How to Add a Watermark to Multiple Images at Once

A watermark is a signal, not a lock. It tells a viewer where an image came from and makes casual reuse slightly less convenient. It does not stop a determined person from cropping, cloning, or regenerating the mark. If you need evidence of ownership or automated detection, a visible corner mark is the wrong tool; that requires metadata, registration, or an invisible watermarking system. What a batch watermark does well is put a consistent, readable credit on a hundred previews without opening a hundred files.

Decide what the mark is for

The design follows the purpose:

Client proofs. Large, semi-transparent, often diagonal and centered. The goal is to make the image unusable for final publication while still showing composition and color.
Public previews. Small, low-contrast, placed in a corner. The goal is attribution without ruining the viewing experience.
Marketplace listings. A name or store URL in a corner, sized so it survives the platform’s own thumbnail crop.
Internal documents. A department name or “Draft” label, usually more visible than aesthetically subtle.

Write down the purpose before choosing settings. A proof mark at 25 percent opacity is useless on a listing thumbnail, and a bold center mark on a portfolio image is obnoxious.

Settings that work

Use	Position	Size	Opacity	Rotation
Client proof	Center	35–50% of width	20–35%	30–45 degrees
Portfolio preview	Bottom-right	6–10% of width	35–55%	0
Marketplace listing	Bottom-right or bottom-left	8–12% of width	45–65%	0
Document label	Top-left	10–15% of width	60–80%	0

These are starting points, not rules. The real test is whether the mark is readable at the size the image will be displayed, and whether it disappears into a busy background.

Make it survive the image

A watermark that sits on a flat sky is easy to remove. A watermark that crosses textured detail is harder to clone out. For proofs, place the mark across the subject rather than in an empty corner.

Contrast matters more than color. White text with a subtle dark shadow works on most photographs; black text with a light shadow works on bright product shots. A logo supplied as a transparent PNG can be switched between a light and a dark version. Avoid pure white at full opacity on a light background and pure black on a dark one.

Keep the mark away from the outer edge. Social platforms crop previews, and a mark 10 pixels from the border can be cut off entirely. A margin of 3 to 5 percent of the image’s shortest side is a safer inset.

Batch watermark with the tool

The watermark tool applies one set of settings to every file in a batch.

Open the tool and add the images. Processing happens in the browser; nothing is uploaded.
Choose the watermark type: text or image. Text is faster and scales cleanly. An image watermark is better for a logo, and it should be a PNG with a transparent background.
Enter the text, or choose the logo file. For text, use the name or domain you actually control; a mark that points nowhere is just clutter.
Set the corner or position.
Set the margin so the mark sits 3 to 5 percent of the shortest side in from the edge.
Set the size as a fraction of the image, not in fixed pixels. This is what keeps the mark proportional across a batch that mixes phone photos and desktop screenshots.
Set opacity. For proofs, 25 to 35 percent; for attribution, 45 to 65 percent.
Set rotation if you want a diagonal proof mark.
Run the batch, open one output at 100 percent, and adjust before exporting the rest.

The size-as-a-fraction control is the important one. A 200-pixel logo looks reasonable on a 2000-pixel image and enormous on a 600-pixel one. Proportional sizing prevents that.

Separate batches by orientation

A bottom-right mark on a landscape image and a bottom-right mark on a portrait image occupy different visual space. If the batch mixes both, run two passes with adjusted margins rather than accepting an inconsistent result. The same applies to images with very different aspect ratios, such as a 16:9 banner and a 4:5 social crop.

If the mark must be identical in absolute pixels across a set, size it for the smallest image first. A mark that fits the smallest file will fit the rest; the reverse is not true.

Format and quality

Watermarking requires re-encoding, so the usual browser-side rules apply. EXIF metadata, including GPS coordinates, is dropped. Transparency is preserved in PNG and WebP; JPEG has no alpha channel, so a transparent logo is composited onto the image and the output is flattened.

For photographs, JPEG or WebP output keeps the file small. For graphics with text, PNG avoids adding another generation of JPEG artifacts around the lettering. If the image is going to a platform that re-compresses uploads anyway, a high-quality JPEG is usually the pragmatic choice.

Do not watermark an already compressed file repeatedly. Export once from the best source you have.

What a visible watermark cannot do

It cannot prevent reuse. Cropping the marked corner takes seconds. A centered proof mark can be obscured or reconstructed. Automated removal tools exist and are improving. Treat the watermark as a deterrent and an attribution cue, not as technical protection.

It also cannot establish ownership by itself. If ownership matters, keep the original files, the capture metadata, and a record of publication. A watermark helps people find you; it does not prove a legal claim.

Use it only on images you have the right to mark

Watermarking someone else’s photograph and presenting it as your own is copyright infringement, regardless of the tool. Removing another photographer’s watermark is also infringement in most jurisdictions. The tool has no way to check ownership, so the responsibility sits with the person running the batch.

If you are marking client work, confirm whether the contract allows a visible credit and where it may appear. Some clients require a clean deliverable and permit a watermark only on proofs.

A quick pre-flight check

The mark is readable at thumbnail size.
It does not sit inside a platform’s crop zone.
It has enough contrast on both light and dark areas.
The size looks consistent across the whole batch.
The output format preserves transparency if you need it.
You have the right to mark every image in the set.

For the export step, the image resizer explains how to bring a mixed batch to consistent dimensions before watermarking, which makes proportional placement far more predictable.