Add image watermark

Provide the base image dimensions, the watermark image dimensions, a scale multiplier, and one of nine familiar anchor positions. The result gives the watermark’s left, top, right, and bottom coordinates together with its placed width and height. Optional proportional scale-down safely fits an oversized watermark within the base image while preserving its aspect ratio.

Describe both images in one coordinate system

Start with the width and height of the base image, then enter the watermark’s natural width and height. Pixels are the usual unit, but the calculation works with any consistent positive unit because it only compares and multiplies dimensions. A base image of 1200 by 800 and a watermark of 300 by 120 therefore describe two rectangles in the same coordinate system. The scale multiplier is applied uniformly to both watermark dimensions, preserving its aspect ratio. A scale of 0.5 makes each dimension half as large, while a scale of 2 doubles both. The calculator rejects zero, negative, missing, nonnumeric, or non-finite dimensions because those values cannot define a usable rectangle. It does not inspect or alter image files, estimate visible artwork inside transparent padding, or account for rotation. Use the actual canvas dimensions of each image. The returned width and height describe the complete placed watermark canvas, including any transparent pixels that belong to that source image.

Choose an anchor and understand the bounding box

The position chooses one of nine anchors: the four corners, the midpoint of each edge, or the exact center. Placement uses a top-left origin, which is standard for raster image coordinates: x increases toward the right and y increases toward the bottom. A top-left watermark begins at x zero and y zero. A bottom-right watermark ends at the base width and base height. Centered variants divide the remaining horizontal or vertical space equally, so their coordinates may contain decimals when the leftover distance is odd or fractional. The response reports x and y for the watermark’s top-left corner, width and height for its placed size, and right and bottom for the opposite edges. Together these values form the bounding box needed by many image libraries and rendering pipelines. No margin or padding is implied. If a design needs a gap from an edge, subtract or add that gap after calculation, or enlarge the conceptual watermark rectangle before positioning it.

Handle oversized watermarks deliberately

Scale is evaluated before placement. If the scaled watermark is wider or taller than the base image, the default behavior is an input error rather than a clipped or surprising result. This makes accidental use of the wrong dimensions visible early in an automated workflow. Set scale_down to true when fitting is the intended behavior. The calculator then selects the smaller of the available width ratio and height ratio, applies that additional reduction uniformly, and guarantees that both placed dimensions fit. It never enlarges a watermark under this option; the adjustment occurs only when the requested placement is oversized. The response includes the final applied scale and a scaled_down flag, so downstream code can record whether the requested size was changed. This is useful when preparing overlays for several image sizes: one placement rule can handle landscape, portrait, and thumbnail bases while preserving the watermark’s proportions. The calculation is deterministic, performs no upload, makes no network request, and costs $0.002 per API request when automated.