Convert JPG to BMP

This conversion has essentially one cause: a program or a device that will accept nothing else. Industrial and laboratory software that reads pixels directly. Point-of-sale and embedded equipment. Small display controllers on microcontrollers. Older Windows applications, and the occasional university assignment specifying the format precisely because it is simple to parse.

What those have in common is that they do not want to implement a decoder. A BMP is a header followed by rows of pixels, and code that reads it can be a page long. That is the format’s entire remaining value, and it is a real one in the places it applies.

It is worth stating plainly because the arithmetic invites the opposite conclusion. A JPG is lossy: detail was discarded when the file was created, according to a model of what the eye would not miss. Decoding it produces exactly what survived, and writing that into an uncompressed container preserves it faithfully.

So the output holds the same picture, artefacts and all, in a file many times the size. Nothing about the format restores what JPEG removed, and no setting anywhere changes that. What is bought is a file the receiving software will read.

A BMP stores every pixel as plain bytes with no encoding scheme wrapped around it, so the file size depends only on the dimensions — not on what the picture contains. A 3000 by 2000 photograph is around 18 megabytes; so is a 3000 by 2000 white rectangle.

That is unfamiliar if you are used to formats where a simple picture is a small file, and it explains why the only lever available is the pixel count. Reducing the maximum width genuinely halves or quarters the output; nothing about the content will.

The specification does allow a run-length encoded form, which would shrink flat graphics considerably. It is almost never used, and a large share of the software that reads BMP does not implement it — which is precisely the software that asked you for a BMP in the first place.

Writing the plain uncompressed form is therefore the right default rather than a limitation: it is the variant that every reader in existence accepts, which is the entire reason anybody wants the format.

BMP has essentially nowhere to record it. The camera model, the exposure settings, the date and — importantly — the GPS location that a phone writes into a JPG have no home in the format, so the converted file arrives without any of them.

That is a loss for a personal archive and a benefit for anything being handed to a machine or published. Either way it is worth knowing rather than discovering: if the date and location matter, keep the JPG as well, because the BMP will not carry them back.

A BMP is for the program that asked for it and nothing else. Email one and the recipient receives a file many times larger than it needed to be, which some mail systems will refuse outright. Put one on a web page and every visitor downloads the full uncompressed image.

For anything a person will receive, look at or store, JPG is the right file — and it is the file you already have. Convert at the boundary, feed the BMP to the software, and keep the original as the copy that travels.

Drop a whole folder and each file converts in turn, keeping its name with the extension swapped, with the results returned as one ZIP. Because nothing is uploaded there is no queue and no daily allowance.

The practical constraint here is unusual: BMP outputs are large, so a batch of fifty photographs produces close to a gigabyte, and it all has to fit in your browser’s memory while it is assembled. Converting in smaller groups is the sensible approach for a large set, which is not something most conversions on this site require.