Interfacing Instruments with Computers

Here's how to transfer the signals from your analog world into the computer's digital world

By Craig Anderton

A guitar, mic, or electric piano doesn’t just plug into a computer the way it would plug into an amplifier. This is because instruments are analog devices that generate a continuously varying audio signal, while a computer is a digital device that processes data. So, the audio must be converted into data before the computer can use it, and then must be converted back into audio so we can hear it – we can’t “hear” data any more than a computer can “hear” a microphone.

THE COMPUTERIZED SIGNAL CHAIN 

Here’s a “map” (Fig. 1) of the computerized signal chain:Fig. 1: Block diagram of the computerized signal chain. Signal source (instrument) > Preamp (usually necessary) > Analog to Digital Converter (translates analog in to digital out) > Computer > Digital to Analog converter (translates digital in to analog out) > Monitoring system

Let’s explain each block in general terms. 

• Signal source (instrument): This is the analog output from your instrument or other electronic signal generator (CD player, Minidisc, etc.).
• Preamp: Many instruments, such as guitar, electric bass, microphones, old electric pianos, and the like put out low-level signals that need to be amplified in order to provide enough level for subsequent stages.
• Analog-to-Digital converter: This is built within a device called an audio interface. The audio interface accepts an analog input, and converts it to a data format the computer can understand. Its input will be audio connectors, and its output will be a digital signal cable that hooks into the computer. Think of the interface as a “bridge” between the analog and digital worlds.
• Computer: This runs the software that processes the data. That software can be anything from a sequencer that acts like the computerized equivalent of a tape recorder, to guitar amp simulations that turn your computer into a virtual amp, cabinet, and pedalboard. And of course, those are just two examples! There are complete virtual studios (like Propellerheads’ Reason), sophisticated tuners like Peterson’s Strobosoft, entry-level programs for beginnings, DJ tools, and much more.
• Digital-to-Analog converter: This converts the processed data back to analog so we can feed it into a monitoring system and hear it. It’s often built into the same audio interface that sends the signal into the computer. Note: Sometimes the audio interface is referred to as “I/O” because it provides the input to the computer and takes the computer’s output.
• Monitoring system: This could be a headphone jack in your audio interface, an amplifier with a set of speakers, powered speakers, a PA system, or anything that lets us hear the results of what the computer does.

THE CONVERSION PROCESS 

Now we need to make a brief journey into the world of digital audio theory. Although this isn’t the world’s most entertaining subject, understanding a bit about how the process works will allow you to make intelligent decisions when setting up your computer-based studio, as well as make better quality recordings…and that’s exactly what we’re trying to do. 

Analog-to-digital conversion is crucial, as the conversion has to be extremely accurate for the highest possible signal quality. (Digital-to-analog conversion is equally important, but technically easier to do.) As a result, your ultimate sound quality depends to a huge extent on the conversion process. 

Conversion takes advantage of the fact that all sound, from a dog bark to a symphony orchestra, is simply a change in air pressure that varies over time. Expressed in electrical terms, this change in air pressure is analogous to a voltage that changes over time. For example, when air pressure hits a dynamic microphone, a diaphragm moves back and forth, and generates a voltage because the diaphragm cuts across a magnetic field. A guitar pickup works similarly, but instead of responding to the vibrations of air, it responds to the vibrations of your guitar strings. They also cut across a magnetic field and generate a voltage. 

However, a computer cannot understand a changing voltage unless it’s presented as a series of numbers. So, a converter works by taking a series of “snapshots” of an incoming analog voltage, measuring the voltage of each snapshot, then converting that number into digital data (binary numbers) the computer can understand. This process is called sampling. 

To accurately convey the input signal level, a converter takes thousands of snapshots every second. This is called the sampling rate. A CD uses audio that was sampled at 44,100 times a second (44,100 Hertz, or 44.1 kHz), and that’s the most common sample rate that most digital studios use. However, there are other popular sample rates, including 48,000 kHz, 88,200 kHz, 96,000 kHz, and even higher (Fig. 2).

Fig. 2: The upper window shows a waveform sampled at 44.1 kHz, while the lower one is sampled at 96 kHz. Clearly, the lower window has greater resolution; but practically speaking, this doesn’t make a huge difference in the sound quality if the rest of the system uses quality components and engineering practices.

Lower sample rates are used too, but mostly for “lo-fi” applications like answering machines, toys, and the like. 44,100 kHz is considered the minimum sampling rate for professional audio applications. 

There is currently a big debate about whether rates higher than 44,100 kHz yield a significant sonic improvement. The only reason for the debate is because you don’t get something for nothing, otherwise everyone would use the highest sample rate possible. One tradeoff is that all the numbers that make up digital audio have to be stored somewhere, and obviously, the more samples per second, the more data there is to be stored. Another tradeoff is that if you double the sample rate, the computer has to work harder because it has to deal with twice as much data. So, you may not be able to record as many tracks, or insert as many plug-ins (virtual signal processors) as you might want. 

Higher sample rates have some other advantages aside from sound quality, but these are mostly technical in nature and probably not worth the time to examine them. So here’s my summary on sample rates, but note that some audio professionals (and certainly, some marketing departments of companies trying to sell devices with higher sample rates!) would disagree: 

• 44.100 kHz. This “lowest common denominator’ sample rate works just fine (CDs sound okay, right?) and is the default standard for most people.

• 48 kHz. Although not used much for purely audio products, a lot of video projects run audio at 48 kHz. If you do audio for video, you might be requested to provide music in this format. Otherwise, the slightly higher sampling rate doesn’t offer much of a sound quality improvement (if any) compared to 44.1 kHz.

• 88.2 kHz. Many people claim this sounds better than 44.1kHz, while others don’t hear much of a difference. If it sounds better to your ears and your gear can handle it, 88.2kHz is a good choice for audio work as it’s easy to convert back down to a CD’s 44.1kHz rate.

• 96 kHz. This is the most common “high” sample rate, and is sometimes used in DVDs and other high-end audio recording processes. Like 88.2kHz, if you can hear a difference and your gear is up to the task, it’s probably worth using…although as with 48/44.1kHz, 96kHz provides no significant advantage over 88.2kHz.

• 176.4 and 192 kHz. I don’t think these ultra-high sample rates are worth the effort or extra storage although of course, some people don't agree.

Another aspect of the conversion process is called bit resolution. This simply states the accuracy with which the converter can measure the input signal. A good analogy is a mosaic: The more tiles in a mosaic of a fixed size, the more defined the picture will be. 

CDs are 16 bit, which means that audio voltages are defined with an accuracy of 1 part in 65,536. However, most audio programs let you record with 24-bit resolution, which provides an accuracy of about 1 part in 16,000,000. Obviously, this is a lot more precise – assuming you’re also using converters with 24 bits of resolution. As with higher sample rates, higher bit resolution requires more storage: a 24-bit file is 50\% larger than a 16-bit file, assuming they use the same sample rates. Unlike the controversy involving higher sample rates, though, few dispute that 24-bit recording sounds better than 16-bit recording. If you can’t decide whether to use higher sample rates or higher bit resolutions, the latter will make a greater sonic difference. 

Most people record at 44.1kHz with 24-bit resolution, although some use higher sample rates. Just remember that ultimately, the music you play will always be the most important element of all.

Craig Anderton is Editor Emeritus of Harmony Central. He has played on, mixed, or produced over 20 major label releases (as well as mastered over a hundred tracks for various musicians), and written over a thousand articles for magazines like Guitar Player, Keyboard, Sound on Sound (UK), and Sound + Recording (Germany). He has also lectured on technology and the arts in 38 states, 10 countries, and three languages.