Why does a color coordinate change after cropping?

The original control

We authored a 240 × 120 opaque RGB pattern with a known coordinate-dependent formula, then cropped the rectangle (20, 10, 220, 110). Pillow’s crop box uses left and top inclusive, right and bottom exclusive. The resulting image measured 200 × 100. We saved both files without resizing or converting them through JPEG.

Source (60, 40) → crop (40, 30) — Results from our authored fixture and the exact comparison described below.

What the saved files measured

The original at (60, 40) stored RGB (148, 24, 244), or #9418F4. Cropped (40, 30) returned the same value. Cropped (60, 40) instead returned (112, 94, 138), or #705E8A. The mapping was (60 − 20, 40 − 10) = (40, 30); the different result at the unchanged numbers was expected.

Original test file	Detected format / mode	Dimensions	Bytes	Distinct decoded RGBA values
uncropped.png	PNG / RGB	240 × 120	692	28,800
cropped.png	PNG / RGB	200 × 100	592	20,000

These are source-file measurements made with Python 3.12.14, Pillow 12.3.0 and libwebp 1.6.0. The manifest records complete file hashes, decoded RGBA hashes and the relevant coordinates. File sizes are measured bytes for these inputs, not forecasts for other images.

How to interpret this comparison

This comparison isolates a location error from a color-conversion error. The crop keeps that chosen source pixel intact, but gives it a new address. A discrepancy report should therefore pair every code with its file and coordinate system. A coordinate remembered from an uncropped reference is not enough to identify a point in its cropped derivative.

Original experiment file previews on white, with detected formats and dimensions — These are previews of authored test files, composited on white for the diagram where necessary. They are not screenshots of a browser test.

Repeat this exact test

Run the crop case and check crop_box, source_xy and mapped_crop_xy.
Compare the source and mapped crop values before checking the deliberately wrong same-number sample.
When reproducing a discrepancy, keep the original file and crop offsets alongside the selected HEX code.

Download and unpack the ZIP. With Python and Pillow installed, run the command below. Compare the generated measurements.json with the included report. The script regenerates fixtures and computes the measurements; it does not obtain values from this article.

python reproduce.py --case cropped-image-pixel-coordinates --out rerun

Scope and limits

The coordinates here are zero-based source pixels in saved files. They are not pointer positions on a browser canvas. A displayed preview can add another scale mapping, and that separate transformation was not measured in this crop experiment.

Our own sample is a diagnostic control. Applying its conclusion to another file requires checking that file and the operations performed on it. We have not measured a global error rate or established that an automatic palette is an exhaustive list of image colors.

Did cropping alter the chosen pixel’s RGB?

The mapped pixel retained exactly the same RGB. Only the deliberately reused coordinate numbers selected a different source point.

Watch the source-file comparison

Original diagrams and synthetic English narration. Caption sentence timings are approximate.