Watermark Your Photos Before You Post Them

A watermark is attribution and friction, not protection. Anyone determined can crop a corner mark off or remove it with content-aware fill in about ten seconds, so the only placement that actually deters reuse is one that costs the thief the picture: a large, low-opacity mark across a low-contrast area of the subject itself. If you tuck your logo neatly into the bottom-right corner, you have made a credit line, not a lock — which is fine, as long as that is what you meant to make.

On Android, Convertlix's Watermark Adder stamps text or a logo on your images on the handset, so the photos never leave the device — no upload, no server-side copy, nothing to retain. The scope, up front: Android 8.0 or later, sign-in required (there is no guest mode), 25 MB per file and batches of 5 on the free tier against a server-set monthly allowance of conversions shown in the app, and the marked copies land in Downloads. The rest of this page is about placing a mark that does the job you actually want.

Decide which job the watermark is doing

These are three different goals and they want three different marks. Most people never separate them and end up with a mark that does none of them well.

• Credit. You want people who see the image to know who made it. A small, legible mark in a consistent position is exactly right, and removability barely matters — the audience seeing it is the point.
• Deterrence. You want casual reuse to be annoying enough that people don't bother. This needs the mark over the subject, because removal has to be more effort than asking you.
• Proof of ownership. You want to be able to establish that the image is yours. A watermark is weak evidence here and it is worth being blunt about that: what actually establishes ownership is your original file with its capture metadata, your archive, and in many jurisdictions registration. A visible mark helps a conversation; it is not the thing that wins one.

The mechanics of a mark that survives

Four things decide whether a watermark reads well on every image rather than disappearing on half of them.

Contrast, not colour. A white mark vanishes on a bright sky; a black one vanishes in shadow. A mid-grey mark at moderate opacity, with a subtle darker outline or drop shadow, stays readable over both. This is the single change that fixes most watermarks.

Opacity around 25–50% if it crosses the subject. Lower and it's decorative, higher and you have vandalised your own photo. If the mark sits on a plain area — sky, wall, out-of-focus background — you can go lower, because there's nothing competing with it.

Scale as a percentage, not in pixels. A mark sized in absolute pixels looks correct on one export and wrong on every other. Size it relative to the image width — somewhere between 15% and 40% of the width for a deterrent mark, under 15% for a credit line — and it looks deliberate at any resolution. This matters more than people expect once you start exporting the same photo for different placements; see Instagram, YouTube and X Image Sizes (2026) for how many sizes one image ends up in.

A watermark is destructive. It burns into the pixels. There is no undo, no layer, no way to lift it back off cleanly later. Always keep an unmarked master, in a separate folder, and mark copies. This is the mistake that costs real money: someone licenses an image and the file you still have is the watermarked one.

The invisible alternative, and why it doesn't replace this

Photographs can carry a copyright and creator credit in their metadata — the EXIF and IPTC fields that professional workflows have used for decades. It is invisible, it doesn't touch a pixel, and it is the correct place for your name, your URL and your licence terms.

The catch: most platforms strip metadata on upload. They re-encode your image to produce their display sizes, and the metadata usually does not survive that. So the field you carefully filled in is gone by the time anyone sees the picture, which is precisely why visible watermarks still exist in 2026.

There is a genuine tension here worth naming. The same metadata block holds your copyright credit and the GPS coordinates of where you were standing. Strip it for privacy and you strip your credit with it; Remove GPS and EXIF Data From Photos (Android) covers that side of the trade. For photos of your home, your children or your workplace, the privacy side usually wins and the visible mark carries the credit instead.

Content Credentials (C2PA) is the emerging standard aiming to fix this properly, by attaching cryptographically signed provenance that survives editing. As of September 2026 support is real but patchy — some cameras, some editing tools, some platforms — so it is worth knowing about and not yet worth relying on.

How to watermark your photos

1. Copy the photos you're marking into a working folder. Mark copies, never originals. If you skip this step once you will eventually regret it.
2. Decide credit or deterrence, and place accordingly. Credit goes in a consistent corner, small. Deterrence goes across the subject, large, at low opacity. Don't try to do both with one mark.
3. Make the mark work over light and dark. Mid-grey with a soft outline, or your logo with a subtle shadow. Test it on your brightest and your darkest image, not your favourite one.
4. Apply it to the batch. In Convertlix, open Watermark Adder, pick the images through Android's picker and stamp your text or logo. Free batches are capped at 5 items and 25 MB per file, so a large shoot has to go in groups — a 200-image set is a Premium batch size, and that is a limit worth knowing before you start.
5. Check the mark on the busiest image in the set. A mark that reads perfectly on a plain sky can disappear entirely over a detailed background. That image sets your contrast, not the easy one.
6. File the unmarked masters somewhere you'll find them. Separate folder, clear name. If you're marking hundreds of files, consistent naming is what keeps marked and unmarked sets apart — Rename Multiple Files at Once on Android covers doing that in one pass.

When this doesn't work

• Stock agency submissions reject watermarked images. Every major agency requires clean files; they add their own preview watermark. Sending a marked file is an instant rejection.
• Client deliverables. Nobody paying for photographs wants your logo on them. Mark the proofs, deliver clean finals.
• Print. A watermark sized for a screen looks crude at A3, and a print buyer will not accept one anyway.
• Legal enforcement. A watermark is not what makes a claim stick. Your originals, your records, and where applicable registration are. Treat the mark as a deterrent and a credit, not as evidence.
• Screenshots of text. A low-opacity mark over text makes the text harder to read and is trivially cropped out. Not worth it.
• Anything a determined person wants. Content-aware fill removes a corner mark in seconds and does a decent job on a low-contrast mark over a plain area too. The honest ceiling of this technique is "casual reuse", and marks placed over the subject are the ones that raise it meaningfully.

FAQ

Does a watermark stop people stealing my photos?

No — it raises the effort, which stops casual reuse and does nothing against anyone determined. A corner mark can be cropped off in one action, and content-aware fill removes a mark over a plain background convincingly in seconds. A large, low-opacity mark placed across the subject is genuinely hard to remove without visibly damaging the image, which is why that placement functions as a deterrent rather than a credit.

Where should I put a watermark?

If you want credit, a consistent corner at small scale is right and looks professional. If you want to deter reuse, put it across the subject at 25–50% opacity and 15–40% of the image width, over an area with moderate detail — a mark on a plain sky is the easiest kind to remove. Keep it out of the outer 10% if the image will be cropped by a social platform.

Will a watermark hurt my image's quality?

The mark itself is composited cleanly, but saving the marked copy re-encodes the file, so a JPEG picks up one more lossy pass. It is not usually visible, and it is a reason to mark once from the original rather than marking an already-marked or already-resized copy. The bigger risk is not quality but permanence: the mark is burned into the pixels, so keep an unmarked master.

Does Convertlix upload my photos?

No. The images are opened through Android's own picker, the watermark is composited and the new file written on the phone, and nothing is sent anywhere — the app has no storage permission, so it cannot read anything you did not select. Its network calls carry identity, licence, quota, configuration and payment data, never a photograph. Sign in once, switch on airplane mode, and it still works.

Is there a free way to watermark photos in bulk?

Yes. ImageMagick is free and made for this — magick mogrify -draw "gravity south text 0,20 '© Your Name'" *.jpg marks a whole folder in one command, and it will handle thousands of files without complaint, which beats any phone batch limit. GIMP does it with a script, and Lightroom's export watermark is the standard in professional workflows if you already pay for it. Snapseed can add text on a phone for free, one image at a time. The reason to use a phone tool is that the photos are on the phone and you want them marked before they go anywhere.

---

If the photos are on your phone and you want your name on them before they go up, stamp copies on the device and keep the unmarked originals somewhere separate.