Is a logo in the corner a watermark?

It depends entirely on why it is there. Put a logo on one photograph for a listing and you have placed an element: you want it looked at, you cared where it went, and you will probably never do it again to that picture. Put the same logo on three hundred photographs at twenty per cent opacity so that anyone lifting one has to deal with it, and you have watermarked them — even though the mark is identical, because now it is about the set rather than the picture.

Can I not just do a watermark here one file at a time?

You can put a single semi-transparent mark on a photograph here and it will look right. For a folder, the logo-on-multiple-photos page places one mark on every file by a rule — one of nine anchors, a gap and a size as shares of each frame — and hands back a ZIP. What neither page does yet is repeat the mark across the whole frame, which is the part of protection that survives a crop.

Which one actually stops a picture being stolen?

Neither, and anybody telling you otherwise is selling something. A repeated mark across the subject makes reuse expensive and visibly damaged, which deters most casual copying; a corner mark deters almost none. Both come off with effort. The honest reason to mark a picture is attribution and friction, not prevention.

Reading

Overlay or watermark: telling the two jobs apart

Both put a graphic on a photograph and they are not the same operation. One is a mark you want read; the other is a mark meant to be ignored across hundreds of files. The intent decides the tool.

Open the placing toolBoth files stay on your device — we only draw one on top of the other.

One question sorts it

Do you want the person looking at this picture to read the thing you put on it? If yes, you are placing an element, and everything about the job follows: it goes once, it goes where you decide, it is sized to be legible, and you will fiddle with it for a minute because this one picture matters. If no — if the mark exists so that copying the picture becomes annoying — you are watermarking, and the job is the opposite shape.

That second job is about a set rather than a photograph. The mark has to land identically on every file regardless of its size, it has to repeat so a crop cannot lose it, it has to be faint enough to live with and stubborn enough to matter, and nobody is going to hand-place it three hundred times. That needs a rule rather than a drag — which is what the folder page on this site does — and, for real protection, a pattern repeated across the frame, which this site does not draw yet.

The same graphic, two different tools around it

How many times: Placed once, here. Repeated across the frame, there.
How many files: One picture at a time on the stage; a folder by rule on the multiple-photos page. A queue and an archive, there.
What you tune: Position, angle, size, mixing mode, shadow and outline. Versus spacing, lattice angle, row offset and instance count.
What is saved: Nothing. Each picture is its own decision. Versus named settings you reuse for years.
What success looks like: Somebody notices the mark and reads it. Versus somebody stops noticing the mark and gives up on the picture.
Where it sits: Usually a corner, deliberately. Versus across the subject, deliberately.

The awkward middle

Plenty of real jobs sit between the two. A photographer putting a name on twelve images from one shoot is somewhere in the middle; so is a shop owner marking a handful of listing photographs a week. The practical test is whether you would be annoyed to place the mark by hand. Twelve times is a mildly tedious afternoon and the placement is better for it. Three hundred times is a different program.

The other honest middle case is a mark you want read that also has to be hard to lift. You cannot have both. Legible means big enough to see and therefore easy to crop out; stubborn means across the middle of the subject and therefore in the way of the viewer. Pick which of those two you are actually buying, and the tool follows from the answer.

What neither of them does

Neither creates a right you did not already have. Copyright starts when the picture does. A graphic on top of it is a claim about that right, made visible, and nothing more.
Neither is invisible or forensic. There is no hidden payload in anything this site writes, because detecting one would need a service to check against and there is no server behind this page at all.
Neither survives a screenshot of a screenshot intact — but then neither does the picture, which is part of why the argument matters less than it feels like it should.

Questions people send about this

Is a logo in the corner a watermark?: It depends entirely on why it is there. Put a logo on one photograph for a listing and you have placed an element: you want it looked at, you cared where it went, and you will probably never do it again to that picture. Put the same logo on three hundred photographs at twenty per cent opacity so that anyone lifting one has to deal with it, and you have watermarked them — even though the mark is identical, because now it is about the set rather than the picture.
Can I not just do a watermark here one file at a time?: You can put a single semi-transparent mark on a photograph here and it will look right. For a folder, the logo-on-multiple-photos page places one mark on every file by a rule — one of nine anchors, a gap and a size as shares of each frame — and hands back a ZIP. What neither page does yet is repeat the mark across the whole frame, which is the part of protection that survives a crop.
Which one actually stops a picture being stolen?: Neither, and anybody telling you otherwise is selling something. A repeated mark across the subject makes reuse expensive and visibly damaged, which deters most casual copying; a corner mark deters almost none. Both come off with effort. The honest reason to mark a picture is attribution and friction, not prevention.