Stereophile's Test CD 2 Tracks 27-29

Track Information, Tracks 27-29, CD Player & Tape Recorder test tones

[27] Spot frequency tones at -20dBFS (L+R) (DDD) 6:15
10Hz (Index 1), 12.5Hz (Index 2), 16Hz (Index 3), 20Hz (Index 4), 25Hz (Index 5), 31.5Hz (Index 6), 40Hz (Index 7), 50Hz (Index 8), 63Hz (Index 9), 80Hz (Index 10), 100Hz (Index 11), 125Hz (Index 12), 160Hz (Index 13), 200Hz (Index 14), 250Hz (Index 15), 315Hz (Index 16), 400Hz (Index 17), 500Hz (Index 18), 630Hz (Index 19), 800Hz (Index 20), 1kHz (Index 21), 1.25kHz (Index 22), 1.6kHz (Index 23), 2kHz (Index 24), 2.5kHz (Index 25), 3.15kHz (Index 26), 4kHz (Index 27), 5kHz (Index 28), 6.3kHz (Index 29), 8kHz (Index 30), 10kHz (Index 31), 12.5kHz (Index 32), 14kHz (Index 33), 15kHz (Index 34), 16kHz (Index 35), 18kHz (Index 36), 20kHz (Index 37)

[28] De-emphasis test tones at -20dBFS, (L+R) (DDD) 0:52
100Hz, 1kHz, 4kHz, 10kHz, 16kHz; each tone lasts 10s.

[29] Noise Modulation Test Tones (L+R) (DDD) 1:02
41Hz at -50dBFS (Index 1), -60dBFS (Index 2), -70dBFS (Index 3), -80dBFS (Index 4), -90dBFS (Index 5), -100dBFS (Index 6) with dither

Track 27 consists of 37 sinewave tones, each lasting 10s, covering frequencies from below audibility to the top of the audio band, all recorded at the same level. These will enable anyone with an accurate voltmeter—we can confidently recommend the handheld Fluke 87 series as being flat in the audio band—to assess the frequency response and crosstalk of their amp, preamp, and tape decks. By contrast, the tones on track 28 have been recorded with a treble boost. Termed "pre-emphasis," this boost is canceled by a complementary cut in the treble when the CD is played back, due to the CD player detecting what is termed the "pre-emphasis flag," if set in the datastream. The result is that HF distortion and noise are reduced in level (though there is an unfortunate tradeoff in high-frequency headroom, which is why most discs are not pre-emphasized). If your CD player's de-emphasis circuitry is operating correctly, these tones should all measure as being identical in level. If not, then any pre-emphasized CDs will not be reproduced with the correct tonal balance. (Track 9 on this CD, for example—Peter Mitchell's organ recording—is recorded with pre-emphasis.)

Track 29 will enable those with access to a spectrum analyzer to examine their CD player's performance with a new test proposed by Audio Precision's Dr. Richard Cabot (footnote 10) at the February 1991 Audio Engineering Society Convention in Paris. The test, based on psychoacoustic principles, attempts to predict the audible performance of a player's D/A conversion by measuring its noise modulation (footnote 11).

The technique is straightforward: the D/A section is driven with the code representing a low-frequency sinewave, the player's output is high-pass filtered to remove the test signal, and a 1/3-octave spectral analysis of the analog output is performed. The result is plotted as noise level vs frequency. The measurement is repeated at different input-signal levels, with each curve overlaid on the previous curves for easy comparison (footnote 12).

The test-signal frequency is 41Hz, chosen because it is not an integer sub-multiple of the sampling frequency. The test signal will thus exercise the greatest number of steps in the DAC. Six signal levels are used, from -50dBFS (FS = Full Scale) to -100dBFS, in 10dB steps (footnote 13).

Basically, the technique measures noise-floor shifts (a result of quantization distortion) as a function of signal level. There is a direct correlation between low-level linearity and performance in this test. In addition to how much the noise floor is modulated by signal level, the measurement reveals shifts in the noise floor's spectral balance with changes in signal level. Ideally, the noise-floor spectrum should remain constant with level, producing curves that exactly overlay each other. Psychoacoustic research by Louis Fielder at Dolby Labs indicates that noise-floor shifts of 2dB are audible. Further, Dr. Cabot's paper asserts that the ear is very sensitive to shifts in the noise floor's spectral balance; changes on the order of 1dB are reportedly audible.

Footnote 10: See Robert Harley's interview with Dr. Cabot in Stereophile, Vol.14 No.1, January 1991.

Footnote 11: "Noise Modulation in Digital Audio Devices" is available from the Audio Engineering Society, 60 E. 42nd Street, Room 2520, New York, NY 10165-0075. Tel: (212) 661-8528.

Footnote 12: See Stereophile, February 1992, Vol.15 No.2, p.143.

Footnote 13: To record digital data representing signals at or below the theoretical resolution limit of the 16-bit CD system, the appropriate amount of shaped dither noise was used.