Convert any photo to the Ray-Ban Meta glasses format

A photo taken with Ray-Ban Meta smart glasses is an ordinary JPEG. What makes it recognisable is not a filter or a signature effect but two much simpler properties: the size of the frame and the block of metadata sitting in its header. Match both and any tool that inspects the file reports the same thing it would report about a genuine capture. Match neither and no amount of colour grading will help.

The frame: 3024 × 4032

The camera in the glasses produces stills at 3024 × 4032 pixels. That is a portrait 3:4 frame and it works out to 12.2 megapixels, which is the number that appears on the spec sheet. It is not an unusual size, since plenty of phones shoot 3:4, but it is a specific one, and it is the first thing that gives away a photo that was cropped by hand or exported at whatever size an editor happened to default to.

Getting to that frame from an arbitrary photo means making a choice, and it is worth understanding which one this tool makes. A source photo almost never has exactly a 3:4 ratio. There are three ways to reconcile the difference. You can stretch, which fills the frame and distorts everything in it. You can letterbox, which preserves the whole image and pads the edges. Or you can centre-crop, which preserves proportions and loses the edges. This converter centre-crops by default, because a stretched photo is obvious to the naked eye long before anyone opens the metadata, and white bars down the sides of a photo are not something a camera produces. If you want the whole frame preserved instead, the resizer on this site offers both options.

Orientation deserves a note of its own, because it is where most naive implementations break. A JPEG can store its pixels sideways and carry an orientation tag telling the viewer to rotate them. Every current browser applies that tag automatically when it decodes the image, so code that reads the tag and then rotates the canvas as well ends up rotating twice. This converter decodes the photo with its rotation already applied, renders upright pixels, and writes an orientation of 1 into the output. That value means "no further rotation", and it is the only correct one for a file whose pixels are already the right way up.

The header: what EXIF carries

EXIF is a block of tagged values that lives in the APP1 segment near the start of a JPEG. Cameras write it, phones write it, and almost every photo you have ever received carries one. It holds the make and model of the camera, the moment of capture, the exposure settings, and very often the coordinates of the place it was taken.

Two of those tags carry the device identity. Make is set to Meta AI, and Model to Ray-Ban Meta Smart Glasses 2. Those are the two values this converter writes, and they are the two an EXIF reader shows when it names the camera behind a photo. Alongside them the converter normalises the technical tags that have to agree with the pixels it just produced: colour space, orientation, and the pixel dimension fields that restate the frame size inside the header.

Everything else from your original file is carried across on purpose. The capture date, shutter speed, aperture and ISO of the photo you started from stay where they are. A header containing a camera name and nothing else reads as synthetic, because real captures are dense with exposure data. Preserving that data is what makes the output read as a photograph.

What is deliberately removed

The same header that identifies a camera also, routinely, identifies a person. Phone photos carry GPS coordinates precise enough to name a building. Many cameras write a body serial number. Editing software leaves its own name in the Software field, and lens tags name hardware that has nothing to do with a pair of glasses.

All of it is dropped. The GPS block is emptied wholesale rather than filtered tag by tag, because partial GPS data is still locating data: a latitude without its reference hemisphere still narrows the world down considerably. Serial numbers, owner and artist fields, the software and host computer fields, lens make, model and specification, and the maker note blob all go with it. The embedded thumbnail is removed too, and that one is easy to overlook: a JPEG can carry a small second copy of itself in its header, and on an edited photo that copy is frequently the original frame. Leaving it in place would ship the pre-conversion image inside the converted file.

Where the limits are

It is worth being precise about what this changes, because there is a persistent belief that EXIF is the key to Instagram's "Made with Meta glasses" label. It is not. That label is attached by the upload path. A photo that arrives in Instagram through the Meta AI app carries it, and one that arrives from your camera roll does not, whatever its header says. No tool that edits a file can produce it, and any tool claiming otherwise is describing something it cannot do.

What metadata does control is what every reader of the file reports: the camera name in your operating system's info panel, in Lightroom, in an EXIF viewer, in the properties dialog of a photo library. That is a real and checkable change, and it is the whole of what this tool claims.

Second limit: resolution is not detail. Rendering a 1200-pixel-wide photo into a 3024 × 4032 frame produces a file with 12.2 megapixels and the detail of a 1.4 megapixel one. The frame will be right and the picture will be soft. Start from the largest original you have.

Trademarks and affiliation

Ray-Ban is a trademark of Luxottica Group, and Meta and Meta AI are trademarks of Meta Platforms. This site is independent and is not affiliated with, endorsed by, sponsored by or connected to either company. Those names appear here to describe a photo format and the metadata values that identify it, which is the only way to describe accurately what this tool produces.