A friend sends you a folder of holiday photos from their new camera, and every file ends in .NEF, .CR3 or .ARW. Your computer shows a blank icon, the photo app shows something dark and flat, or nothing opens at all. These are RAW files, and they are not quite photos yet.
What is inside a RAW file
A camera sensor does not see colour. Each photosite measures only how much light hit it, and a grid of red, green and blue filters laid over the sensor, usually the Bayer pattern with two green filters for every red and blue, means each one measured only one colour. A RAW file stores those measurements, typically at 12 or 14 bits per photosite, along with notes about how the camera was set.
Turning that into a picture takes several steps, together called developing:
- Demosaicing: estimating the two missing colours at every pixel from its neighbours.
- White balance: deciding what counts as neutral grey under that light.
- Tone and colour: applying a contrast curve and the camera maker's colour rendering.
- Sharpening and noise reduction.
Every one of those is a decision, which is exactly why photographers shoot RAW. A blown sky or a wrong white balance can often be recovered from the extra data in a RAW file, where a JPG has already thrown it away.
Why there are so many RAW formats
There is no single RAW format. Each manufacturer has its own: NEF for Nikon, CR2 and CR3 for Canon, ARW for Sony, RAF for Fujifilm, ORF for Olympus and OM System, RW2 for Panasonic, PEF for Pentax. Adobe's DNG is an openly documented alternative used by some cameras, many phones in their pro modes, and by people who convert their archives for long-term safety. Each new camera model can change the details, which is why editing software regularly needs updates just to open a new camera's files.
The JPG that is already in there
Here is the useful part. Almost every camera also stores a finished JPG inside the RAW file. It is what the camera shows on its own screen when you review a shot, rendered with whatever picture style, white balance and sharpening the camera was set to. On most modern cameras that embedded preview is full size, the same pixel dimensions as the sensor.
So if what you want is simply to see, share or print the photos as they looked on the camera, you do not need to develop anything. You need that preview. The RAW to JPG converter finds it, picks the largest one in the file, and saves it as JPG, PNG or TIFF, in the browser. It reads DNG, NEF, CR2, CR3, ARW, RAF, ORF, RW2, PEF and several older formats.
When you need real editing instead
The embedded preview is the camera's interpretation, fixed at the moment of shooting. If the white balance is wrong, the shadows are too dark, or you want to pull detail back from a bright sky, extracting the preview will not help, because those decisions are already baked into it. That is when a RAW developer earns its place: Adobe Lightroom and Capture One on the commercial side, darktable and RawTherapee as free and open-source options. They go back to the sensor data and let you make every decision again.
A sensible workflow for most people is both: extract previews to sort and share a shoot quickly, then open only the handful of photos that deserve it in a proper RAW editor.
Before you share
Full-size camera JPGs are large, often 20 to 40 megapixels and several megabytes each. For email, chat or a website, resize them to the size they will actually be viewed at, and compress if needed. RAW files and their previews also carry detailed camera metadata, and photos from cameras with GPS or a paired phone can carry location; the image metadata viewer shows what is there before you send anything.