Dynamic range
Overview
The term "dynamic range" is used to define the difference between the highest and lowest level an audio signal can have in a circuit or device. The lowest level is typically limited by noise in the system. At some point the signal level will be low enough for the signal to be obscured by the noise. At the other end of the scale, the loudest signal level is typically defined by a certain level of distortion. In analog audio exactly what level of distortion is used varies with the application; and in the case of digital audio, corresponds to full scale or "digital clipping" level where all bits in the digital word which represent audio are "1."
Basics
Dynamic range in audio is quite often specified as the "signal-to-noise ratio" (SNR). It is typically specified in decibels (dB) which is a relative measurement- the ratio of the power of the signal to the power of the noise. The "signal" means the highest level signal the circuitry can handle without distortion and the "noise" is typically present in the circuitry whether or not the signal is present. The highest level the circuitry can handle is usually the point where a specific level of distortion is used to signify the onset of distortion. This means that the signal level will only need to be increase slightly for the distortion level to increase significantly. If the onset of distortion is more gradual, this level represents when the distortion becomes audible.
Exactly what is considered to be audible distortion can vary with the type of device being measured. For example; in loudspeaker testing levels as high as 3-5 % distortion are not unusual compared to one-half a percent or less for high quality audio amplifiers. As a result; the distortion level used to determine the peak level as well as the type of distortion is usually also specified as part of the SNR or dynamic range specification.
The other factor is the rate of the on-set of distortion. In digital audio or amplifier circuits with op-amps; the highest signal level is quite often just below "clipping level" and the signal below that is very low in distortion. Due to the way in which none of the original information is retained above the clipping level, the onset of distortion is quite rapid beyond this level. This means a very low level of distortion can be used to signify the peak level because even a small increase in distortion will indicate being very close to the gross distortion level. In the case of a speaker or analog tape recording; the onset of distortion is much more gradual. This makes defining "the point" at which the distortion increases to an unacceptable level more arbitrary and subjective.