Number Systems and Bases
Base systems like binary and hexadecimal seem a bit strange at first. The key is understanding how different systems “tick over” like an odometer when they are full. Base 10, our decimal system, “ticks over” when it gets 10 items, creating a new digit. We wait 60 seconds before “ticking over” to a new minute. Hex and binary are similar, but tick over every 16 and 2 items, respectively.
Try converting numbers to hex and binary here:
Way back when: Unary Numbers
Way back in the day, we didn’t have base systems! It was uphill both ways, through the snow and blazing heat. When you wanted to count one, you’d write:
l
When you wanted 5, you’d write
lllll
And clearly, 1 + 5 = 6
l + lllll = llllll
This is the simplest way of counting.
Enter the Romans
In Roman numerals, two was one, twice. Three was one, thrice:
one = I
two = II
three = III
However, they decided they could do better than the old tradition of lines in the sand. For five, we could use V to represent lllll and get something like
l + V = Vl
Not bad, eh? And of course, there are many more symbols (L, C, M, etc.) you can use.
The key point is that V and lllll are two ways of encoding the number 5.
Give each number a name
Another breakthrough was realizing that each number can be its own distinct concept. Rather than represent three as a series of ones, give it its own symbol: “3″. Do this from one to nine, and you get the symbols:
1 2 3 4 5 6 7 8 9
The Romans were close, so close, but only gave unique symbols to 5, 10, 50, 100, 1000, etc.
Use your position
Now clearly, you can’t give every number its own symbol. There’s simply too many 
But notice one insight about Roman numerals: they use position of symbols to indicate meaning.
IV means “subtract 1 from 5″
and VI means “add 1 to 5″.
In our number system, we use position in a similar way. We always add and never subtract. And each position is 10 more than the one before it.
So, 35 means “add 3*10 to 5*1″ and 456 means “4*100 + 5*10 + 6*1″.
Our choice of base 10
Why did we choose to multiply by 10 each time? Most likely because we have 10 fingers.
One point to realize is you need enough digits to “fill up” until you hit the next number. Let me demonstrate.
If we want to roll the odometer over every 10, so to speak, we need symbols for numbers one through nine; we haven’t reached ten yet. Imagine numbers as ticking slowly upward - at what point do you flip over the next unit and start from nothing?
Enter zero
And what happens when we reach ten? How do we show we want exactly one “ten” and nothing in the “ones” column?
We use zero, the number that doesn’t exist. Zero is quite a concept, it’s a placeholder, a blank, a space, and a whole lot more. Suffice it to say, Zero is one of the great inventions of all time.
Zero allows us to have an empty placeholder, something the Romans didn’t have. Look how unwieldly their numbers are without it.
George Orwell’s famous novel “1984″ would be “MCMLXXXIV”! Rolls right off the tongue, doesn’t it?
Considering other bases
Remember that we chose to roll over our odometer every ten. Our counting looks like this:
1
2
3
4
5
6
7
8
9 (uh oh, I’m getting full!)
10 (ticked over - start a new digit)
What if we ticked over at 60 when we counted, like we do for seconds and minutes?
1 second
2
3
4
5
…
58
59
1:00 (60 seconds aka 1 minute. We’ve started a new digit.)
Everything OK so far, right? Note that we use the colon (:) indicate that we are at a new “digit”. In base 10, each digit can stand on its own.
Try Base 16
If we want base 16, we could do something similar:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15 (we’re getting full)
1:00 (16 - we’ve started a new digit)
However, we don’t want to write hexadecimal numbers with the colon notation (even though we could). We’d rather cook up separate symbols for 10-15 so we can just write numbers like we’re used to. We’ve run out of numbers (1-9 already used, with 0 as a placeholder) so we need some other symbols. We could use some squiggly lines or other shapes, but the convenions is to use letters, Roman style. Just like 5 became V, programmers use letters A-F to get enough digits up to 16. That is,
1
2
3
4
5
6
7
8
9
A (10 - we’re using the symbol “A”)
B (11)
C (12)
D (13)
E (14)
F (15 - uh oh, we’re getting full)
10 (16 - we start a new digit)
Ahah! Now we can use one digit per “place”, and we know that 10 actually means we’ve “ticked over to 16″ once.
20 means we’ve ticked over to 16 twice (32).
25 means we’ve ticked over to 16 twice (giving us 32) and gone an extra 5. The total is 32 + 5 = 37.
Quick review
With me so far? This is pretty cool, right? We can count in any system we want. Also notice that base 16 is more “space efficient” in the sense we can write a number like 11 in a single digit: B.
Base 16 really isn’t that different from base 10, we just take longer to fill up.
The wonderful world of binary
We’ve seen plenty of base systems, from over-simple unary, to the unweidly Roman numerals, the steady-going base 10 and the compact base 16.
What’s great about binary? In the spirit of keeping things simple, it’s the simplest number system that has the concept of “ticking over”. Unary, where we just write 1, 11, 111… just goes on forever. Binary, with two options (1 and 0) looks like this:
1: 1
2: 10 (we’re full - tick over)
3: 11
4: 100 (we’re full again - tick over)
5: 101
6: 110
7: 111
8: 1000 (tick over again)
…
and so on.
Because binary is so simple, it’s very easy to build in hardware. You just need things that can turn on or off (representing 1 and 0), rather than things that have 10 possible states (to represent decimal).
Because it’s so simple, binary is also resistant to errors. If your signal is “partially on” (let’s say 0.4), you can assume that’s a zero. And if it’s mostly on (say 0.8), then you can assume it’s a 1. If you’re using a system with 10 possible states, it’s difficult to tell when an error occurred. This is one reason digital signals are so resilient to noise.
Other examples of bases
We use other bases all the time, even dynamically changing bases. We usually don’t think of it that way:
Hours, minutes, seconds: 1:32:04
- We know this is 1 hour, 32 minutes, 4 seconds. In seconds, this is 1*60*60 + 32*60 + 4.
Feet and inches: 3′ 5″
- This is 3 feet, 5 in or 3 * 12 + 5 inches.
Pounds and ounces: 8 lbs, 5 oz
- Since a pound is 16 oz, This is 8 * 16 + 5 oz. We’ve been using a base 16 number system all along!
Parting thoughts
“10″ in any number system indicates the base, and means we’ve ticked over once. 10 in binary means two, 10 in decimal means ten, and 10 in hexadecimal is sixteen.
How do you keep these numbers apart? Programmers will often write “0b” in front of binary numbers. So 2 in binary is
0b10
Similarly, they’ll write 0x in front of hex numbers. So 16 in hex is:
0×10
If there aren’t any symbols (0b or 0x) in front, we assume it’s base 10, a regular number.
Now go forth and enjoy your new knowledge!
Tags: bases, binary, decimal, hexadecimal, number
Related Posts:
RSS
11 Comments »
Trackbacks & Pingbacks
-
Pingback by The Quick Guide to GUIDs | BetterExplained — April 18, 2007 @ 11:03 pm
-
Pingback by Another Look at Prime Numbers | BetterExplained — September 9, 2007 @ 11:32 am
-
Pingback by Me’s blog » GUID — March 25, 2008 @ 2:24 am
Comments
RSS feed for comments on this post. TrackBack URI
Leave a comment
Have a question? Know an explanation that caused your own a-ha moment? Write about it here.
wow. thanks, i feel smarter albeit nerdier already.
Jeff — January 11, 2007 @ 10:37 pm
hey, this is a wonderful site, its really helped me in my assignment in college, but there is a question that i can’t answer and that is, Why do we use different number bases
Alex Heap — March 25, 2007 @ 3:55 am
Hi Alex, good question. I think we use binary (base 2) in computer systems because they are the easiest to build. It’s easier to make a switch that turns on or off (1 or 0) rather than one that has to go between 10 states.
For our choice of base 10, it’s probably because we have 10 fingers. Though some ancient civilizations used base 60, base 10 is pretty natural to us as we count off items on our hands.
Kalid — March 25, 2007 @ 12:51 pm
i want to know the history of the number system o to 9, when and why did it start
jacki forster — May 15, 2007 @ 6:10 am
Hi Jacki, I don’t know that much about the actual history, but there’s some good information here:
http://en.wikipedia.org/wiki/Decimal#History
Kalid — May 17, 2007 @ 7:28 am
Negative bases can be used, too. For example, base-negative-10. 0-9 are as in base-10, but 10[-10] is -10, 11[-10] is -9, …, 99[-10] = 81, 100[-10] = 100 (because -10 x -10 = 100). Seems silly, but I actually read a paper once in which using base-negative-two was useful for reducing the size of some electronic circuits.
David Mercer — September 10, 2007 @ 11:47 am
Hi David, thanks for the info, that’s pretty interesting. I searched around and think it’s called “negabinary”: http://en.wikipedia.org/wiki/Negabinary
and there are some nice features like not having to store a separate sign bit. There’s even a number system with imaginary base, which is a bit mind-boggling:
http://en.wikipedia.org/wiki/Quater-imaginary_base
Appreciate the pointers!
Kalid — September 10, 2007 @ 1:19 pm
this place is cool and helpful at the same time!!
Anonymous — December 12, 2007 @ 12:54 pm