Columns Retired Columns & Blogs |
Excellent review, I have recently plumped for the revised Uniti2 which has a little more power than the original, very pleased with the musical performance despite it being a one box system.
I examined the Naim Uniti's measured behavior using mainly Stereophile's loan sample of the top-of-the-line Audio Precision SYS2722 system (see the January 2008 "As We See It" and www.ap.com); for some tests, I also used my vintage Audio Precision System One Dual Domain. To test the digital playback performance, I played WAV files on an iPod Touch, the same files loaded onto a USB flash drive or burned to CD-R, or fed to one of the Uniti's optical S/PDIF inputs from the RME soundcard in one of my test-lab PCs. All tests of the Uniti's digital section were taken from its Line Out jacks, which come before the volume control. I didn't have to worry, therefore, about blowing up the amplifier's output stage with high-level signals, as I could keep the volume control set to zero.
Fed a 1kHz signal at 0dBFS from CD, the maximum level from the Line Out jacks was 1.965V. However, I could raise only 1.795V from the front-panel headphone jack without clipping. With a full-scale digital signal, that was equivalent to a volume-control setting of "82" rather than the maximum setting of "100." While the headphone output preserved absolute polarity, the Line output inverted signal polarity. The source impedance was close to the specification, at a moderately high 555 ohms across the audioband from the Line Out jack, but a very low 1 ohm at high and middle frequencies from the headphone jack, and 4 ohms at 20Hzstill very low. The Uniti's headphone output thus uses a true buffer stage rather than just tapping the amplifier's main output stage via a series resistor. Error correction for the Uniti's CD transport section was good rather than great, the player not muting its output until the gaps in the data spiral of the Pierre Verany test disc reached 0.7mm in length. The Naim's digital input locked to data with sample rates ranging from 32 to 96kHz. The D/A frequency response (fig.1) basically followed the same overall shape regardless of sample rate: flat within the audioband, a gentle rolloff above 20kHz, and then a sharp cutoff just below half the incoming data's sample rate. This behavior remained the same when WAV files were played back from a USB drive. Channel level-matching was excellent, and channel separation from the Line Out jacks was >100dB in both directions below 1kHz, this decreasing to an okay 70dB at 20kHz due to the usual capacitive coupling between the left and right circuits.
Fig.1 Naim Uniti, D/A frequency response at 12dBFS into 100k ohms with sample rate set to: 96kHz (left channel cyan, right magenta), 44.1kHz (left green, right gray), 32kHz (left blue, right red). (0.25dB/vertical div.)
Playing back a dithered 1kHz tone at 90dBFS from CD, the spectrum of the Uniti's output featured a peak just kissing the 90dB line, as expected, and was free from harmonic- or power-supplyrelated spuriae (fig.2). However, switching to 24-bit data fed to the S/PDIF input dropped the noise floor by just 6dB at high frequencies and not at all at low frequencies, suggesting that the resolution of the Uniti's D/A section is limited by self-noise to around 17 bits. The lowest trace in fig.2, which was taken with 24-bit data representing a dithered tone at 120dBFS, therefore barely resolves the tone. Fig.2 was taken with a swept bandpass filter; fig.3 repeats the measurements using an FFT techniquethe result is the same, which will be fine for CD and iPod playback, of course. The background noise obscures to some extent the shape of an undithered tone at exactly 90.31dBFS (fig.4), though a 24-bit version of the data gave a relatively good sine waveform (fig.5).
Fig.2 Naim Uniti, 1/3-octave spectrum with noise and spuriae of dithered 1kHz tone at 90dBFS with 16-bit data (top) and 24-bit data (middle at 2kHz), and dithered 1kHz tone at 120dBFS with 24-bit data (bottom at 1kHz). (Right channel dashed.)
Fig.3 Naim Uniti, FFT-derived spectrum with noise and spuriae of dithered 1kHz tone at 90dBFS with 16-bit data (left channel cyan, right magenta) and 24-bit data (left channel blue, right red).
Fig.4 Naim Uniti, D/A waveform of undithered 1kHz sinewave at 90.31dBFS, 16-bit data (left channel blue, right red).
Fig.5 Naim Uniti, D/A waveform of undithered 1kHz sinewave at 90.31dBFS, 24-bit data (left channel blue, right red).
Harmonic and intermodulation distortion from the Uniti's D/A section were both low. The second harmonic was the highest in level with a full-scale tone, but at 96dB (0.0015%) will not be bothering anyone. The 1kHz difference tone resulting from an equal mix of 19 and 20kHz tones, each at 6dBFS, lay at 100dB (0.001%). Any jitter-related spuriae with CD playback was below the threshold of the Miller Jitter Analyzer; however, I did measure 78 picoseconds peakpeak when the same J-Test signal was played back from a USB drive, this rising to 236ps pp for iPod playback, and to 454ps pp for an external TosLink datastream. Fig.6 shows the spectrum of the Uniti's output when the signal was played back from an iPod Touch. While only a pair of jitter-related sidebands are visible either side of the 11.025kHz tone, the central peak representing that tone is broadened at the base, suggesting the presence of very-low-frequency noise-like jitter, this most likely stemming from the iPod itself.
Fig.6 Naim Uniti, high-resolution jitter spectrum of analog output signal, 11.025kHz at 6dBFS, sampled at 44.1kHz with LSB toggled at 229Hz, 16-bit data sourced from iPod Touch. Center frequency of trace, 11.025kHz; frequency range, ±3.5kHz (left channel blue, right red).
I then turned to the Uniti's performance as an amplifier. Before I test an amplifier, I run it for 60 minutes at one-third its specified power into 8 ohmsthermally, this preconditioning is the worst case for an amplifier with a class-B or -AB output stage, and taxes the amplifier's ability to dissipate waste heat. With the Naim driving 16Wpc into 8 ohms, the amplifier turned off after 25 minutes, with the chassis comfortably warm and the front-panel display telling me, "Too Hot: cooling down." The amplifier turned itself on again 10 minutes later with no harm done.
Measured at the speaker terminals, the maximum gain for an analog input was 36.7dB into 8 ohms, and the amplifier preserved absolute polarity; ie, was non-inverting. The line-stage input impedance was a moderate 21k ohms at low and middle frequencies, dropping slightly but inconsequentially to 20k ohms at 20kHz.
The Uniti's output impedance was higher than the norm for a solid-state design, at 0.27 ohm across most of the band, rising to 0.28 ohm at 20kHz. As a result, there was a modest modulation of the amplifier's frequency response by ±0.2dB when it drove our standard simulated loudspeaker (fig.7, gray trace). Otherwise, the Naim offered a basically flat response up to 10kHz, with an increasing ultrasonic rolloff with decreasing load impedances. Into 8 ohms, the output was down 3dB at 53kHz, which correlates with the slowed-down leading edges of the Uniti's reproduction of a 10kHz squarewave (fig.8); no overshoot or ringing is visible, however. Channel separation at the speaker terminals was good, at 77dB across the band, decreasing to 70dB above 20kHz. The wideband signal/noise ratio, taken with the input shorted and the volume control at its maximum, was a good 68.7dB ref. 1W/8 ohms, improving to 79.8dB when A-weighted.
Fig.7 Naim Uniti, frequency response at 2.83V into: simulated loudspeaker load (gray), 8 ohms (left channel blue, right red), 4 ohms (left cyan, right magenta), 2 ohms (green). (1dB/vertical div.)
Fig.8 Naim Uniti, small-signal 10kHz squarewave into 8 ohms.
Distortion products were buried under the amplifier's noise floor up to actual waveform clipping, as indicated by the downward slopes of the traces in fig.9. This graph also shows that, at 1% THD+noise, the Uniti just exceeded its specified output power of 50W into 8 ohms (17.0dBW), delivering 55Wpc into that load (17.4dBW). However, the amplifier met its specified 4 ohm power of 90W (16.5dBW) at only 3% THD+N, rather than the usual 1%. I didn't test the Uniti's clipping power into 2 ohms because it quickly overheated and shut off before the output reached its maximum into this punishing load.
Fig.9 Naim Uniti, distortion (%) vs 1kHz continuous output power into (from bottom to top below 10W): 8, 4 ohms.
Fig.10 shows that the small-signal THD+N percentage doesn't vary significantly with frequency, at least in the audioband. Though it does increase as the load impedance drops, this is not to anything alarming, even into 2 ohms (green trace). At low powers, the distortion itself seems predominantly third-harmonic in nature (fig.11), though at high powers the third is joined by the second harmonic, as well as by higher-order odd harmonics (fig.12), albeit at a low level. A spectral spike can be seen at 120Hz and 84dB (0.006%) in fig.12; I experimented with the grounding between the Naim and the Audio Precision test set, and also tried lifting the Uniti's chassis ground with the rear-panel switch, but could not eliminate this supply-related component.
Fig.10 Naim Uniti, THD+N (%) vs frequency at 4V into: 8 ohms (left channel blue, right red), 4 ohms (left cyan, right magenta), 2 ohms (green).
Fig.11 Naim Uniti, 1kHz waveform at 4.4W into 4 ohms (top), 0.018% THD+N; distortion and noise waveform with fundamental notched out (bottom, not to scale).
Fig.12 Naim Uniti, spectrum of 50Hz sinewave, DC10kHz, at 55W into 4 ohms (left channel blue, right red; linear frequency scale).
Finally, fig.13 shows the spectrum of the Naim Uniti's output while it drove an equal mix of 19 and 20kHz tones into 8 ohms at a level a little below visible waveform clipping on the oscilloscope display. The highest-level intermodulation product is the difference tone at 1kHz, which lies at 74dB (0.02%). All the other tones are lower in level. The picture didn't change significantly into 4 ohms.
Fig.13 Naim Uniti, HF intermodulation spectrum, DC24kHz, 19+20kHz at 30W peak into 8 ohms (linear frequency scale).
Although the Naim Uniti didn't excel in any way on the test bench, overall the balance of its measured performance impressed me. It was also delightfully intuitively easy to use, which is something I am not able to say as much as I would like. I also liked the fact that when an iPod was plugged into the Uniti's dedicated input, the amplifier's display showed its contents, sorted by Playlist, Genre, Artist, etc., with navigation of that content now provided by the Uniti's remote.John Atkinson
Excellent review, I have recently plumped for the revised Uniti2 which has a little more power than the original, very pleased with the musical performance despite it being a one box system.