Digital-Audio Myths
Page 1, 2, 3, 4, 5, 6

However, recording from one digital device to another is different; as long as the data stays in the digital domain, jitter is not recorded. The only thing actually captured is a sequence of amplitude values; digital media simply have no provisions for storing information about individual sample timing. The timing is based implicitly on the sampling rate and is freshly re-created by the digital-to-analog converter’s clock every time the audio is played back.

Even digital-audio tape systems don’t play audio directly from the tape. Instead, they pass the data through a RAM buffer from which a clock pulls individual samples and sends them to the outputs. As a result, variations in tape speed or data spacing aren’t reflected in the output data.

Although jitter causes distortion on playback—and can certainly generate unalterable distortion during the A/D process when recording from an analog source—it is not recorded when making a digital dub or when recording between digital devices. On the other hand, if the jitter of one device involved in a digital dub is bad, it can cause problems of a different sort, as I will discuss in the next myth.

Myth No. 6: Digital dubs are perfect copies. If you copy audio between DAT decks, CDs, or modular digital multitrack (MDM) tape machines, you might expect the copy to be perfect. After all, digital audio is just ones and zeros, right? As I described previously, copies of audio files on computers are indeed perfect. However, when you use tape and CD-audio media, those zeros and ones are being assisted, and sometimes created, by error correction.

FIG. 5: For simple error correction, the linear interpolation draws a straight line between the points on either side of the error (top). Higher-order error correction takes into account the curvature around the error and uses multiple points on each of the sides to reconstruct a more complex, and hopefully more accurate, shape (bottom).

Here’s the problem: If an error occurs when a computer reads a file from a hard drive, the computer can go back and try again. A backup tape drive can do the same thing, rewinding the tape as necessary. But with an audio DAT and other digital media, the tape or disc must always keep moving; otherwise, you hear a pause in playback. When errors occur, and they always do, going back and trying again simply isn’t an option. Instead, the DAT and CD formats include several methods for real-time error correction.

The first and most effective error correction is recorded onto the tape in the form of Reed-Solomon error-correction codes. These codes take up more than 25 percent of the data on a DAT tape or CD, and they allow most errors to be completely corrected, yielding data that is byte-for-byte perfect.

Occasionally, even this sophisticated error-correction mechanism can’t recover the data, resulting in one or more completely blank samples. In that case, the second level of error correction comes into play. This technique, called interpolation, considers the data before and after the blank sample or samples and then makes a guess as to what value might have been in that blank space.

Say you have the following sequence of samples: 2, 6, <error>, 14, 15. Simple interpolation might draw a straight line between the samples just before and just after the error so that the sequence becomes 2, 6, (10), 14, 15. More sophisticated interpolation constructs a curve based on two or more samples on either side of the error. This procedure recognizes that the two samples after the error have closer values than the two before the error, and it tries to reflect that curvature in the corrected data. Its reconstruction of the data might look more like 2, 6, (12), 14, 15 (see Fig. 5). Generally, the more points you look at during interpolation, the more accurate of a guess you can make as to the error’s original value, though it’s still just a guess.

Reprinted with permission from

[an error occurred while processing this directive]