RSS 2.0 FeedMay 2009, Cover Stories, Technology
Less is More - How Digital Trumped Analog
When you tap your feet to a Beyonce song, it is easy to forget that her voice went through a long journey before reaching your ears. A single three minute track like “Single Ladies” gets translated into several millions of 0’s and 1’s sitting on a hard drive or an MP3 player, as opposed to a few grooves on a vinyl disk or a few inches of a cassette tape. Why then is the world moving away from analog to digital?
I was driving my daughter to her guitar lessons the other day when she tuned the car radio to Kiss FM. Pretty soon the car was filled with a voice urging all the single ladies to put a ring on the thing they loved. My daughter started bopping her head, and I asked her who the artist was. She pulled out her iPhone, pointed it to the speaker, and turned to me with a triumphant grin. She had this application called Shazam, which was able to recognize the song. This application told us that the artist was Beyonce, and that the song was “Single Ladies”.
When you tap your feet to Beyonce’s song, it is easy to forget that her voice went through a long journey before reaching your ears. The recording studio captured her song along with all kinds of sound effects on their hard drives, after going through an almost magical electronic transformation on their digital audio workstations. This digitally stored piece of music then got copied on to thousands (and in the case of super star artists like Beyonce, millions) of CDs, and – increasingly in recent years – to the repertoire of online music stores like iTunes, from where avid fans like my daughter could download directly to their MP3 players, computers or smart phones, one track at a time. Once downloaded, the bits and bytes making up that file are reassembled into the vibrating patterns that reach your ears through the iPod headphones, recreating the voice of Beyonce.
A single three minute track like “Single Ladies” gets translated into several millions of 0’s and 1’s sitting on a hard drive or an MP3 player, as opposed to a few grooves on a vinyl disk or a few inches of a cassette tape. That is, if you are only talking about the audio portion of the song. If you add the video, that number jumps to hundreds of millions of 0’s and 1’s as opposed to a few feet of a VHS tape. In general, representing voice, video or data in digital format requires a lot more storage space as well as network bandwidth than the equivalent analog format. Why then is the world moving away from analog to digital?
To answer this question one has to understand the history of how communication technology evolved over the last century. Whether it is sound, video, text or any other form of data, in order to send it from one location to another, you need to represent it in a form that can be transmitted, and if necessary stored for a period and converted back to the original format for the receiver. The technologies used for this purpose could be electrical, mechanical, optical, magnetic or some combination of all of these. The original telephony and telegraph technologies for instance used electrical circuits to represent the sound waves of the voices being transmitted from one speaker to another, with devices on either end that translated sound waves to electrical signals and vice versa. The representation of sound in the electrical signals was proportionate to the source sound wave properties. In other words, louder sounds were represented by higher currents. Similar technology was used in radio transmission, where sound waves were converted to electrical signals of proportionate or analog strength and then converted back to sound waves of equivalent strength on the receiving end. Early television technology was also similar, with the combination of the picture elements and the sound elements represented in multiple analogous electrical signals.
Around the mid twentieth century, computers brought another dimension to the communication paradigm – the sending and receiving of data, as opposed to voice or video over long distances. Early computers also followed the analog path, with data represented by proportionate values of electrical, mechanical, magnetic or optical signals. In fact, many early versions of computers were analog computers built for special purpose computing within a specific field of study. Typically built with combinations of potentiometers, operational amplifiers, and function generators, they used the basic properties of electric circuits like voltage, resistance, current etc. to represent the input data, and by manipulating these values through different circuits, basic arithmetic operations like addition, subtraction, multiplication and division were derived based on how these properties varied across the circuits.
Because analog computers represent data using equivalent electrical or mechanical properties, there is greater flexibility in managing complex calculations. For example, a single electrical signal might represent a single day’s production of a machine, or the gross domestic product of China, by merely turning up or down the voltage on the circuit. By contrast, in a digital computer the two numbers will have dramatically different representations, as everything gets reduced to a string of 0’s and 1’s.
As the use of computers grew, however, it became apparent that the precision required to maintain the integrity of data in an analog computer made it cost prohibitive for largescale use of such computers. Wear and tear of the mechanical components would over a period of time change the electro-mechanical properties of the components (like increase the resistance), which in turn would change the value of the data that one could represent within those components. Also, analog restricted the use of computers to computationally intensive “real time” applications such as numerical integration, with little or no need for long term storage of the results or the input data. By contrast, applications like payroll or general ledger involve relatively simple calculations, but require consistent and reliable results repeatedly over a long period, with both the inputs and the results needed to be available long term. Analog computers proved to be ill suited for such applications as they are not designed for precision and reliability over time, but rather for the speed and accuracy of processing complex calculations.
Digital computers on the other hand rely on representing data in Boolean terms – that is to say, everything is converted into a pattern of 0’s and 1’s, which can be represented as the presence or absence of a characteristic such as current, light, magnetic charge, a hole on a punched card, etc, as opposed to the value of that characteristic. Thus, whether a circuit at a given time carries one milli ampere of current or 10 milli amps is not as significant as whether it carries any current or not. This means that wear and tear of the components does not impact the reliability of data, as an increase of resistance or reduction of voltage would not change the fact that current is still flowing through the circuit. This engineering simplicity made it possible to design versatile computers that could perform a wide range of arithmetic and logical operations reliably over a long period of time.
Once everything was reduced to the presence or absence of a property as opposed to the value of that property itself, mass production of components that could store or process 0’s and 1’s became much easier. From vacuum tubes to integrated circuits to microprocessors, technology evolved rapidly to make computing faster, cheaper and more versatile. My daughter’s iPhone is perhaps a million times more powerful than the ENIAC, which occupied more than 2000 cubic feet of space, weighed about 30 tons, and consumed 150 KW of electrical power. The iPhone on the other hand weighs about 5 ounces, and holds 30 gigabytes worth of data.
Reducing everything to binary makes not just the storage of data simpler, but also its transmission from source to destination. Granted that having to convert a song or a picture or someone’s voice into 0’s and 1’s means that you need more bandwidth to convey the same information than if you used an analog means of transmitting that data, but that is more than compensated by the reduced overhead for managing the transmission. If voice, video and data are all converted to binary format, there is no need to have different networks and different equipment to manage the transmission and delivery of any form of content from one place to another.
A common pipe can carry all sorts of traffic, leaving the translation to equipment on either end to worry about. A cable company delivering TV signals to your house can also deliver telephony or high speed internet. Likewise the familiar twisted pair copper network connecting your telephone to the rest of the world can also carry video or high speed internet to your home. All of a sudden, your friendly neighborhood telephone company and your entertainment service provider can each do what only the other could do previously. As for the challenge of bandwidth, that’s a matter of technology – optical fiber today can carry a thousand times more traffic than twisted pair copper and today’s 4G technologies deliver live video to your smart phones, allowing you to watch American Idol while still stuck in traffic.
Digital transmission has another very significant advantage over analog transmission. In analog transmission, if anything happens to the signal on the way (such as a drop in voltage, for example) there is no way to recover from the error on the other end. Because digital transmission consists of 0’s and 1’s, it is possible to provide additional information along with the data (such as checksums) can help in validating the signal on the receiving side and recovering from any transmission errors. This means greater clarity in the voice, video or other content you receive.
The engineering simplicity of dealing with just two variables at a time trumps the mathematical limitation of dealing with just two variables at a time. Because each component of the network delivering the data from the source to the destination at any given point has only two degrees of freedom, it is easier to standardize the equipment, the interfaces, the signaling, and the error checking.
All in all, creation, storage and transmission of content (voice, video or data) is more efficient in the digital format than in an analog format. Less is indeed more sometimes.
Mr. Uday Bhaskar Nandivada is an enterprise architect and an authority on emerging trends in communications technology and a leading evangelist of TeleManagement Forum’s New Generation Operational Support Systems (NGOSS). He has over twenty three years of experience in architecting and implementing complex enterprise level solutions for a wide range of industries and business models. For the last nine years he has been working exclusively with telecommunications, media and entertainment service providers, providing strategic advisory services to Tier one carriers around the globe including AT&T, Sprint, Comcast, Directv, Telstra, Sprint, Telkom South Africa, FarEastern Telecom, etc.
He is currently the Global Technology Solutions Lead with Amdocs Consulting, in which capacity he has led the development and successful launch of five holistic world-class technology solutions to address critical business needs of the world’s leading telecommunications and media providers. Prior to joining Amdocs, Mr. Nandivada was a part of the Cable & Media Leadership team for Capgemini Inc., where he successfully led the implementation of several complex OSS/BSS Integration projects for more than ten major service providers. Earlier, he was a solution architect for AT&T’s Texas Local Factory. He may be reached at uday_n@sbcglobal.net.