Stand Loudspeaker Reviews

Measurements
Whereas the 1995 samples of the Mani-2 had a low sensitivity, the Y2K samples measured 84.5dB(B)/2.83V/m. While this is still below average, it is a more reasonable figure. Figs.1 and 2 show the impedance plots of the 2000 and 1995 samples, respectively. They overlay almost exactly, which reveals commendable production consistency. (The two Y2K samples were also very well pair-matched.) However, the use of paralleled bass units means that the impedance magnitude falls to 3 ohms at 200Hz, which will mandate a good high-current amplifier be used.

Fig.1 Totem Mani-2, 2000 sample, electrical impedance (solid) and phase (dashed). (2 ohms/vertical div.)

Fig.2 Totem Mani-2, 1995 sample, electrical impedance (solid) and phase (dashed). (2 ohms/vertical div.)

The individual drive-unit and port responses are shown in fig.3. As with the 1995 samples, the acoustic crossover point appears to be around 3kHz rather than the specified 4kHz, and the woofer shows some peakiness in its top octave and a half. (The 1995 samples were significantly smoother in this region.) The SEAS tweeter, which has its pole-piece vent at the back of the magnet covered by a small, clear-plastic enclosure, features a huge ultrasonic peak at 26kHz. While this will, of course, not be excited to anything like its full extent with CD sources, which have no energy above 22kHz, I did wonder if the presence of top-octave noise on CDs, due to the use of noise-shaping, was getting the resonance going, which in turn led to the HF "wispiness" I occasionally noted.

Fig.3 Totem Mani-2, 2000 sample, acoustic crossover on tweeter axis at 50", corrected for microphone response, with the nearfield woofer and port responses plotted below 300Hz and 1kHz, respectively.

The compound woofer rolls off smoothly below 80Hz or so to reach its minimum-motion point at 36Hz, a little higher in frequency than the maximum port output. And as with the 1995 samples of the Mani-2, very-high-Q resonant peaks can be seen in the port's response at 305Hz, 550Hz, and 640Hz. But the fact that the Mani-2's port faces away from the listener significantly reduces the audibility of this behavior. Even though I knew these resonances were present, I couldn't hear them unless I stood behind the speaker and played pink noise.

The overall response of the Mani-2 on its tweeter axis, averaged across a 30 degrees horizontal window, is shown in fig.4 (Y2K sample) and fig.5 (1995 sample). The graphs have very similar shapes, with very flat overall balances. However, a sharp peak can be seen just above 1kHz with the more recent speaker, which might correlate with the touch of character I noted on saxophone and clarinet. The very sharp spike at 305Hz is due to the lowest in frequency of the port modes and was measurable only in the nearfield. In my room, the spatially averaged response (fig.6) was very even, though with a slightly uptilted nature and some energy excesses in the upper midrange and low treble. The low frequencies are superbly extended, and the low bass can be seen to actually be a little exaggerated in my preferred room positions!

Fig.4 Totem Mani-2, 2000 sample, anechoic response on tweeter axis at 50", averaged across 30 degrees horizontal window and corrected for microphone response, with the complex sum of the nearfield woofer and port responses plotted below 300Hz.

Fig.5 Totem Mani-2, 1995 sample, anechoic response on tweeter axis at 50", averaged across 30 degrees horizontal window and corrected for microphone response, with the complex sum of the nearfield woofer and port responses plotted below 300Hz.

Fig.6 Totem Mani-2, 2000 sample, spatially averaged, 1/3-octave response in JA's Brooklyn room.

Finally, the Mani-2's cumulative spectral-decay plot (fig.7) is superbly free from resonances between 2kHz and 20kHz, this correlating nicely with the speaker's freedom from grain and its clean, clear presentation of recorded detail. There is a large ridge of delayed energy at 26kHz associated with the tweeter's "oil-can" mode, but of more subjective significance is the similar ridge at 1150Hz, the frequency of the on-axis peak. The 1995 sample's waterfall plot (not shown) had neither a peak nor a resonant mode in this region.

Fig.7 Totem Mani-2, 2000 sample, cumulative spectral-decay plot at 50" (0.15ms risetime).

Summing Up
In my original review, I concluded that while Totem's Mani-2 appears to cost a lot for a conventional-looking if well-finished minimonitor, its combination of clean treble, transparent midrange, natural dynamics, and powerful, extended bass allowed the music to communicate in a most effective manner. Although the current sample is a little more forward-balanced in the upper midrange, the Mani-2 is still enthusiastically recommended, particularly if you have a smallish room, value bass extension, and are prepared to pay for it. A solid Class B loudspeaker in this magazine's "Recommended Components."—John Atkinson