Watermark image files with enough visibility to communicate authorship without making the picture unusable. This page combines one base image with custom text or a logo, then lets you compare a single position with a repeated layout before creating the final file.

The tool accepts one JPG, PNG or WebP base image up to 40 MB, 12,000 pixels per side and 40 million pixels. A logo watermark can be another JPG, PNG or WebP file up to 10 MB. The output keeps the source pixel dimensions and can be saved as JPG, PNG or WebP.

Choose text or logo watermark

A text watermark is fast and remains editable on this page. It can show a name, brand, handle, website, copyright notice or review status. The current tool provides clean sans, editorial serif and technical monospace choices, a custom color and an optional contrast outline.

A logo watermark uses an existing image. This is appropriate when a consistent visual mark, signature or business identity already exists. A transparent PNG is usually the cleanest input because only the logo pixels are drawn; JPG logos always include their rectangular background.

The watermark becomes part of the exported pixels. It is not saved as a separate editable layer, and this page does not create reusable brand presets or store a logo for later sessions.

How to watermark an image on this page

Choose the base image. Add one JPG, PNG or WebP file. The preview reports dimensions, bytes and format.
Choose text or logo. Enter up to 80 characters, or add a logo image. For text, select a font, color and optional outline.
Choose a layout. Use a single watermark in one of nine positions, or repeat the mark across the image.
Tune appearance. Adjust size, opacity and rotation. Single marks include edge margin; repeated marks include tile gap.
Choose an output format. Select JPG, PNG or WebP and set quality when the format supports it.
Create and inspect. Check the mark at full result size, read the output report and download the generated file.

Every setting is reflected in the live preview. Changing a setting clears the previous result so the download always follows a deliberate create step.

Single position or repeated layout

The single layout provides nine familiar anchor points: each corner, each edge center and the image center. Edge margin is measured as a percentage of the shorter image side, helping the placement remain proportional across landscape and portrait files.

Corner placement is restrained and works for branding or attribution when cropping risk is low. Center placement is harder to crop away but can interrupt the subject. Check where the important visual content sits before choosing an anchor.

The repeated layout draws the watermark in staggered rows across the image. Rotation and gap control how the pattern moves through the frame. Repetition is more difficult to remove with one crop, but high opacity or dense spacing can make the image unpleasant to view.

Start with lower opacity and a moderate gap. Increase coverage only when the sharing context justifies it.

Balance size, opacity and rotation

These controls work together:

Size: text is based on the shorter image side; logo width is based on image width. This makes values proportional rather than tied to one fixed pixel size.
Opacity: Canvas globalAlpha applies a value from fully transparent to fully opaque before text or the logo is drawn.
Rotation: a small angle can distinguish the mark from image edges. Strong diagonal tiling increases coverage but also visual interruption.
Margin or gap: margin controls a single mark's distance from the edge. Gap controls spacing between repeated marks.

For a corner brand mark, try 60–80% opacity and enough size to remain readable after the image is displayed smaller. For a repeated proof watermark, try 20–45% opacity, a diagonal angle and generous spacing before deciding whether stronger coverage is necessary.

These are design starting points, not fixed rules. The correct balance depends on contrast, subject placement, expected display size and how much visual interruption is acceptable.

Choose a readable text watermark

Text must remain recognizable against both light and dark parts of the image. White works well on many photographs, but it can disappear over highlights. Dark text has the opposite problem.

The optional outline draws a contrasting stroke before the fill. White text receives a dark outline; other colors receive a white outline. This is a simple readability aid, not a full typographic effects editor.

Keep the content concise. Long text creates a wide marker that can run close to the edge or produce a crowded repeated pattern. Use a short name, handle or domain when the image may be viewed on a phone.

The font choices intentionally use the site's available browser fonts. If a long text line would exceed the image, the Canvas renderer reduces its effective font size to fit within the available width; the result panel still reports the requested percentage. This version does not upload custom fonts, curve text, adjust letter spacing, create shadows, add background pills or save text presets.

Prepare a logo watermark

Use a logo with clean edges and enough source resolution for the intended size. A tiny logo stretched across half of a large image can look soft. A very large logo file adds decode work without improving the result when the displayed watermark is small.

Transparent PNG is the practical default for signatures and logos with irregular outlines. Transparent WebP can also work when the browser decodes it correctly. JPG has no alpha channel and therefore brings its background rectangle into the final image.

The tool preserves the logo's aspect ratio. Logo width ranges from 5% to 60% of the base image width. Tiled logos at the upper end can overlap heavily, so adjust both size and gap.

What a visible watermark can and cannot do

A watermark communicates attribution and can discourage casual reuse. Canva's watermark guidance describes logo, name and signature marks, plus size, position, rotation and opacity controls. Its own localized guidance also acknowledges that watermarking can make copying harder without completely preventing it.

A visible mark is not encryption, access control, proof of registration or a guarantee against editing. Someone may crop a corner watermark, cover it, retouch it or use only an unmarked portion of the image. Repeated marks raise the editing effort but do not make removal impossible.

Keep original files, publication records and licensing information separately. Use the watermark as a visible communication layer, not as the only ownership or enforcement mechanism.

Choose an output format

JPG: suitable for opaque photographs. Any transparency in the base image is flattened onto white, and the quality control uses lossy encoding.
PNG: useful when source transparency or sharp graphic edges matter. The Canvas output is lossless and has no quality slider.
WebP: supports transparent output and adjustable quality in this workflow.

The source metadata is not copied into the Canvas-created output. That can remove EXIF and GPS fields, but the visible watermark does not replace metadata management. Inspect the generated file separately when provenance or privacy depends on file information.

Check the generated image

Review the output at both fit-to-screen and 100% size:

Readability: can the name, domain or logo still be identified after the image is displayed smaller?
Subject interference: does the mark cover a face, product detail or important text?
Cropping risk: could a single corner mark be removed with a simple crop?
Pattern density: for repeated marks, are opacity and gap tolerable across light and dark areas?
Logo edges: confirm transparent edges do not show an unexpected rectangle or halo.
Dimensions and format: verify the source pixel size is retained and the destination accepts the chosen file type.
Bytes: use the generated size rather than assuming the watermark has little effect on compression.

Illustrative watermark examples

Add a quiet portfolio signature

Assume a 2400 × 1600 JPG needs a creator name. Choose text, bottom right, 6% size, 70% opacity, 3% margin and the readable outline. Check the subject does not sit under the mark before exporting JPG.

Create a repeated proof image

Assume a client preview should not be mistaken for a final delivery. Enter “PROOF”, choose Repeat across, set approximately 8% text size, 28% opacity, -24° rotation and a 12% gap. Inspect the subject and raise or lower opacity based on actual readability.

Place a transparent brand logo

Assume a square PNG logo will sit on a product photograph. Choose Logo image, bottom center, 18% width and 80% opacity. Export WebP or PNG when the base image contains transparency.

Illustrative examples. Exact placement, bytes and appearance depend on the source image, logo, dimensions and browser encoder.

Scope and alternatives

Adobe Photoshop Express documents text and image watermarks, nine placement points, and size and opacity controls in its mobile workflow. Canva provides a broader design editor with drag-and-drop positioning, fonts, rotation, opacity and saved online projects.

base97 focuses on one export task: text or logo, nine anchor positions, repeated tiling, proportional size, opacity, rotation, spacing, three output formats and a detailed result report, all in the current browser. It does not provide accounts, cloud projects, arbitrary free-drag positioning, batch watermarking, saved presets, custom font uploads, SVG input or invisible digital watermarking.

Troubleshoot watermark problems

The watermark disappears on part of the image

Enable the text outline, choose a different color, raise opacity or move the single mark to an area with steadier contrast.

The logo has a white rectangle

The logo source is probably JPG or an opaque PNG. Use a PNG or WebP logo with real transparent pixels.

The repeated pattern is too crowded

Reduce marker size, increase tile gap or lower opacity. A repeated layout should remain identifiable without overwhelming the image.

The single mark is cut off

Reduce its size or increase edge margin. Very long text can be wider than the available area even when centered on an edge anchor.

The result file is larger

New text, edges and repeated patterns can make compression less efficient. Choose an appropriate output quality or use Image Compressor after confirming the watermark remains readable.

Watermark image FAQ

Can I place the watermark anywhere by dragging it?

Not in this version. Single marks use nine predictable anchors plus an edge margin. Repeated marks use a staggered tiled layout.

Does the watermark change image dimensions?

No. The output keeps the decoded width and height of the base image.

Can I save the watermark as a reusable preset?

No. The page does not create an account, store a logo or save presets between sessions.

Watermark Image

Place the mark where it stays readable.

Build a visible, deliberate mark.