Free TIFF to BMP converter

TIFF is a container. One file extension holds many different ways of storing an image inside, and the tags placed at the start of the file say which one was used. Revision 6.0 of 3 June 1992 allows the data to sit uncompressed and allows PackBits, LZW, CCITT and JPEG compression, and next to that greyscale, RGB and CMYK colour with 8 or 16 bits per channel.[1] That is why two files with the same extension can look entirely different inside.

The format was developed in 1986 by Aldus Corporation together with Microsoft and Hewlett-Packard, and the rights later passed to Adobe.[2] That flexibility is the reason TIFF took hold in print shops and archives: one format serves both a black and white document scan and a photograph in full colour.

BMP works the other way round. It has one fixed layout: a header with the size of the file, a second header with the dimensions and the number of bits per pixel, and behind them an array of points written one after another.[3] This converter writes the most widely read variant: 24 bits per pixel, uncompressed, with a 40-byte BITMAPINFOHEADER.

Turning a TIFF into a bitmap therefore brings the image down to a single form. Whatever was inside the source file, what you get is an array of pixels that a program reads with no decoding library at all. The way back is handled by the BMP to TIFF converter.

The converter opens the TIFF file in your browser, moves the image onto a canvas and writes a bitmap out of it. Reading is handled by UTIF.js, a TIFF decoder written in JavaScript and running on your own computer.[4] Among browsers, TIFF files are displayed by Safari alone, so the decoder reads the file no matter what you open this page in.

The writing is done by our own encoder as well. A browser canvas does not hand back a BMP file: the MDN documentation describes this plainly — with a format the browser does not know you get a PNG file and no warning at all.[5] So we assemble the header and the array of points ourselves, and the image stays on your disk from the first click to the end of the work.

We checked how many ways of storing a TIFF pass through this decoder. Through the converter went files with no compression and files compressed with LZW, PackBits, Deflate, CCITT Group 4 and JPEG. On top of that came greyscale, 16 bits per channel, CMYK colour and a file with two pages. Every one of them came out as a bitmap of the same weight, because the weight of the result follows from the size of the image alone.

When you want a light file, the same image is written by the TIFF to JPG converter, and when every pixel has to stay, by the TIFF to PNG converter.

You can work out the weight of the finished bitmap before the conversion, and the result depends neither on how heavy the TIFF file was nor on what the image shows.

Every pixel takes three bytes: one each for red, green and blue. Every row of the image is padded with zeros to a full four bytes, and the Microsoft documentation gives a ready formula for that.[6] On top of all the rows come 54 bytes of header.

width in pixels × 3 B, rounded up to 4 B, × height + 54 B = weight of the BMP file

Let us follow that on an A4 page scanned at 300 dots per inch, which is 2480 by 3508 pixels. Start with a single row: 2480 pixels of three bytes make 7,440 bytes, and that number divides by four, so there is nothing to pad. There are 3,508 rows, so the pixels alone take 26,099,520 bytes. Together with the header the file weighs 26,099,574 bytes, which is 24.9 MB in the notation the tool uses in its summary.

The padding shows up only at widths that do not divide by four. A row of an image 1001 pixels wide carries 3,003 bytes of data, and the file reserves 3,004 bytes for it. Over a thousand rows that adds up to exactly a thousand extra bytes.

Which way the weight moves therefore depends on how the TIFF was stored. A scan compressed with CCITT Group 4 grows more than thirteenfold, a file with LZW compression by roughly two thirds, and an uncompressed CMYK file even gets lighter, because four printing inks turn into three colour components.

When you need one document out of the same scans instead of separate images, it is assembled by the TIFF to PDF converter.

A TIFF file can carry more than an array of points, so here is the shape each of those things takes on the way out of the converter. We checked all of it on our own test files.

The size of the image and the pixel values pass through unchanged. Storage at 16 bits per channel comes out at 8 bits per channel, the form the bitmap header describes. CMYK colour is converted by the decoder into RGB, so a file from a prepress studio opens as an ordinary screen image. Transparent areas come out white, because 24-bit storage describes every pixel as fully opaque. Out of a multi-page file the converter takes the first page, and you prepare each further page by saving it as a separate file first.

In the header of the finished bitmap we write a resolution of 96 dots per inch, given there as 3,780 pixels per metre. When the file is going to print at a different resolution, you set that in the graphics program you open the bitmap in.

An image that keeps its transparency is prepared by the TIFF to PNG converter, and an image for a web page by the TIFF to WebP converter.