Bytes to MB Converter

bytes to megabytes — quick reference

The formula, worked line by line

One division by a million, or one by 1,048,576 if you want the binary answer. Both are given below, because the useful thing about starting from bytes is that you can produce either and say which is which.

That is the practical argument for recording byte counts in anything you build. A size stored as a rounded megabyte figure has thrown away both precision and provenance, and neither can be recovered later from the number itself.

megabytes = bytes / 1,000,000
mebibytes = bytes / 1,048,576
1 MB = 1,000,000 bytes
1 MiB = 2^20 = 1,048,576 bytes
1,474,560 bytes = 1.47456 MB = 1.40625 MiB

Why the byte is eight bits

The eight-bit byte is a convention that won rather than a law of nature, and the standards reflect that. ISO/IEC 2382 defines the octet as exactly eight bits and treats it as equivalent to an eight-bit byte, which is the unambiguous term. Machines with other byte widths existed; the DEC PDP-10 allowed bytes of any width up to its 36-bit word.

The two ladders, from the byte up

One consequence worth stating: if you are designing a limit, express it in bytes. A cap of 5,000,000 bytes cannot be misread, whereas a cap of "5 MB" will be interpreted differently by the person uploading and the server receiving, and the difference will land on somebody at the worst moment.

How to use the bytes to MB converter

A megabyte is a million bytes, so the conversion is six decimal places to the left. What makes this page worth reading is the direction it runs in. You have started from an exact byte count, which is the one figure on this whole subject that nobody can dispute, and that puts you in a position to settle the ambiguity rather than inherit it.

Byte counts come from machines that had no reason to round: a web server log, a storage API, the properties panel of a file. Because the count is exact, you can check it against both scales and see which one produced it. That is not possible once a figure has been rounded to "4.7 MB" and the divisor has been lost.

The test is quick. Divide the count by 1,048,576. If the answer is a clean number — a whole one, or a half, or a quarter — the figure came from a binary system. If dividing by a million gives the clean answer instead, it came from a decimal one. Very few counts look tidy on both scales, so the answer is usually unambiguous.

Byte counts also protect you from a subtler problem: rounding that has already happened. A figure reported as 4.7 MB could be anything from 4,650,000 to 4,749,999 bytes, a spread of a hundred thousand. When you are comparing against a hard limit, that spread is often wider than the margin you were relying on.

The 1.44 MB diskette, which was neither

The best illustration of how tangled the labelling became is a format most people have handled. The high-density 3.5-inch diskette was sold as 1.44 MB. Its actual capacity is 2,880 sectors of 512 bytes, which is 1,474,560 bytes. That figure is 1.475 million bytes and it is 1.406 mebibytes. It is not 1.44 of either.

It is a useful thing to remember when a figure seems not to reconcile on either scale. Hybrid units are rare but they are not unheard of, and the only way to detect one is to work back to the byte count and see which divisors were applied in what order.

Reading a byte count in the wild

Web server logs record response size in bytes, which is why bandwidth reports built from them are exact rather than estimated. Storage APIs return byte counts for the same reason. So do file properties dialogs, if you look past the friendly rounded figure at the top to the exact number underneath.

That last point catches people. A file listing usually shows two numbers: the size of the content and the space it occupies once the filesystem has allocated whole clusters to it. A small file can occupy several times its own size. Upload limits measure content, so the smaller of the two is the one to convert.

Every bytes value, worked out

67 common bytes figures each get their own page, with the answer at full precision, the arithmetic, what rounding costs, and the nearest real-world reference point on the scale.

Questions people ask