Sound Track: Speaking of Specs

by Jerry Whaley
July 1996

Specifications capture the essence of equipment performance. Understanding them will make you a savvy shopper.

When you're shopping for audio or video equipment, you should base your decision on three basic criteria: price, features and performance. Price and features are relatively easy to compare. Once you decide on a price range and select the features you need, you're "in the ballpark." The next step is to compare specifications. They are the true indicators of the equipment's performance, and they help you get the most bang for your buck.

But be careful. Comparing specifications can be a simple and straightforward process, or it can be a misleading mess. The trick is to make sure you're comparing apples with apples, not apples with oranges.

All specifications are measured relative to certain standards. The problem is that not all manufacturers use the same standards. Some manufacturers select a standard, or even create one, that makes their product look better than the competition's. Most independent testing laboratories use identical test methods and reference points, and are better information sources.

To determine which specifications are useful and which aren't, let's take a closer look at how we measure audio performance.

Units of Measurement
Sound loudness, or level, can range from nearly inaudible to ear-shatteringly loud. Measuring such a wide range on a linear scale would require very large numbers. So to make the numbers more manageable, engineers measure sound on a logarithmic (log) scale. The fundamental unit for measuring sound level is the decibel (dB).

Engineers usually measure sound levels relative to a given sound level. The decibel (dB) is actually a ratio, a measure of the difference between two levels. A difference of one "Bel" (named after Alexander Graham Bell), is equivalent to twice (or half) the loudness of a given sound. But the Bel is a rather large unit, so engineers use the decibel, which is one tenth of a Bel. Engineers also express audio levels as a ratio above or below an established reference point in terms of loudness, power or voltage.

While you don't need to become an engineer to compare audio specifications, there are three important two-to-one ratios you need to understand and remember:

• A doubling or halving of loudness represents a 10dB change,

• A doubling or halving of power, or wattage represents a 3dB change, and

• A doubling or halving of voltage represents a 6dB change.

Remembering these ratios is fundamental to understanding audio specifications.

You will often see measurements of electrical signal levels within a device or chain of devices. Signal levels are measured in volts. The volt is an absolute unit. But since we are interested in the difference between two or more signal levels, we measure and express these signal levels as ratios, in dB.

Frequency Response
The fundamental unit of audio frequency is cycles per second, or Hertz (Hz). One kilohertz (kHz) is a thousand cycles per second. In measuring audio, we are interested in frequencies from 20Hz to 20kHz because that's the maximum range of frequencies humans can hear. Realistically, we only need to consider audio frequencies from about 80Hz to about 16kHz. Any audio signal below or above this range is likely to be unwanted hum, rumble or hiss. The low and high ends of the audio spectrum require the most attention to render them in a pleasing manner. Low frequencies give your audio track "punch" and impact. Highs give your audio a nice, clear, "open" sound.

When comparing frequency response specs, be aware that the loudness of the device may vary at different frequencies. For example, a high-quality audio device typically has a frequency response specification of 80Hz-16kHz +/-3dB. This means that the device's loudness at one frequency in the given range and its loudness at another frequency in that range cannot differ by more than 6dB. A mid-quality audio device, in contrast, might have a frequency response spec of 80Hz-16kHz +/-10dB. The frequency range for both devices is the same, but the loudness variations of the second device are much larger within the specified frequency range.

The dB variation associated with the frequency response range is important to note. With the possible exception of microphones, the greater the dB variation within the frequency range, the poorer the audio performance. Beware of the solitary "20Hz-20kHz" frequency response specification. It is meaningless without the +/- dB specification.

Most purely electronic devices (such as preamplifiers and amplifiers) have excellent frequency response specs. Analog tape recorders are another matter. When comparing audio tape recorders, video camcorders and VCRs, be sure to look closely at the audio frequency response specs. Don't expect a cassette tape recorder to have the same specs as a reel-to-reel recorder or a digital recorder. Likewise, a VHS camcorder or VCR may have a high-frequency response of 10kHz or less, while 8mm and VHS hi-fi devices may have a high-frequency response approaching 20kHz.

Microphones are a special case because they have level variations within their response curves that provide them with special characteristics. A "presence peak" (in the 2kHz to 6kHz area) can improve the clarity of the human voice. Still, the overall frequency response needs to be within the 80Hz-16kHz range to produce good, professional sound.

Sensitivity, Efficiency and Gain
Sensitivity is the microphone specification that tells you how much output to expect with a given level of input. Generally, condenser mikes and professional dynamic mikes are very sensitive and have a "hot" output (plenty of signal). Standard dynamic mikes don't have a lot of output, and require more amplification (signal gain) from the device you plug it into.

Microphones have a relatively low signal output, so their sensitivity is usually measured against a 0dB reference level. Thus, a microphone with a sensitivity specification of -40dB has 20dB more output than a microphone with a -60db specification. The closer the sensitivity spec is to 0dB, the better the sensitivity. There's really no need to get more technical than this with microphone sensitivity specs. If you do want to compare numbers, be certain the product specs you're comparing use the same reference standards and units.

The term sensitivity also describes the amount of output you obtain from a loudspeaker when applying a given amount of power. Fortunately, there is an accepted standard for measuring speaker sensitivity: 1 watt/1 meter. This means that the testing lab applies one watt of power to the speaker and measures the level of sound output with a microphone positioned one meter away. If you don't see the sensitivity rating given at 1 watt/1 meter, the manufacturer probably has something to hide, like an inefficient speaker.

Inefficient speakers can have ratings below 90dB, while efficient speakers rate at around 100dB or more. If speaker A has a sensitivity rating that's 3dB lower than speaker B, speaker A requires twice the amplifier power to give the same loudness as speaker B. If speaker A's sensitivity rating is 6dB lower, it requires four times the power as B to produce the same loudness. That's quite a difference!

Gain generally refers to the ability of a device to amplify a signal. A mixer's microphone inputs must have a great deal of "gain" to bring the low-level microphone signal up to the operating level of the other circuits in the mixer. This is usually not a point of concern unless you really "cheap-out" on your mixer.

Signal-to-noise, Dynamic Range and Crosstalk
Signal is something you want, and noise is something you don't want. Unfortunately, you have to deal with noise in any electronic system. Audio devices can generate noise (hum, hiss or rumble), or pick it up from other electronic equipment.

There is one spec that tells you how much noise a device generates, and how low the noise is relative to the desired signal. We call this specification the "signal-to-noise ratio" (S/N). It is the ratio of the desired signal level to the internal electronic noise that the product generates. A good S/N (70dB or greater) means that noise will not be a problem for you. A poor S/N (50dB or less) means the device will generate noise that is clearly audible and highly objectionable. To compensate for a poor S/N, you'll have to record at the highest signal level possible without distorting the signal.

Headroom, sometimes given as a specification, is the amount of level above nominal (normal) that an audio device can reproduce. If you add the S/N to the headroom, you have the dynamic range of your device. Dynamic range is the maximum range of audio levels, from noise floor to maximum output, that an audio device can reproduce. So a device with a S/N of 80dB and 10dB of headroom has a dynamic range of 90dB.

Most microphones have tremendous dynamic range. You should look at the specs of any mike you might be considering, and compare them to others. Most electronic devices have fairly good S/N specs, though they do vary from unit to unit and are worth reviewing.

Just as with frequency response, devices using magnetic tape represent the most challenged group with regard to S/N and headroom. The S/N of various videotape formats range from 40dB for VHS to 90dB for Hi8 PCM and VHS hi-fi. It's definitely worthwhile to compare these specs between machines. Generally, the more tape area the audio recording head covers in a given amount of time, the better the S/N and overall dynamic range.

Crosstalk is a measure of how much one audio signal "bleeds" over into another. Audio track crosstalk increases distortion and affects stereo separation. If a time code signal bleeds onto another audio track, it basically becomes a noise source. The lower the crosstalk, the better. You may also see this spec listed as "separation," in which a higher number is better.

Power
The watt is the fundamental unit of power. Wattage ratings usually indicate how much power an amplifier can deliver to a speaker. A power handling rating indicates how many watts a speaker can handle for a certain amount of time without damage.

When looking at power amplifier ratings, try to determine if you are looking at a broadband (20Hz-20kHz) rating or a single frequency (1kHz) rating. Some manufacturers give both. The broadband rating is more meaningful, but the 1kHz rating yields a larger number. So which do you think most amplifier producers want to toss out at you? That's right--the high number. If a manufacturer doesn't specify a broadband or single-frequency reference, and instead quotes a "music power" rating or something similar, beware. They're not telling you what the amplifier's power really is.

Distortion
All audio devices alter the audio signal in small, unwanted ways. We call this alteration "distortion," and measure it as total harmonic distortion (THD) or as intermodulation distortion (IMD). You want the THD and IMD numbers to be as low as possible. IMD is more objectionable to the ear than THD, at a given percentage level.

Each component adds its own distortion to the signal. With that in mind, you can see why it's important to keep the distortion in each component as low as possible. You can expect to see .01%THD specs for most signal processing equipment such as audio mixers and equalizers. Amplifiers routinely specify their power ratings at from .1% to 1% distortion, though they generally operate at lower distortion levels.

Specs and the Audio Chain
It is sometimes hard to see why specifications for individual audio devices receive so much emphasis. But it is easier to understand if you think of an audio system as a chain of audio devices. If you were hanging, suspended by a metal chain, you would be depending on the weakest link in that chain to support your weight. If the weak link can't hold all your weight, down you go.

An audio system is similar. Just as a chain is only as strong as its weakest link, your audio system's performance is only as good as your worst audio component. If your mixer has a S/N of 80dB, and the mixer is part of the signal chain, the system's S/N cannot be greater than 80dB.

The Informed Buyer
With your new knowledge of audio specifications, you can now go out and be an informed buyer when you need a new piece of audio equipment. Keep in mind, however, that specs are not everything. A good recommendation from a friend, or from a competent salesperson, can sometimes be more helpful than depending solely on specs. This is especially true if you are unfamiliar with the product group you are considering.

In any case, it pays to look before you leap. Audio specifications provide a convenient, if not totally dependable, way of comparing audio devices before you lay down your hard-earned cash.

• Digg This!
• del.icio.us
• Technorati
• StumbleUpon
• Reddit