How to Choose Image Compression Settings for JPEG, PNG, and WebP

There is no single image-quality number that works across JPEG, PNG, and WebP. Choose a format based on the image and whether you can accept visual changes, resize it to the dimensions it will actually be displayed at, then compare a few encoded versions. Use the smallest file that still looks right and meets your transparency, metadata, and compatibility needs.

Choose the format by image type and fidelity needs

Image content matters as much as format. Photographs, flat-color graphics, text, sharp edges, and transparency can react differently to lossy encoding. A source preset may provide a useful first pass, but it does not replace inspecting the output.

#	Preview	Product	Price
1		Compressed Image File Formats: JPEG, PNG, GIF, XBM, BMP (SIGGRAPH Series)	$50.15	Buy on Amazon

JPEG: use quality values as encoder-specific starting points

JPEG is lossy: encoding discards information, and re-encoding cannot restore it. In the libjpeg-turbo cjpeg tool, quality runs from 0 (worst) to 100 (best), with a default of 75. The project says photographic images generally fall between 50 and 95 and recommends finding the lowest setting that still looks visually indistinguishable from the source. These are cjpeg guidelines, not a scale that transfers exactly to every editor or encoder. Its documentation also cautions that a value near 100 can substantially increase file size for little visible benefit. libjpeg-turbo cjpeg usage documentation

Google’s deprecated PageSpeed image-optimization guidance historically suggested JPEG quality 85 or lower when reducing a higher-quality source, 4:2:0 chroma sampling, and progressive encoding for larger files. Treat those as experiments to compare, not current universal rules. Google PageSpeed image optimization guidance (deprecated)

#1 Best Overall

Sale

Compressed Image File Formats: JPEG, PNG, GIF, XBM, BMP (SIGGRAPH Series)

$50.15

Buy on Amazon

• For black-and-white content, consider grayscale encoding if color information is not needed.
• During editing, keep a lossless working copy and export to JPEG once the edit is complete rather than repeatedly decoding and re-encoding JPEG files.

PNG: compression effort does not lower visual quality

PNG compression is lossless. Changing its compression effort can change file size and processing time, but not the decoded pixels. ImageMagick’s documentation explains that its PNG “quality” setting controls zlib compression and filtering behavior rather than acting as a visual-quality slider. A rewrite may even make an already optimized PNG larger, so check the output size against the original. ImageMagick command-line quality option

• Remove an alpha channel only when every pixel is opaque and transparency is not needed. Google’s optimization guidance recommends removing unused alpha. Google PageSpeed image optimization guidance (deprecated)
• Keep metadata when the delivery use case needs provenance, color-profile, or camera information; otherwise, remove only what is unnecessary.
• If a limited-palette graphic can tolerate changes, test a palette-based lossy workflow separately. Raising or lowering a standard PNG compression level does not make PNG lossy.

WebP: distinguish lossy quality from lossless effort

WebP supports both lossy and lossless encoding, transparency, and optional metadata. In Google’s cwebp tool, -q ranges from 0 to 100 and defaults to 75. In lossy mode, lower settings generally produce smaller files with lower quality. In lossless mode, the same control is an effort-versus-size choice: lower settings encode faster and usually produce larger output, while higher settings spend more time for smaller output. Google cwebp documentation

The -m encoding-method option ranges from 0 to 6; higher values spend more time evaluating encoding possibilities. For transparent images, cwebp defaults to alpha quality 100, and lowering it allows lossy alpha compression. In lossless mode, RGB values under fully transparent pixels are not preserved unless -exact is used. Metadata copying is configurable and defaults to none. Check those behaviors when the transparent pixels or metadata have meaning for your workflow. Google cwebp documentation

Use lossy WebP when its visible difference is acceptable; choose lossless WebP when exact pixel reconstruction is required. Compare at the same dimensions against JPEG or PNG and inspect edges, fine texture, gradients, and transparency. Google lists Chrome, Safari, Firefox, Edge, and Opera among browsers that support WebP; verify support for the actual audience and software environment. Google WebP overview

Compare candidates with a repeatable workflow

1. Preserve the source. Make test exports from the same original, and record the encoder and version so you can reproduce the result.
2. Set delivered dimensions first. Resize to the pixel dimensions the image needs on the page or in the application. Comparing an oversized source with a correctly scaled delivery image obscures the effect of compression. Google’s optimization guidance recommends considering resolution and appropriately scaled assets. Google PageSpeed image optimization guidance (deprecated)
3. Choose a mode. Use lossy encoding when visual change is acceptable, lossless when pixel preservation matters, and an alpha-capable format or mode when transparency is required.
4. Export a small set of candidates. Try several quality or effort settings rather than relying on a supposed magic number. Keep other options—such as chroma sampling and metadata handling—consistent and record them.
5. Inspect the intended result. Compare file size and appearance at the rendered size, then zoom into risky areas: text, sharp edges, gradients, texture, and transparency. If you use an objective image metric, record the metric and apply it consistently; a score is not a substitute for checking the image in its actual use.
6. Select and repeat. Choose the smallest output that meets the visual and delivery requirements. Test representative photographs, graphics, and transparent assets instead of assuming one image predicts every other image.

What published WebP size comparisons do—and do not—show

Google’s WebP overview, last updated in 2025, reports WebP images as 25–34% smaller than comparable JPEG images at equivalent SSIM quality, and WebP lossless images as 26% smaller than PNG. These are Google’s comparison claims, not guaranteed savings for an individual file. The page update date does not by itself establish that the underlying comparison method or dataset is new. Google WebP overview

Google’s separate comparison study reports an average 25–34% smaller file size at equivalent SSIM index across its named datasets and settings. It identifies libwebp 0.1.2, released in Q1 2011, and libjpeg 6b; those historical encoder versions make the result unsuitable as a current benchmark for your own assets. Google WebP comparison study

For a real delivery decision, compare the actual outputs with the encoders and settings you plan to use. Judge fidelity and artifact character alongside file size, dimensions, transparency, metadata needs, encoding time, and format support; quality numbers alone do not rank the choices.

Quick Recap

SaleBestseller No. 1

Compressed Image File Formats: JPEG, PNG, GIF, XBM, BMP (SIGGRAPH Series)

$50.15

Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.