BMP to PDF converter

A BMP file stores an image point by point. Every pixel gets its own place in the file and its own colour value, with nothing shortened. The format was created for OS/2 and later became part of the Windows family.[1] In its most common variant a single pixel takes 24 bits, one byte each for red, green and blue. That gives 16.7 million shades.

PDF describes something else: a whole page together with its dimensions and everything standing on it. Adobe Systems created the format and released it on 15 June 1993.[2] Since July 2008 PDF has been an open international standard under the number ISO 32000-1, based on version 1.7 of the format.[2] The same document therefore lays out identically on every computer, phone and printer.

Our converter does both of those things in one step. It takes the image out of the BMP file and places it on an A4 page, and every uploaded file comes out as its own PDF document with a single page.

Drag your BMP files onto the Add files box, or click that box and pick the images from your disk. You can add several files to the queue at once. Then press the Convert button.

You can save the finished documents one at a time with the Download button next to a queue row, or all at once with the Download all button. The Clear all button empties the whole queue, and the Remove label next to a row takes a single file out of it.

The column on the right is headed Files in queue. Next to each image it shows the weight of that image and its status: Pending, Processing…, Done or Error. A progress bar sits above the buttons, and above it the In queue line together with the Completed counter.

The document itself is built by jsPDF, a library written in JavaScript and run on your own computer.[3] Its name then stays in the metadata of the document: the Producer field of our files holds the entry jsPDF 4.2.1, with the creation date next to it.

A format switcher labelled BMP to PDF sits under the converter. Clicking either of the two formats opens a list, and picking from it takes you to the converter page for that pair. We have also gathered every pair from this group on the Format conversion subcategory page.

Let us start with the weight of the incoming file, because that one can be worked out before the image is even opened. A single pixel takes three bytes, and every row of the image is padded with zeros up to a full four bytes; Microsoft documentation gives a ready formula for it.[4] On top of that comes a 54-byte header, made of 14 bytes describing the file and 40 bytes describing the image.[1]

Let us follow that through on a photo measuring 3,000 by 2,000 pixels. One row is 3,000 pixels at three bytes each, which is 9,000 bytes, and at that width nothing needs padding. There are 2,000 rows, so the pixels alone take 18,000,000 bytes. With the header the file weighs 18,000,054 bytes, which is 17.2 MB in the notation the tool queue uses.

On its way into the document the image goes through one extra step. The browser canvas returns it as a JPEG at quality 0.92, and that JPEG is what enters the PDF file.[5] The scale for that quality ends at 1, so the converter works close to its upper end.

JPEG compression splits the image into blocks and records above all the detail that the eye picks out best. Recommendation ITU-T T.81 from September 1992 sets out the rules of that encoding.[6] The encoding is lossy, meaning part of the image data is replaced by an approximation of it.[7] That is where the whole difference in weight between the incoming file and the finished document comes from.

The pixel count stays untouched through all of this. We looked inside the finished files: the 3,000 by 2,000 pixel photo is stored there as an image of exactly those dimensions, under the DCTDecode filter and in the DeviceRGB colour space. When the same image needs to keep every piece of data pixel for pixel, the BMP to PNG converter will prepare it.

Every document from this converter has one page in A4 format. The ISO 216 standard gives that sheet the dimensions 210 by 297 millimetres, which is 8.27 by 11.69 inches, and the ratio of its sides is 1 to the square root of two.[8]

Inside the file the page is described in points, because that is the unit PDF works with. One point is one seventy-second of an inch. An A4 page therefore measures 595.28 points across and 841.89 points down, and those are exactly the two numbers we found in the MediaBox field of our documents.

Page orientation follows the shape of the image. A photo wider than it is tall produces a landscape page, and every other image leaves it portrait. A margin of 20 points is left around the image on each side, so on a portrait page the image gets an area of 555.28 by 801.89 points. The converter fits it into that area whole and centres it.

An image smaller than that area keeps its own size. Our 400 by 300 pixel file landed on the page as a rectangle of 400 by 300 points, which is 5.56 by 4.17 inches. That works out at exactly 72 points per inch, because an inch holds precisely that many points in the measure PDF uses.

For a reader in the United Kingdom, Ireland or Australia the A4 page matches the paper already in the office printer. In the United States and Canada the default sheet is Letter at 8.5 by 11 inches, that is 215.9 by 279.4 millimetres,[8] so a printer driver there scales the A4 page at the moment of printing. Picking Fit to page in the print dialog gives the same picture on the shorter Letter sheet.