1. How Fax Machines Split a Picture into Dots to Send It by Wire
Electric wires can only carry signals of varying strength. They cannot transport a physical picture itself. To solve this, facsimile machines divide a picture into a grid of tiny squares, known as pixels, and scan them one by one. The shade (light or dark) of each square is converted into an electrical signal for transmission [1]. In modern systems, a row of small light sensors reads the width of the paper at once. Then, the original document moves slightly so the machine can read the vertical direction, line by line [1]. Thinking of it as reading one line and stacking them makes the process easier to understand.
The main problem lies at the receiving end. The receiver must record the dots at the same speed and in the same position as the sender scanned them. If the timing is off, the picture will not return to its original form. One of the first people to tackle this difficult challenge was an inventor from Scotland.
2. Alexander Bain's 1843 Patent for Sending Pictures with Electricity and Two Pendulums
On May 27, 1843, 33 years before the telephone was invented, Alexander Bain of Scotland received a patent for a system to send pictures using electricity. This patent is considered the starting point of the modern facsimile [2]. Bain placed two pendulums, driven by clockwork, one at the sending end and one at the receiving end. As the pendulums swung, they traced a line of the picture. At the sender, a needle traced a picture made of metal pins, turning the electric current on and off to create a signal. At the receiver, the traces were recorded on chemically treated paper [2].
However, the two devices did not truly stay in sync. According to the English Wikipedia, the image quality was poor [2]. While the patent marks the origin, practical use required further improvements by later inventors. To match the swinging speed, what else was needed?
3. Giovanni Caselli's Pantelegraph: Sending Bank Signatures in 1865
Giovanni Caselli, an Italian physicist, succeeded in using pendulums to match the speed of transmission. On May 16, 1865, a commercial service called the Pantelegraph began using telegraph lines between Paris and Lyon. This happened about 11 years before the telephone was invented [4]. The system scanned characters written in non-conductive ink on a metal plate. At the receiving end, paper that darkened when electric current flowed through it recorded the traces. Pendulums were used to synchronize the movement of the sender and receiver [3].
Sending one sheet containing about 25 words took approximately 108 seconds [3]. The most common use was verifying that signatures were genuine in bank transactions [3]. Because it sent the shape of the writing rather than the meaning of the text, it was well-suited for checking signatures. When did pictures first travel across wires in Japan?
4. Japan's 1928 Photo Transmission System and the Coronation Photos
In Japan, Yasujiro Niwa and others developed the "NE-style" photo transmission system in 1928 (Showa 3). In November of that year, the Mainichi Shimbun newspaper company used this system to send photos of Emperor Showa’s enthronement ceremony from Kyoto to Tokyo, and it is reported that the transmission was successful [5]. Some records describe the route as "wired transmission between Osaka and Tokyo." Because descriptions of the route beyond Kyoto differ, the exact path is not stated definitively here [5].
This success was celebrated as a victory for domestic technology [5]. Rather than importing and using foreign technology, Japanese people used their own equipment to publish photos of a major ceremony in the newspaper. This was received as a proud achievement. When did the fax machine, originally a tool for newspapers, spread to homes and offices?
5. Why Sending a Fax Took Only One Minute: Sending Lengths of White and Black
Fax machines spread rapidly in Japan after the telephone network was opened up by a legal change in 1972 [1]. Around that time, usage for sending architectural drawings and proposals began [6]. Fax generations became faster over time. G1 (1968) and G2 (1976) were "3-minute faxes," taking about 3 minutes per page. The G3 standard in 1980 was called the "1-minute fax" [6].
The key to this speed is encoding that sends only the length of white and black sections. In text, white spaces often continue for long distances. Sending information as lengths, such as "100 white, 3 black," reduces the amount of data. This basic method is called MH encoding. A rough guideline on English Wikipedia gives an average MH compression ratio of about 20 for text pages, though it varies with the content [8]. In 1980, the MR method, which also uses differences from the neighboring line, was incorporated into an international standard (ITU-T Recommendation T.4), based on proposals from Japan’s KDD and Nippon Telegraph and Telephone Public Corporation [6]. The MR method is described as having about twice the compression performance of MH encoding [6].
This standardization was recognized as an IEEE Milestone in 2012 [7]. How are faxes recorded today?
6. How to Simulate a Fax Machine Yourself with Grid Paper
You can try a "fax game" using grid paper in your notebook. Draw a small picture using only white and black. Starting from the top line, write down the length of consecutive colors as numbers, like "3 white, 2 black, 1 white." Hand only these numbers to a family member and ask them to recreate the picture. Comparing pictures that need few numbers versus many helps you feel why text pages are shorter to send than pictures.
To learn more, the ETHW article (IEEE Milestone explanation) describes the history of G3 standardization in English. The NTT and KDDI news release describes the 2012 IEEE Milestone certification [7].