Dual Transmission: Breakdown

A Look Under the Hood

Adam Sultanov
3 min readFeb 14, 2020

MICR breakdown, E-13B

The Magnetic Ink Character Recognition typeface is perfectly readable, if a little odd-looking. Here is a visual representation of the process of character recognition.

Broken down into its constituent parts,

condensed,

and interpreted into numbers.

Barcodes

One-dimensional barcodes operate like binary digits, either 0 or 1, on or off, in a straight line.

UPC

Breakdown of a UPC:

F LLLLL RRRRR C

F: the first number is a special indicator

LLLLL: the manufacturers code

RRRRR: the product code

C: the check digit, a form of error correction determined by putting all the previous numbers through a specific calculation

But how does the computer know left from right? Couldn’t it go in either direction?

A really cool part of this code is its method of self-orienting. A digit on the right is the inverse pattern of the same digit on the left. Designed specifically that each representation is a unique combination of two (variable-width) black stripes and two (variable-width) white stripes. If there are seven total ‘bars’ for each symbol, notice that all the left-hand numbers have an odd number of black bars, while all the numbers on the right have an even number of bars.

Another type of barcode is used in one form or another by many postal systems around the world.

The US used to use a system called POSTNET, which was eventually superceded by the Intelligent Mail barcode. They look similar but I think it’s an interesting development from a binary- to a quaternary-coded symbology.

POSTNET:

11000 00011 00101 00110 01001 / 01010 0110 10001 10010 10100

Intelligent Mail-type barcode:

0033 0213 0231 2013 2031 / 2211 0123 0303 0321 2103

This is actually the RM4SCC, a similar code from the UK, as IMb encoding is proprietary it seems. Even though the newer code is one less character long, the increased information density means it contains eight times the data of the old code.

The addition of two more symbols increases the information density of the code and allows for many more available characters (alphabetic instead of just numerical). The allows the codes to be shorter overall, as well as, presumably, increasing machine-reading speed.

00       10        01        11         Hmmm...

Next time: 2-D barcodes!

--

--