Getting Your Own Photographs Onto a Samsung Frame TV

The Samsung Frame is a television that pretends to be a picture frame, and it’s very good at it. Naturally I wanted my own photographs on the wall rather than the stock art.

Naturally, that turned out to be harder than it should be.

Art Mode has real requirements it doesn’t document and doesn’t explain. Get them wrong and it rarely tells you — it just hangs partway through the upload, or takes the image and displays it in a way you didn’t intend. The debugging experience is “wait, then guess.” Samsung has a very conspicuous lack of documentation around this.

So I built the tool. It’s headed for an open-source release — not published yet, because I’m still tweaking the UI and installation — and this is what it knows.

What the Frame actually wants

sRGB JPEG. Non-progressive. Any other color space and you’ll get colors you didn’t shoot.

3840×2160, when the image is genuinely 16:9. That’s the TV’s natural 4K resolution.

Never a crop. This is the part I care most about. A tool that “fits” your image by cropping or stretching it has made an editorial decision on your behalf, and if you’re a photographer you already made that decision at the time. My tool preserves the native ratio and fits it inside the target box. Your composition survives contact with the television.

Where it fails silently

The genuinely nasty one: the TV hangs on WebP, GIF, HEIC and HEIF. Not rejects — hangs. The upload starts and never finishes and nothing tells you why.

Which matters more than it sounds like, because HEIC is what a lot of modern cameras and phones produce by default. Apple’s .HEIC and Canon’s .HIF files land right in this trap. You do everything correctly, you send the file the camera gave you, and the transfer quietly dies.

So the tool always transcodes to JPEG. Not conditionally, not per-format — always. It’s the one place I chose a blunt rule over a clever one, because the clever version has to be right every time and the blunt version can’t be wrong.

Inputs it will take: JPEG, PNG, TIFF, BMP, HEIC/HEIF (including Apple .HEIC and Canon .HIF), WebP, and GIF. Animated GIF and WebP contribute their first frame. Camera RAW files that it can’t decode are skipped with a clear message up front, rather than being accepted and failing mysteriously at upload — a small thing that saves a genuinely annoying half hour.

Using it

All of the above is what the app handles so you don’t have to. You never think about resolutions, file formats or colour profiles — you pick photos, and they show up on the wall.

One requirement up front: the TV and the computer running the app need to be on the same network segment (subnet). The Frame only agrees to pair with devices it considers local — if your network is split into VLANs, I explain the workaround in The TV That Would Only Talk to Its Neighbors.

The TV does need to be paired once, so it trusts the app. The app walks you through that the first time: you accept a prompt on the TV, and you never need to touch it again.

After that it’s a web page. Browse to your photos, pick one or a whole folder, see how each one will look matted, and send them. Re-sending a folder doesn’t fill your TV with copies of the same sunset — it knows what’s already there.

Compare that with the usual routes — fiddling with the TV remote, or Samsung’s SmartThings app on your phone. This is the same job in a few clicks, from a browser on your computer or phone.

Why bother open-sourcing it

Because the hard part wasn’t the code. The hard part was a week of finding out what the Frame silently dislikes, and there’s no reason for anyone else to spend that week.

None of the code is clever. The knowledge — that HEIC hangs rather than fails, that progressive JPEG isn’t accepted, that the panel is 3840×2160 and the ratio is yours to keep — is what actually cost something. That part should be free.

A future release will add a paid option to use mattes and colors beyond Samsung’s own limited set — additional mattes I’m making available myself — which is the natural next thing to want once your own work is up there.

The repo isn’t public yet. When it is, it’ll be under Cairo Labs on GitHub.

The projects, experience and opinions here are mine. AI helped me turn my notes and build records into this piece and polished it for Cairoglyphics.ai.