It trims to a shape, from the top-left
This is the first thing to know, because it is not what most people assume.
The [msm_tool slug="image-crop"] works out how large a region of the chosen ratio fits inside your image, keeps that region starting at the top-left corner, and discards the rest. There is no region to drag and no position to set.
Nearly every other cropper centres the region, so the difference matters. A portrait with the face in the middle of the frame will lose the bottom of the head to a square crop here, where a centring cropper would have kept it.
The practical consequence: check where your subject sits before you crop. If it is not toward the top-left, either recompose first in an editor or expect to lose part of it.
When trimming to a shape is the right job
Despite that constraint, a fixed-position trim is exactly what you want in several common cases:
- Screenshots, where the content you care about is usually at the top-left by construction.
- Batch work where every image was shot or exported the same way, so one crop rule applies to all of them.
- Getting a ratio quickly when the composition already suits it and you only need the shape to match what a platform expects.
For anything where the composition matters, an editor with a draggable crop box is the better tool, and there is no shame in saying so.
Nothing is scaled
Pixels are kept at their original size and the rest are discarded. The result is smaller in dimensions, not resampled — so what remains is exactly as sharp as it was.
That is worth knowing because it distinguishes cropping from resizing. Cropping never costs quality; resizing always involves a choice about it. If an image is both too large and the wrong shape, crop first and resize second, so the resize is working with fewer, more relevant pixels.
PNG output is lossless and large
The result is written as PNG, which means nothing is re-compressed and a transparent background survives.
It also means a photograph that arrived as a 400KB JPEG can leave as a 3MB PNG despite now containing fewer pixels. That surprises people. If file size matters, convert the result back to JPEG or WebP afterwards — the one re-compression at the end is far better than compressing at every step.
Choosing a ratio
Let the destination decide, not the image:
- 1:1 — avatars, product grids, most social feed posts.
- 16:9 — video thumbnails, banners, embeds.
- 1.91:1 — the shape most platforms crop link previews to.
- 4:5 — the tallest most feeds will display without cropping again.
- 9:16 — vertical video and story formats.
Cropping to a ratio the destination does not expect means it will be cropped a second time, somewhere you cannot see and did not choose. That double crop is the usual reason a carefully framed image appears beheaded on a social card.
Crop for the smallest place it will appear
An image that works at full width often fails as a thumbnail, because the subject becomes too small to identify. If the same file serves both, crop tighter than feels right at full size — the large version tolerates a tight crop far better than the small version tolerates a loose one.
Keep the original
Cropping is destructive here: the discarded pixels are gone from the exported file. Always work from a copy, particularly when cropping to a platform's ratio, because platforms change their ratios and you will want to redo it from the full frame rather than from an already-cropped version.