or AVIF for smaller web images?

Both are modern options.

MDN describes WebP and AVIF as efficient web image formats, while also emphasizing compatibility considerations.

Do not choose based only on a benchmark from another image.

Test your own source set and delivery environment.

A practical test

Take one source photograph.

Create:

  • JPG high quality;
  • JPG medium quality;
  • WebP;
  • AVIF.

Keep pixel dimensions identical.

Record:

  • output file size;
  • encoder setting;
  • visible artifacts;
  • opening compatibility in your target app/browser.

That gives you evidence for your own workflow.

ReFormatic workflow

  1. Add the original image.
  2. Set final pixel dimensions.
  3. Choose a format.
  4. Adjust quality if the format is lossy.
  5. Decide metadata policy.
  6. Preview.
  7. Export.
  8. Compare file size and appearance with the original.

Bottom line

The cleanest image optimization process is:

right dimensions → right format → sensible quality → optional metadata cleanup

not:

maximum compression → hope it still looks okay

Optimize for the real destination.

Try it with ReFormatic

Open the related tool

How to Reduce Image File Size Without Making It Look Bad · ReFormatic