Image Compression Benchmark

1. Choose a source image

Runs JPEG and WebP at 60%, 80% and 90%, plus lossless PNG. Dimensions stay unchanged; all formats use a white background, matching ToolzHive’s image compressor.

No measurements yet. Choose an image to begin.

2. Compare measured results

3. Add another tool’s output

Process the exact same original in another tool, then select its downloaded result. These are user-supplied comparisons: source identity and settings cannot be independently verified. Resized outputs are rejected. External processing time is not measured.

Methodology and limits · version 1

ToolzHive uses the browser’s Canvas encoder, the same approach as its image compressor. Each setting gets one warm-up and three timed encodes; the median excludes decoding and file reading. Unsupported formats are skipped. Encoding quality values are browser settings, not visual quality scores.

File size reduction is (original bytes − output bytes) / original bytes. Negative savings mean a larger file. Inspect downloaded outputs for artifacts; this benchmark does not calculate perceptual quality. PNG preserves the rendered pixels, but metadata and transparency may change. Animated inputs are treated as a single rendered frame.

Browser, hardware and source image affect results. Use several representative photographs, screenshots and illustrations before drawing conclusions. Imported outputs are labelled separately, and no overall winner is declared.

Reports include a SHA-256 source fingerprint when supported, exact byte counts, dimensions, settings, timestamp and browser. Results are not uploaded or published automatically.