Vutral
How to use this tool

Image Watermark Tool & Logo Stamper

Position

Your first image will appear here

How to use the image watermark tool

Back to the tool

One mark, sized as a share of each image rather than in pixels, so a batch of mixed shapes and resolutions comes back looking the same.

  1. 01

    Add the images you want to mark

    Drag them onto the page, use Upload images, or paste one in. Up to 25 MB each · JPG, PNG or WebP. Nothing is uploaded: the marking happens in this tab, which is also why a very large image can run out of browser memory where a smaller one will not.

  2. 02

    Type your text, or switch to a logo

    Text opens first and takes a single line, up to 100 characters. Logo takes a PNG, JPEG or WebP instead — a PNG with a transparent background is what you want here. The two tabs keep separate settings and you can use both marks at once, which is why they start in opposite corners.

  3. 03

    Size it as a share of the image

    Size, Opacity and Margin are percentages of each image’s shorter side, never pixels. That is the whole point: one setting marks a 6000 × 4000 photo and a 4000 × 6000 portrait identically, where a pixel size would look bold on one and lost on the other.

  4. 04

    Place it, then make it readable

    Nine anchors, and the mark is held one Margin away from each of the two edges it touches — anchor it to a middle and that distance stops applying on that axis, because there is no edge left to hold off. Color and Effect are what carry the mark across a photo that changes from light to dark underneath it.

  5. 05

    Watch the preview, then run

    The preview shows the mark before anything is written, and it works with an empty tray too — a grey stand-in stands for the image until you add one, so you can set the mark up first. Once images are there it draws the first one at its real proportions, which is the only honest test: a mark that reads at preview size can vanish on the full-size file. The run button then becomes Pause with a count, so a long batch can be stopped and picked up again.

  6. 06

    Read the rows, then collect the images

    Each row keeps its pixel size, because marking does not resize anything. Download ZIP takes the batch; the buttons on a row take one image. Note that copying puts a PNG on the clipboard whatever the row says — only the download button gives you the file in its own format.

How to read the watermark results table

The pixel size will not move, so the two things to read here are whether the row finished at all and what the re-encode did to the file size.

Result table columns

Image
The thumbnail and the name of the file you will get: the original name with -watermark added. Press the thumbnail to see the mark full size, which is the only view that tells you whether it is actually readable. A second line appears under the name when the row has something to report.
Dimensions
The same number twice, with an arrow between them. That is not the table failing to fill in: marking never changes the pixel size, and showing the pair keeps every row readable the same way as in the other image tools.
File size
File size before and after, and the only number that does move. It can go up: the image is decoded and written again at a fixed quality, so a photo that arrived heavily compressed leaves heavier than it came, mark or no mark.
Result
One badge per row. Rows still queued leave it empty, which says more plainly than a badge would that nothing has happened to them yet.
Actions
Download and copy once a row is finished, Retry once it has failed. Nothing on a row that is still waiting, because there is no file behind it yet.

Result badges

Processing…
This image is being worked on right now. Images are processed one at a time, so exactly one row carries this badge during a run.
Done
The image was marked and the file is ready. Every image that gets this far produces one — there is no quieter third outcome here, because the tool never decides your original was the better file.
Failed
The one image could not be drawn — hover the badge for the reason. It is almost always the browser running out of room for an image that large, which is a limit of the machine rather than of your file; resizing it first and marking the result gets past it. The rest of the batch is untouched and Retry reruns only this row.

What each control actually does

Size
For a logo, the length of its longer side; for text, the height of the letters — not the width of the line, which follows from the words you typed. Both are read as a percentage of the image’s shorter side.
Opacity
Held per mark, so a logo and a line of text on the same image can sit at different strengths. The mark is composed at full strength first and faded once, which is what keeps an outline from showing through its own letters.
Margin
Distance from the edges the mark is anchored to, again as a share of the shorter side, and taken off both sides of the axis. Push it far enough and there is no room left for the mark at all: the row still finishes, and the image comes back with nothing on it.
Effect
What makes the mark survive a background it was not designed for. Text can take a shadow or an outline; a logo is offered a shadow only, since a bitmap has no letterforms to trace.