Linn Majik CD player Measurements
The Linn Majik's maximum output level was 2.01V RMS and the player preserved absolute polarity; ie, was non-inverting. The player's error correction was superb: the Majik produced audible glitches in its output only when the gaps in the data spiral on the Pierre Verany test CD reached an enormous 2.5mm in length. The output impedance was to specification at 298 ohms at low and midrange frequencies, dropping very slightly but inconsequentially to 291 ohms at 20kHz.
The Majik's frequency response was flat, with just a very slight, –0.2dB droop at 20kHz (fig.1, top pair of traces). However, when it played back pre-emphasized data (fig.1, bottom traces), there was a slight positive error in the treble, which will be just audible with those few CDs that were mastered with pre-emphasis. Channel separation (not shown) was superb, at better than 115dB below 1kHz, and still an excellent 92dB at 20kHz.
Fig.1 Linn Majik, frequency response at –12dBFS into (from top to bottom at 100Hz): 100k and 600 ohms without de-emphasis, 100k ohms with de-emphasis (right channel dashed, 0.5dB/vertical div.).
Spectral analysis of the Majik's output while it decoded data representing a dithered 1kHz tone at –90dBFS (fig.2), and the plot of the player's DAC linearity error (fig.3), really show only the properties of the dither noise recorded on the test CD. Any failings the Linn might have are less than the intrinsic performance limitations of the 16-bit CD medium. Plotting the spectrum of the Majik's output signal while it decoded data representing a –1LSB DC offset revealed (in fig.4) that while there was a rise in ultrasonic noise due to the D/A converter's noiseshaping, this was relatively mild compared with some other modern DAC chips. And the player's low analog noise floor permitted the accurate reproduction of an undithered 1kHz sinewave at exactly –90.31dBFS (fig.5).
Fig.2 Linn Majik, 1/3-octave spectrum with noise and spuriae of dithered 1kHz tone at –90dBFS, 16-bit data (right channel dashed).
Fig.3 Linn Majik, left-channel departure from linearity, 16-bit data (2dB/vertical div.).
Fig.4 Linn Majik, 1/3-octave spectrum with noise and spuriae of –1LSB, 16-bit data (right channel dashed).
Fig.5 Linn Majik, waveform of undithered 1kHz sinewave at –90.31dBFS, 16-bit data.
Harmonic distortion was extremely low: a full-scale 1kHz tone was reproduced with just 0.0009% THD in the left channel and 0.0011% in the right (true sum of the harmonics). The second and third harmonics are highest in level (fig.6), but at just –103dB, they will be inaudible. This graph was taken into a high impedance; dropping the load to 600 ohms increased the third harmonic by 10dB, to –93dB, but this won't bother anyone either. Intermodulation distortion was also vanishingly low (fig.7), even into 600 ohms (not shown).
Fig.6 Linn Majik, spectrum of 1kHz sinewave at 0dBFS into 8k ohms (linear frequency scale).
Fig.7 Linn Majik, HF intermodulation spectrum, 19+20kHz at 0dBFS peak into 8k ohms (linear frequency scale).
When it came to the rejection of word-clock jitter, the Majik's performance was as superb as it had been everywhere else. Tested with the Miller Audio Research Analyzer, the Majik produced just 193 picoseconds peak–peak, with almost all the data-related sidebands in fig.8 (red numeric markers) at nearly the residual level of the test signal. The ubiquitous low-frequency sidebands at ±15.6Hz and ±31.2Hz can still be seen (purple "1" and "2," respectively), but the main jitter sidebands lie at ±567Hz (purple "4"). These are of unknown origin.
Fig.8 Linn Majik, high-resolution jitter spectrum of analog output signal (11.025kHz at –6dBFS, sampled at 44.1kHz with LSB toggled at 229Hz), 16-bit CD data. Center frequency of trace, 11.025kHz; frequency range, ±3.5kHz.
Other than that modest de-emphasis error, the Linn Majik's measured performance constitutes the state of the art for a 16-bit device.—John Atkinson