WEBVTT - The Birth of the Modem 0:00:04.400 --> 0:00:07.760 Welcome to text Stuff, a production from I Heart Radio. 0:00:12.119 --> 0:00:15.000 Hey there, and welcome to tech Stuff. I'm your host, 0:00:15.240 --> 0:00:18.400 Jonathan Strickland. I'm an executive producer with I Heart Radio 0:00:18.520 --> 0:00:22.159 and I love all things tech. And when I was 0:00:22.200 --> 0:00:26.800 growing up, we had a couple of computers at my house. 0:00:27.320 --> 0:00:30.000 One was an Apple to E that was our our 0:00:30.040 --> 0:00:34.120 first real personal computer, and the other, which came later, 0:00:34.520 --> 0:00:39.040 was an IBM Compatible. And at that point, I think 0:00:39.040 --> 0:00:41.839 I was almost a teenager or maybe a young teenager 0:00:41.840 --> 0:00:44.720 when we got the IBM Compatible. I don't remember exactly 0:00:45.440 --> 0:00:51.200 when my dad got our first modem, or whether it 0:00:51.320 --> 0:00:54.600 was on the Apple to E or on the two 0:00:54.600 --> 0:00:57.280 eight six IBM Compatible or one of the later ones, 0:00:57.360 --> 0:00:59.840 because we got a three eight six from there. But 0:01:00.080 --> 0:01:04.120 one thing I do remember really well was a particular sound. 0:01:19.280 --> 0:01:21.280 Any of you guys out there around my age or 0:01:21.280 --> 0:01:24.880 so we'll remember that noise too. Some of you younger 0:01:24.920 --> 0:01:28.120 folks might have heard it before, and a few of 0:01:28.160 --> 0:01:31.320 you might still use one. I mean, they're not obsolete, 0:01:31.440 --> 0:01:34.880 but they are the exception rather than the rule. And 0:01:34.959 --> 0:01:38.400 to the rest of you that probably sounds entirely foreign, 0:01:38.760 --> 0:01:42.320 and it is the sound of a dial up modem 0:01:42.360 --> 0:01:45.039 making a connection. So in today's episode, I'm going to 0:01:45.080 --> 0:01:49.480 talk about modems, specifically dial up ones. Before I launch 0:01:49.560 --> 0:01:53.720 into my trademarked history section, let's talk about what a 0:01:53.880 --> 0:01:59.560 modem actually does. A modem's purpose is to facilitate communication 0:02:00.000 --> 0:02:04.960 between computers, and the word modem is a portmanteau, which 0:02:05.040 --> 0:02:08.680 means I get to revel in my background in English lit. So, 0:02:08.720 --> 0:02:11.840 a portmanteau is a word that combines both the meaning 0:02:12.280 --> 0:02:16.760 and the sounds of two or more other words, and 0:02:16.800 --> 0:02:21.639 typically we do this to convey a particular concept or meaning. So, 0:02:21.720 --> 0:02:26.600 for example, brunch, something I miss dearly during this time 0:02:26.600 --> 0:02:31.280 of isolation, is a portmanteau. It combines the sounds from 0:02:31.320 --> 0:02:35.520 the words breakfast and lunch, and it also combines the 0:02:35.760 --> 0:02:39.360 meaning of those two words. Brunches a meal that occurs 0:02:39.480 --> 0:02:43.120 later than breakfast but earlier than lunch, and typically the 0:02:43.200 --> 0:02:46.520 menu for brunch includes foods that you might encounter at 0:02:46.520 --> 0:02:52.000 either of those other meals. So what words and ideas 0:02:52.040 --> 0:02:55.639 combine to form the word modem, Well, that would be 0:02:55.800 --> 0:03:00.360 modulation and demodulation. I'll explain that a little bit later 0:03:00.400 --> 0:03:04.040 in this episode, but first let's dive into some history 0:03:04.080 --> 0:03:08.160 to understand the genesis of modems for computers. It actually 0:03:08.200 --> 0:03:11.400 helps for us to go even further back to talk 0:03:11.440 --> 0:03:15.799 about the telegraph, which I know sounds crazy, but hear 0:03:15.840 --> 0:03:20.960 me out way back in the mid nineteen century. So 0:03:21.040 --> 0:03:25.520 the mid eighteen hundreds, inventors, including the famous Samuel Morse, 0:03:25.960 --> 0:03:29.320 we're developing a method of communication that depended upon sending 0:03:29.400 --> 0:03:35.440 electric signals over wires, and it was ingenious in its simplicity. 0:03:35.640 --> 0:03:38.480 So on one end, you've got a switch. If you 0:03:38.560 --> 0:03:41.920 close the switch, you let current pass through a circuit 0:03:42.120 --> 0:03:45.200 into a wire. If you open a switch, you cut 0:03:45.200 --> 0:03:48.520 off that path. The current cannot flow through. Now, this 0:03:48.640 --> 0:03:53.280 particular switch was spring loaded. If you pushed down on 0:03:53.320 --> 0:03:58.000 a little button, you closed the switch. If you let 0:03:58.040 --> 0:04:00.960 off pressure, if you took your hand the way, it 0:04:00.960 --> 0:04:04.320 would spring back into the open position. So it was 0:04:04.360 --> 0:04:08.440 only closed if you held it closed. On the opposite 0:04:08.480 --> 0:04:12.640 side of that wire was the receiver, and the receiver 0:04:13.160 --> 0:04:16.560 had an electro magnet in it, so when current flowed 0:04:16.680 --> 0:04:19.719 to the electro magnet, it would generate a magnetic field. 0:04:19.839 --> 0:04:22.640 That's all electro magnets were, right. The current flows through, 0:04:22.920 --> 0:04:26.160 it generates a magnetic field. This magnetic field would then 0:04:26.279 --> 0:04:30.359 pull down a ferromagnetic lever, And so if you hold 0:04:30.400 --> 0:04:33.839 down the switch on the sending end, the lever on 0:04:33.920 --> 0:04:38.400 the receiving end would stay down because the electromagnetic force 0:04:38.400 --> 0:04:40.840 would continue to pull on it. If you let go 0:04:40.880 --> 0:04:43.200 of the switch, the lever on the opposite side would 0:04:43.240 --> 0:04:47.640 pop back up once that electro magnetic attraction stopped. You 0:04:47.640 --> 0:04:51.279 can also have it just be some sort of buzzer 0:04:51.480 --> 0:04:54.840 or alarm. So you hold down the switch and it 0:04:54.920 --> 0:04:58.720 makes an electro magnet sound, a bell or a buzzer. 0:04:59.320 --> 0:05:03.800 Now that by itself is not terribly useful, but one 0:05:03.800 --> 0:05:08.159 thing you could do is establish a pattern and determine 0:05:08.160 --> 0:05:11.080 what meaning there is in the pattern. So, for example, 0:05:11.120 --> 0:05:13.040 if you were to hold down the switch for a 0:05:13.080 --> 0:05:17.280 little bit, that might be a dash. If you just 0:05:17.720 --> 0:05:20.640 tapped the switch, you would get a dot. And then 0:05:20.640 --> 0:05:25.360 by establishing what dots and dashes mean, you could build 0:05:25.360 --> 0:05:28.240 out a code. That's exactly what Morse did. He developed 0:05:28.279 --> 0:05:33.839 his eponymous code, the Morse Code, to encode characters into 0:05:33.839 --> 0:05:37.360 collections of dots and dashes. You could take a message 0:05:37.400 --> 0:05:42.000 written in a language like English. You convert every letter 0:05:42.120 --> 0:05:45.919 in that message into its respective dots and dashes. You 0:05:46.000 --> 0:05:49.719 send it off by telegraph by tapping this message out 0:05:49.800 --> 0:05:52.839 on the sending end, and an operator on the receiving 0:05:52.960 --> 0:05:57.680 end would listen to those taps, knowing whether it's a 0:05:57.720 --> 0:06:00.720 dot or a dash. They would take down the message 0:06:00.760 --> 0:06:04.159 and to code it back into its original language. You know, 0:06:04.160 --> 0:06:05.840 if they were really good, they could do it letter 0:06:05.920 --> 0:06:08.839 by letter. Otherwise they were actually writing down dots and dashes. 0:06:09.320 --> 0:06:11.920 A little bit later, you had engineers who created a 0:06:12.040 --> 0:06:15.880 version where the receiving end had a lever where it 0:06:16.080 --> 0:06:18.760 ended with a little wheel that was coded in ink, 0:06:19.600 --> 0:06:23.000 and there was paper tape that would constantly move at 0:06:23.000 --> 0:06:26.040 a speed built neath the lever. So when the sender 0:06:26.120 --> 0:06:30.640 pressed the sending key down, the receiving key would come 0:06:30.680 --> 0:06:33.560 down and that wheel would make contact with the paper. 0:06:33.880 --> 0:06:36.960 And then based upon the length of the switch press, 0:06:37.000 --> 0:06:39.440 you would get a physical dot or dash. You would 0:06:39.440 --> 0:06:43.960 actually get a printed version of the coded message which 0:06:44.000 --> 0:06:47.479 you could then decode. Then thus you would say the 0:06:47.560 --> 0:06:50.200 transcription step where someone would have to take down the 0:06:50.240 --> 0:06:53.120 dots and dashes by hand. Now trust me, this is 0:06:53.120 --> 0:06:55.080 all going to lead into modems, but we do have 0:06:55.120 --> 0:06:56.960 a couple of other stops. We have to make first. 0:06:57.400 --> 0:06:59.320 There was a lot of work in the field of 0:06:59.440 --> 0:07:03.520 teleprint ters in the nineteenth century, including stuff like stock 0:07:03.600 --> 0:07:06.719 tickers that would take in an electric signal and then 0:07:06.800 --> 0:07:10.160 print out results based on that. But we're gonna jump 0:07:10.200 --> 0:07:12.080 over all of that for right now to get to 0:07:12.120 --> 0:07:16.480 the early twentieth century. In nineteen o two, an electrical 0:07:16.560 --> 0:07:20.400 engineer named Frank Peern was experimenting with a method to 0:07:20.600 --> 0:07:24.400 create a printing telegraph system that would let people send 0:07:24.640 --> 0:07:29.880 and receive text messages, essentially using typewriter like devices that 0:07:29.920 --> 0:07:32.600 would connect to each other via wires. You could have 0:07:33.040 --> 0:07:36.679 dedicated wires between two of these things, or you could 0:07:36.760 --> 0:07:41.440 have them tap into some other wired system. He encountered 0:07:41.520 --> 0:07:44.000 some challenges along the way, and after a few years 0:07:44.040 --> 0:07:46.960 of frustration, Peern decided he's going to piece out of 0:07:47.000 --> 0:07:50.240 this whole endeavor. But then his work was carried on 0:07:50.280 --> 0:07:54.840 by another engineer named Charles Crumb. The goal was again 0:07:54.880 --> 0:07:58.680 to make an automated system that could receive incoming messages 0:07:59.040 --> 0:08:02.960 and print them in alpha numeric characters on paper, so 0:08:03.120 --> 0:08:05.720 instead of getting those dots and dashes, you would get 0:08:05.800 --> 0:08:09.360 the original message. As it was intended, it was skipping 0:08:09.480 --> 0:08:13.040 the whole encoding decoding step. You wouldn't have to put 0:08:13.040 --> 0:08:16.440 a message into morse code and then decode it. At 0:08:16.480 --> 0:08:18.680 least it was skipping in as far as the human 0:08:18.760 --> 0:08:22.120 operators were concerned. It was taking all of those sort 0:08:22.160 --> 0:08:27.760 of intermediate steps required by traditional telegraphy and automating them. 0:08:27.800 --> 0:08:30.760 But how well, it comes back to the idea of 0:08:30.880 --> 0:08:34.360 using electric pulses to indicate a letter, So it really 0:08:34.360 --> 0:08:38.800 comes back to that closed and open switch description. The 0:08:38.880 --> 0:08:43.199 engineers coded each character as a series of five potential 0:08:43.280 --> 0:08:47.720 electric pulse states, and the states were either on meaning 0:08:47.760 --> 0:08:51.079 current was flowing through the circuit at that instant, or 0:08:51.240 --> 0:08:54.960 off meaning no current was flowing. And they represented each 0:08:55.040 --> 0:08:59.240 letter as a sequence of either on or off, and 0:08:59.720 --> 0:09:03.040 that's sequence would include five pulses. They referred to the 0:09:03.280 --> 0:09:07.720 on current state as marking and the off current state 0:09:07.840 --> 0:09:12.520 as spacing. So the letter D, for example, would code 0:09:12.600 --> 0:09:18.000 into mark space space mark space, meaning the current would 0:09:18.040 --> 0:09:23.360 be on, off, off, on off as its sequence of 0:09:23.400 --> 0:09:28.640 five pulses. The receiving machine would interpret the incoming signals 0:09:28.679 --> 0:09:31.559 as letters. It would say, all right, we're looking at 0:09:31.600 --> 0:09:34.160 this span of time and we see that the current 0:09:34.320 --> 0:09:40.040 is on off, off, on, off. That's the letter D. 0:09:40.640 --> 0:09:42.720 The code itself dated all the way back in the 0:09:42.760 --> 0:09:47.520 eighteen seventies. It was invented by Emil Baudoux Bado that's 0:09:47.640 --> 0:09:52.000 b a U d O. T made numerous contributions to 0:09:52.040 --> 0:09:55.720 the field, including a method of multiplexing, that is, being 0:09:55.720 --> 0:10:00.000 able to send multiple messages simultaneously through a careful system 0:10:00.040 --> 0:10:04.280 of clockwork switches so that the correct information would go 0:10:04.400 --> 0:10:08.040 to the correct teleprinter. His work was important enough that 0:10:08.120 --> 0:10:11.559 we take his name Baudo, and we created the unit 0:10:11.800 --> 0:10:15.839 for data transmission with modems, the baud b a U D. 0:10:16.320 --> 0:10:19.000 More on that in just a moment. But this approach 0:10:19.160 --> 0:10:24.199 required one other very important component. The sending and receiving 0:10:24.240 --> 0:10:28.640 machines had to be quote in step, meaning they had 0:10:28.640 --> 0:10:32.040 to be synchronized with each other. If they weren't, then 0:10:32.120 --> 0:10:36.200 the receiving machine might start interpreting signals out of step, 0:10:36.360 --> 0:10:39.959 out of sync with the sending machine, and thus print 0:10:40.040 --> 0:10:44.200 the wrong letters. Because remember every character coded into a 0:10:44.320 --> 0:10:47.680 signal of five pulses. If the receiving machine is off 0:10:47.720 --> 0:10:51.600 by even one pulse, it's going to misinterpret what's being 0:10:51.720 --> 0:10:54.040 sent to it. So if you were sending a message 0:10:54.080 --> 0:10:56.920 that was five letters long. Those five letters would be 0:10:56.960 --> 0:11:02.120 represented by twenty five short pulses, five pulses each, and 0:11:02.160 --> 0:11:04.880 the first letter would be pulses one through five, the 0:11:04.920 --> 0:11:08.320 second letter would be pulses six through ten, and so on. 0:11:08.679 --> 0:11:12.080 But let's say the receiving machine only starts to pick 0:11:12.160 --> 0:11:15.319 up at pulse three. It misses the first two pulses. 0:11:16.160 --> 0:11:20.679 It thinks pulse three is actually pulse one, which means 0:11:21.040 --> 0:11:24.400 it's going to use the pulses three through seven to 0:11:24.520 --> 0:11:27.600 be the first letter, and the pulses eight through twelve 0:11:27.679 --> 0:11:29.760 will be the second letter, and so on, with the 0:11:29.760 --> 0:11:33.120 final letter having two spaces at the end because no 0:11:33.280 --> 0:11:36.959 more signal is being sent. While the engineers were able 0:11:37.000 --> 0:11:39.640 to create a way to send and receive signals, it 0:11:39.720 --> 0:11:43.520 was this synchronization stuff that would be a really big challenge. 0:11:43.520 --> 0:11:46.640 They had to make sure that both machines knew quote 0:11:46.679 --> 0:11:51.280 unquote when messages were coming through, so that any errors 0:11:51.320 --> 0:11:54.720 that were detected would be known and it wouldn't just 0:11:54.840 --> 0:11:58.520 print gibberish. The solution turned out to be fairly elegant. 0:11:58.880 --> 0:12:03.720 Howard Crumb, the son of Charles Crumb, proposed that every 0:12:03.840 --> 0:12:09.120 code combination representing a letter would first be preceded by 0:12:09.200 --> 0:12:13.839 a start pulse and followed by a stop pulse. So 0:12:13.960 --> 0:12:16.680 every single letter would have a special pulse at the 0:12:16.760 --> 0:12:19.160 very beginning and the very end of that sequence of 0:12:19.240 --> 0:12:22.360 five to indicate this is the beginning of a letter, 0:12:22.640 --> 0:12:25.520 this is the end of a letter, ignore anything else. 0:12:25.960 --> 0:12:28.760 That way, the receiving machine would detect the true beginning 0:12:28.880 --> 0:12:32.080 and the true end of each combination, and would more 0:12:32.160 --> 0:12:35.080 likely print the appropriate letter. It was a simple way 0:12:35.120 --> 0:12:38.920 to achieve full synchronization, though there was still no error 0:12:38.960 --> 0:12:41.960 correction at this point. Now I'll have to dedicate a 0:12:41.960 --> 0:12:45.040 full episode to teletype and the advancements that made it possible, 0:12:45.360 --> 0:12:48.480 because there's a whole lot more than that brief overview 0:12:48.480 --> 0:12:51.040 I just gave. But what does this all have to 0:12:51.120 --> 0:12:55.880 do with modems. Well, the process of encoding information and 0:12:55.920 --> 0:13:00.160 sending it through a signal is modulation. The process us 0:13:00.240 --> 0:13:04.200 of receiving a signal and decoding that to get at 0:13:04.240 --> 0:13:08.680 the original information is demodulation. The same process is used 0:13:08.679 --> 0:13:12.520 in lots of other technologies like radio. With radio signals, 0:13:12.760 --> 0:13:15.880 you take a carrier wave, that is a radio wave 0:13:16.200 --> 0:13:20.200 with a consistent amplitude and a consistent frequency. It would 0:13:20.240 --> 0:13:22.599 just be a pure tone if you could hear it, 0:13:22.640 --> 0:13:26.080 if it were in the range of human hearing. So 0:13:26.120 --> 0:13:30.679 you get this pure radio wave. Now you encode information 0:13:30.960 --> 0:13:35.360 by laying on another wave, by altering that wave in 0:13:35.400 --> 0:13:37.960 some way, you can change the amplitude. If you do 0:13:38.040 --> 0:13:41.760 it through those changes, you get what is amplitude modulation 0:13:42.000 --> 0:13:45.239 or a M radio. Or you can change the frequency 0:13:45.720 --> 0:13:49.760 and you get frequency modulation or FM radio. The receiver 0:13:49.840 --> 0:13:53.400 on the other end follows the same process in reverse. 0:13:53.960 --> 0:13:56.000 It takes these waves in and then converts them back 0:13:56.000 --> 0:13:59.000 into the original information that was sent out, like you know, 0:13:59.160 --> 0:14:01.400 your your pop radio station or whatever it is you're 0:14:01.440 --> 0:14:04.440 listening to. So really this is all about creating a 0:14:04.480 --> 0:14:08.040 reversible process that lets you translate information into some other 0:14:08.120 --> 0:14:12.000 format and then back again on the other side. Now, 0:14:12.000 --> 0:14:15.600 in the nineteen twenties, countries around the world slowly began 0:14:15.640 --> 0:14:19.960 building out telephone infrastructure, and there was now the opportunity 0:14:20.000 --> 0:14:24.400 to piggyback the infant teletype technology onto the phone infrastructure, 0:14:24.760 --> 0:14:27.680 except for the fact that phone companies were pretty dead 0:14:27.720 --> 0:14:32.200 set against it. Previously, these machines relied upon purpose built wires, 0:14:32.200 --> 0:14:34.760 but that limits their use right, and you can only 0:14:34.800 --> 0:14:38.800 send messages to whichever machines are hardwired to your machine. 0:14:39.360 --> 0:14:41.560 That doesn't really give you a whole lot of options. 0:14:41.800 --> 0:14:43.800 What if you wanted to send it to somebody else 0:14:43.840 --> 0:14:47.720 whose machine wasn't directly connected to yours, You'd be stuck. 0:14:48.120 --> 0:14:51.280 The telephone network would allow for a lot more connectivity 0:14:51.440 --> 0:14:54.119 and had a lot more infrastructure that was already established. 0:14:54.520 --> 0:14:58.680 One early approach was to use non switched telephone lines 0:14:58.720 --> 0:15:02.040 that were dedicated for the sorts of connections. These are 0:15:02.080 --> 0:15:06.320 called least lines because these lines were meant only for 0:15:06.400 --> 0:15:09.320 that kind of process and would not be used for 0:15:09.760 --> 0:15:13.680 phone to phone conversations. But that was really expensive. It 0:15:13.720 --> 0:15:16.560 would be way more cost effective and more useful to 0:15:16.600 --> 0:15:20.480 tap into the general telephone network, but there needed to 0:15:20.480 --> 0:15:22.400 be a couple of things for that to happen. One, 0:15:22.480 --> 0:15:25.600 you had to get it past the telephone company, which 0:15:26.080 --> 0:15:27.760 essentially in the United States was a T and T, 0:15:28.440 --> 0:15:29.960 and another one you had to find a way to 0:15:30.040 --> 0:15:33.720 modulate and demodulate the information so it could transmit over 0:15:33.760 --> 0:15:36.440 phone lines. There had to be a way to have 0:15:36.480 --> 0:15:39.280 two teletype machines establish a connection with one another and 0:15:39.360 --> 0:15:44.080 transmit and receive messages. At the time the phone infrastructure 0:15:44.080 --> 0:15:47.479 in the United States was capable of carrying sound frequencies 0:15:47.520 --> 0:15:51.880 between three hurts and three killer hurts, also known as 0:15:51.960 --> 0:15:56.240 the voice band. Now, that's a pretty narrow range of frequencies, 0:15:56.440 --> 0:15:59.120 and you could think of frequencies as being related to pitch. 0:15:59.280 --> 0:16:02.520 Lower frequent sees have a lower pitch. Higher frequencies have 0:16:02.520 --> 0:16:05.320 a higher pitch. But the range of human hearing goes 0:16:05.400 --> 0:16:08.880 from twenty hurts to twenty killer hurts, So the voice 0:16:08.880 --> 0:16:12.720 band represents only a small slice of the full range 0:16:12.760 --> 0:16:16.080 of human hearing. It also means that any solution that 0:16:16.080 --> 0:16:19.960 would convert information into audio to transmit over phone lines 0:16:20.320 --> 0:16:23.960 would have to work within those limitations. Oh and here's 0:16:23.960 --> 0:16:28.120 another fun fact. We call this system the plain old 0:16:28.160 --> 0:16:32.440 telephone service or POTS. Now it's been updated since then, 0:16:32.480 --> 0:16:34.520 but it took a long time for that to happen. 0:16:34.600 --> 0:16:37.760 The United States was relying on POTS until the late 0:16:37.840 --> 0:16:43.080 nineteen eighties, so it was a venerable technology by the 0:16:43.080 --> 0:16:46.120 time we finally got off of that. The teletype example 0:16:46.160 --> 0:16:49.680 is also important not just because of modulation demodulation, but 0:16:49.760 --> 0:16:52.680 because it would help inform future engineers as they began 0:16:52.720 --> 0:16:55.640 to think about creating methods for computer systems to communicate 0:16:55.680 --> 0:16:59.240 with one another. Keep in mind that the early computer 0:16:59.280 --> 0:17:03.680 systems were all independent. They were mainframe systems that performed 0:17:03.720 --> 0:17:07.320 like an isolated island. You needed physical access to the 0:17:07.359 --> 0:17:11.200 machines or two dumb terminals that were connected to those 0:17:11.240 --> 0:17:14.440 machines in order to make any use of them. Communication 0:17:14.520 --> 0:17:17.960 with other computers wasn't really a possibility for numerous reasons. 0:17:18.359 --> 0:17:22.560 One was that different computers effectively spoke different languages, so 0:17:22.600 --> 0:17:25.159 a program written for one type of computer would not 0:17:25.280 --> 0:17:28.520 work on another, just as a program written for a 0:17:28.680 --> 0:17:32.280 Mac computer won't run natively on a Windows based machine. 0:17:32.640 --> 0:17:35.359 But another reason was just that there was no interface 0:17:35.440 --> 0:17:39.240 through which computers could send or receive information with one another. 0:17:39.720 --> 0:17:42.400 When we come back, we'll talk about how that would 0:17:42.440 --> 0:17:45.320 begin to change, But first let's take a quick break. 0:17:52.840 --> 0:17:57.359 The development of modems, which weren't yet called modems, or 0:17:57.520 --> 0:18:01.520 were they actually for computers just yet, would coincide with 0:18:01.680 --> 0:18:05.800 other trends in technology. Now. As we've learned in many episodes, 0:18:06.240 --> 0:18:10.040 the development of tech rarely follows a straight timeline where 0:18:10.280 --> 0:18:13.840 event A leads to event B, which leads to events see. 0:18:14.160 --> 0:18:18.280 Usually you're actually talking about multiple timelines of multiple technologies 0:18:18.440 --> 0:18:22.000 and they ultimately either converge or at least cross paths, 0:18:22.359 --> 0:18:24.879 so it makes it complicated to tell the history. So 0:18:24.920 --> 0:18:28.600 while companies were working with teleprinters and teletype machines as 0:18:28.640 --> 0:18:32.960 well as fax machines, which are somewhat related, other engineers 0:18:33.000 --> 0:18:36.520 were working on building out computer systems. World War Two 0:18:36.560 --> 0:18:40.720 demonstrated how computer systems could be really important. In addition, 0:18:41.040 --> 0:18:44.480 the British developed radar technology, and it was obvious that 0:18:44.560 --> 0:18:49.119 being able to share radar information quickly across great distances 0:18:49.160 --> 0:18:53.040 like coast to coast in the United States, would be advantageous. 0:18:53.680 --> 0:18:56.159 It would be really useful if a radar station and 0:18:56.240 --> 0:18:59.639 say Hawaii, could send information back to the mainland in 0:18:59.720 --> 0:19:02.640 real time. That all that stuff could be monitored from 0:19:02.640 --> 0:19:07.240 a centralized location. Now here's a big challenge. Ever since 0:19:07.359 --> 0:19:11.000 Aniak hit the scene in the nineteen forties, computers have 0:19:11.080 --> 0:19:15.840 primarily fallen into the digital category, and digital information is 0:19:15.880 --> 0:19:19.600 different from the way information would transmit over phone lines. 0:19:20.160 --> 0:19:22.600 This leads us to talk about the difference between digital 0:19:23.000 --> 0:19:27.479 and analog. This is incredibly important for modems. So we 0:19:27.560 --> 0:19:30.439 can think of the analog world as being the world 0:19:30.440 --> 0:19:36.520 of infinite variability. Everything is infinitely variable. You can change 0:19:36.600 --> 0:19:40.600 something like the brightness of a light to an infinite 0:19:40.680 --> 0:19:44.280 degree of variability. You can always go a little less bright, 0:19:44.560 --> 0:19:48.680 or maybe a little more bright, or maybe half as 0:19:48.720 --> 0:19:51.080 bright as you just went the last time, or half 0:19:51.080 --> 0:19:54.360 as dim. You can keep dividing that up into finer 0:19:54.480 --> 0:19:59.159 and finer adjustments with no limitations. If we took the 0:19:59.240 --> 0:20:01.720 number ten and we cut it in half, we would 0:20:01.720 --> 0:20:03.679 have five. If we cut in half again, we have 0:20:03.760 --> 0:20:05.800 two point five, and we can keep cutting that in 0:20:05.840 --> 0:20:11.080 half forever. We can do that until we're just exhausted. 0:20:11.880 --> 0:20:15.439 So an analog recording is one that captures all the 0:20:15.520 --> 0:20:21.080 variations in a continuous signal. The digital world doesn't do that. 0:20:21.280 --> 0:20:25.879 A digital signal deals with discreete It deals with the finite. 0:20:26.440 --> 0:20:31.480 Computer's capabilities will determine how many values any given signal 0:20:31.680 --> 0:20:35.600 can possess. The more powerful the computer, the more values 0:20:35.640 --> 0:20:39.119 it can handle. With enough values, the signal can seem 0:20:39.119 --> 0:20:42.520 to be almost like an analog one, because you've got 0:20:42.640 --> 0:20:46.800 enough information to describe that signal that upon casual glance 0:20:46.880 --> 0:20:50.240 or even careful examination, it looks the same as an 0:20:50.240 --> 0:20:53.880 analog signal. But this requires an awful lot of information 0:20:54.240 --> 0:20:57.560 to accomplish. One way to think about this is to 0:20:57.600 --> 0:21:02.720 consider a film photograph versus a digital photograph. Film is analog. 0:21:03.160 --> 0:21:05.639 Digital photography is well, I mean it's in the name, right, 0:21:05.680 --> 0:21:10.240 it's digital. So we've all heard about megapixels, right. Megapixels 0:21:10.240 --> 0:21:13.560 referred to a camera's ability to compose an image with 0:21:13.600 --> 0:21:17.600 a certain number of points of color or light, and 0:21:17.800 --> 0:21:21.159 it's all about a camera's resolution. If we had a 0:21:21.200 --> 0:21:25.680 camera with a really low resolution, I mean, like ludicrously low, 0:21:25.760 --> 0:21:29.679 let's say eight by eight pixels, that would mean that 0:21:29.720 --> 0:21:33.399 the images would consist of eight blocks across an eight 0:21:33.440 --> 0:21:37.280 blocks tall, which means that the entire image would consist 0:21:37.280 --> 0:21:41.640 of just sixty four blocks. That would be a really 0:21:42.160 --> 0:21:44.399 blocky image. It would be hard to even know what 0:21:44.440 --> 0:21:48.440 you were looking at unless it was a super simple shape. 0:21:48.760 --> 0:21:52.280 If you increase the resolution, that means you're decreasing the 0:21:52.320 --> 0:21:55.760 size of those pixels. You're decreasing the size of those blocks, 0:21:56.160 --> 0:21:59.080 and you're cramming more of them into there so that 0:21:59.160 --> 0:22:03.480 you can represent the image with more blocks, smaller blocks, 0:22:03.520 --> 0:22:08.439 creating a smoother, less blocky image. If you keep doing that, 0:22:08.520 --> 0:22:11.320 eventually you get to a resolution that's high enough that 0:22:11.400 --> 0:22:14.160 our human eyes can't really pick up on the pixel 0:22:14.200 --> 0:22:16.320 blocks at all, and to us it just looks like 0:22:16.400 --> 0:22:19.360 a photograph on film. But again, it takes a lot 0:22:19.359 --> 0:22:23.160 of information to get to that point. Here's another challenge. 0:22:23.160 --> 0:22:26.120 How do you take the information a computer deals with, 0:22:26.280 --> 0:22:31.879 which is in digital binary format, these discrete packets of information, 0:22:32.160 --> 0:22:34.280 and then how do you send that out over a 0:22:34.280 --> 0:22:39.600 phone line which carries an analog signal over physical copper wire. 0:22:40.200 --> 0:22:43.399 You have to create a way to translate the information 0:22:43.640 --> 0:22:48.240 from digital into analog. You have to modulate the computer 0:22:48.400 --> 0:22:52.920 data so it can pass over onto an analog transmission system. 0:22:52.960 --> 0:22:55.720 Then on the other end, on the receiving end, you 0:22:55.800 --> 0:22:59.800 need a device that can accept this incoming analog transmission 0:23:00.160 --> 0:23:04.520 and demodulated translating it back into binary information for the 0:23:04.600 --> 0:23:08.240 receiving computer to process. This is true for tons of 0:23:08.240 --> 0:23:11.800 different input and output scenarios, not just computers communicating across 0:23:11.840 --> 0:23:15.280 phone lines. Anytime you're using a computer system or you know, 0:23:15.359 --> 0:23:18.879 a digital system to record or analyze stuff out in 0:23:18.920 --> 0:23:22.240 the world, you're typically relying on a digital process to 0:23:22.280 --> 0:23:25.400 measure an analog phenomenon. Same is true if you were using, 0:23:25.440 --> 0:23:28.520 say an analog joystick to play a video game. The 0:23:28.600 --> 0:23:32.680 analog controls, which might include a potentiometer that tells the 0:23:32.720 --> 0:23:36.200 device how far you're pushing the joystick in any given direction. 0:23:36.760 --> 0:23:39.359 That has to be converted into digital information for the 0:23:39.400 --> 0:23:41.840 computer to do anything useful with it. But let's get 0:23:42.000 --> 0:23:45.960 back to modems, particularly so the way a modem actually 0:23:46.040 --> 0:23:49.359 modulates data starts with a carrier wave, just like with 0:23:49.400 --> 0:23:52.840 the radio. Now, you could technically communicate in a very 0:23:52.880 --> 0:23:57.080 basic way just by either turning the wave on or off. Right, 0:23:57.400 --> 0:23:59.400 you could say, all right, well, when it's on, that's 0:23:59.400 --> 0:24:01.600 a one, and it's off it's a zero. But that's 0:24:01.640 --> 0:24:05.359 not necessarily the best option. Another way is to alter 0:24:05.600 --> 0:24:08.919 the carrier wave in some fashion. You can tweak the 0:24:08.960 --> 0:24:11.879 amplitude or the frequency, just like with a M or 0:24:12.000 --> 0:24:16.879 FM radio. So modems take a basic carrier wave function 0:24:17.280 --> 0:24:21.280 and then alter it in some predetermined way to communicate 0:24:21.280 --> 0:24:25.200 digital data. The receiver on the other side gets this 0:24:25.440 --> 0:24:28.840 carrier wave with alterations, and it understands the whole process 0:24:28.880 --> 0:24:31.359 that was made to encode that information, so it just 0:24:31.440 --> 0:24:35.240 reverses it. It decodes the information and gets those delicious 0:24:35.359 --> 0:24:38.520 zeros and ones at the heart of everything. The speed 0:24:38.560 --> 0:24:42.040 at which a modem can transmit information is measured in 0:24:42.160 --> 0:24:46.000 bits per second. A bit, remember, is one unit of 0:24:46.040 --> 0:24:48.840 binary information, so it's either a zero or a one, 0:24:49.200 --> 0:24:52.000 and the unit we used to measure modem speed is 0:24:52.040 --> 0:24:54.880 the baud be a u D named after a meal 0:24:55.000 --> 0:25:00.080 bodo bad refers to a symbol rate, and really that 0:25:00.160 --> 0:25:04.360 means how many times a transmission signal changes every second. 0:25:04.880 --> 0:25:08.640 More changes per second indicate more data carried by that 0:25:08.760 --> 0:25:12.360 signal per second. Now, typically we think of this as 0:25:12.400 --> 0:25:16.040 a faster transmission speed because it reduces the time it 0:25:16.119 --> 0:25:19.480 takes to transmit any given file, but really what it 0:25:19.520 --> 0:25:22.439