MANUFACTURER'S SPECIFICATIONS:
Type: Three-way acoustic suspension.
Driver Complement: 12-in. (30 cm) cone woofer, 3 1/2-in. (8.8 cm) beryllium dome tweeter, 1-in. (3 cm) beryllium dome tweeter.
Crossover Frequencies: 500 Hz and 6 kHz.
Nominal Impedance: 8 ohms.
Output Sound Pressure Level: 90 dB at one meter for one watt input.
Dimensions: 28 in. (71.1 cm) H x 15 1/2 in. (39.4 cm) W x 14 1/2 in. (36.2 cm) D.
Weight: 89 lbs. (39 kg).
Price: $1450.00 per pair.
= = = = = =
The NS-1000 is a recent addition to the Yamaha Natural Sound series of loudspeakers. This system comes in fairly similar versions, the NS-1000 with decorative grille, and the NS-1000M, a version which Yamaha calls the monitor and which is devoid of a front cover except for a metal grid placed over the woofer.
While of a size that it could be considered a bookshelf type speaker system, the NS-1000 is exceptionally robust and quite heavy. At 39 kg (89 lbs.) I would never recommend placing this smooth-sided system on any overhead bookshelf, no matter how sturdy the shelf might be. In my opinion this speaker should be floor mounted in consideration of user safety. The robust nature of this speaker can be appreciated by the ratio of its weight to volume, almost 24 pounds per cubic foot. More than normal precautions in lifting and moving this speaker should be employed, particularly in view of the fine furniture finish on all exposed surfaces.
Careless attempts to move this speaker around the house can translate into instant hernia or a mar in the finish. And, in my opinion, this speaker should never be placed where a small child could accidentally cause it to topple or slide off a smooth surface.
This is a three-way speaker system, using a 30-cm woofer in a sealed enclosure, complemented by an 8.8 cm midrange and a 3 cm tweeter. The midrange and tweeter are of unusual design since they use a thin beryllium diaphragm which, be cause of its higher ratio of stiffness to mass, is considered to offer superior diaphragm performance. Beryllium, however, is quite brittle and protective metal covers are used to protect them against physical contact.
A removable decorative grille covers the front of the NS-1000, and the sides are ebony-wood veneer, hand rubbed to a fine finish with no mounting feet or protuberances. The rear of the enclosure is unfinished black.
Fig. 1--Magnitude of impedance for the three equalizer settings.
Fig. 2--Polar plot of impedance for Normal equalizer settings.
Fig. 3--One-meter axial, anechoic amplitude response for constant voltage
drive corresponding to one-watt average power in an 8-ohm resistor with equalizer
settings at Normal.
Fig. 4--One-meter axial phase response corrected for the air-path delay
of the midrange.
Speaker connection is made to push-type terminals mounted in a recessed cavity in the rear of the enclosure.
Polarity is clearly marked, and users should have no difficulty hooking these speakers up if the instructions in the accompanying manual are followed. However, there is one important point which may need further clarification. These speakers come as a mirror-image pair one for left channel and one for right channel. The manual makes a peripheral comment that there is a left and a right speaker but does not say how to identify which is which. In the units delivered for review, there was a smudged letter R following the serial number on following the serial number on one speaker and a slightly less smudged L on the other speaker. This is quite out of keeping with the neat and explicit construction of these speakers and I hope that is not the case with all NS-1000s.
The midrange and tweeter crossover at 500 Hz and 6000 Hz respectively. Continuous-rotation attenuator controls are mounted on the front panel for the midrange and tweeter, and three positions are marked for suggested settings. These are-3, Normal, and +3, and their effect on frequency response is shown in the owner's manual. The manual itself is moderately sparse in detail, but adequate for proper connection and operation of the speakers.
Fig. 5-One-meter axial phase response corrected for air-path delay of the
tweeter.
Fig. 6-Three-meter room response.
(left) Fig. 7-Horizontal polar energy response with grille in place.
(right) Fig. 9-Effect of grille on horizontal polar energy response.
(left) Fig. 8-Vertical polar energy response with grille in place. (right)
Fig. 10-Effect of grille on vertical polar energy response.
Technical Measurements
The measured magnitude of impedance as a function of frequency is shown in Fig. 1. In order to show the variation of impedance with various control settings, three measurements are presented. The curve marked Normal is the impedance which an amplifier sees when both the mid and high frequency front-panel speaker controls are set to the position which Yamaha indicates as Normal. The +3 curve is the impedance when both controls are rotated to their indicated +3 position. Similarly for the -3 curve. Figure 2 is a complex impedance plot for the Normal setting.
The bass resonance peak occurs at 38 Hz and has a value around 41 ohms. The lowest impedance for all settings occurs at around 100 Hz and has a value around 5.5 ohms. Both impedance plots show that there should be no difficulty driving the NS-1000 from any good power amplifier. Reactance effects at higher frequencies are minimal, and the worst case reactive value occurs at 700 Hz with a 33-degree lagging phase and a magnitude of impedance of 11 ohms.
With the exception of a small glitch at 70 Hz, the smoothness of the complex impedance curve shows the NS-1000 has no indication of structural resonance in the enclosure. The box is rigid and well built, a fact one can verify by striking the side of the NS-1000 enclosure with the heel of one's hand.
There is no drum resonance which is often encountered in less rigidly built enclosures.
The measured one-meter, anechoic, axial sound pressure response is shown in Figs. 3 and 4 for the Normal position of both equalizers. The free-field response showed the existence of a small problem which is surprising in a loudspeaker with the overall quality of the NS-1000 the grille assembly noticeably alters the frequency response. Both the amplitude response of Fig. 3 and the phase response of Fig. 4 are plotted twice, once with the grille firmly in place and once with the grille removed. The problem is not the grille cloth, but the metal frame on which the cloth is placed. The frame extends in front of the drivers and scatters sound. Although the effect is not substantial in this, the axial position, the off-axis measurements, which Audio also checks, shows a more considerable perturbation due to the presence of the grille assembly.
My personal recommendation for obtaining the very best sound from the NS-1000 is to remove the grille entirely.
Audio's frequency response measurements agree quite well with the response which Yamaha publishes for this type of system. The mild interference dip at 7 kHz and the drop in top end above 19 kHz are effects seen with the sound measurement made on the geometric axis of the system, which is our standard measurement position. Sound pressure measurements made at one meter directly in front of the tweeter are identical to those furnished by Yamaha. Free-field, low frequency response extends down to 55 Hz with a uniform fall below that frequency. Frequencies up to 5 kHz are handled quite smoothly with a minimum amount of irregularities in sound pressure from 5 kHz to the upper cutoff around 19 kHz.
Fig. 11--Harmonic distortion for the tones E, (41.2 Hz), A2 (110 Hz), and
A, (440 Hz).
Fig. 12--Intermodulation distortion of A4 (440 Hz) by E1 (41.2 Hz) mixed
one-to-one.
There is, however, one important technical fact which the Audio phase measurements reveal. The phase shift of the midrange is neither zero nor 180 degrees, but is 90 degrees leading through the frequency range from 2 to 10 kHz. Figure 4 is the phase measurement made with the air-path time delay corrected for the acoustic position of the NS-1000 in the 2-to-10-kHz range. What this shows is the net phase change due solely to the loudspeaker. The value of this phase is dependent somewhat upon the equalizer control settings, and these plots are made with the controls carefully set to the Normal positions. Figure 5 is the phase measurement corrected for the air-path time delay from the tweeter.
Here it can be seen that the frequency range above 10 kHz has a sound pressure which is in phase with the driving volt age. Low frequency measurements, not shown here, reveal that the woofer is also in phase with the driving voltage.
The effect of this 90-degree phase change in the midrange can also be observed in the impulse response, not included as a standard plot but also routinely made in our Audio evaluations. The sound from the tweeter arrives approximately 0.08 milliseconds before the sound from the midrange. But the initial pressure peak due to the tweeter is just completing its first period when the midrange pressure peak arrives. The positive 90-degree frequency response phase of the midrange means that the pressure due to the midrange swings first to an overpressure then to a comparable under-pressure before decaying to equilibrium. The result of all this is that the tweeter and midrange complement each other to produce an initial wave packet which is well behaved. Highly technical, but the result is a synergism giving good transient response.
The measured three-meter room response is shown in Fig. 6. Unlike the anechoic response, the room response represents the frequency spectrum of the first 11 milliseconds of sound which one hears from the system when seated three meters away from the front of the speakers and in a typical room. For this measurement the NS-1000 was mounted in the room used for the listening test and in a fashion recommended by Yamaha with the speaker raised slightly off the carpeted floor on cinder blocks. The back of the speaker was placed against a hard wall. Two listening positions are shown for this measurement. One position is three meters directly in front of the speaker, and the other is at the same distance but 30 degrees off center, representing a normal stereo listening configuration for this system. The grille was in place and both equalizers set to the Normal position.
Both positions provide a very smooth response, with the normal stereo listening position somewhat more uniform.
There is some ceiling scatter in evidence, which gives rise to the small irregularities above 4 kHz. A check of the phase response (not shown here) showed an essentially constant average phase about 1 kHz. This indicates a good transient response.
Horizontal and vertical polar energy response is shown in Figs. 7 and 8 respectively. Three curves are shown, corresponding to both equalizer controls set to the-3, Normal, and +3 positions. The NS-1000 comes as a set, with one speaker intended for left channel use and the other for right channel use. The left channel speaker was used for this measurement. The horizontal plot shows that indeed the left channel speaker has a more uniform dispersion when heard at a position to the left of its central axis. Similarly, the vertical measurement shows that sound is launched in an upward direction, as befits a floor-mounted system. Horizontal dispersion is quite good, and the measurements indicate that little is to be gained by rotating the speakers toward the listening area. The energy difference between the settings is almost exactly the 3 dB indicated on the controls.
The influence of the grille assembly on both the vertical and horizontal dispersion is strong. Figures 9 and 10 are plots of horizontal and vertical dispersion patterns with and without the grille in place, and it will be seen that the grille structure introduces lobes in both patterns. Again, I recommend removal of the grille assembly for most accurate sound.
Harmonic distortion for the tones of E, (41.2 Hz), A2 (110 Hz), and A4 (440 Hz) is quite low up to the maximum average level which the speaker is intended to handle. Particularly noteworthy is the low value of harmonic distortion for A2 and A4, though E1, 41.2 Hz, has a bit higher level of distortion than I would like to see in a speaker of this overall high quality, particularly at the lower sound pressure levels. Loudspeaker overload became evident with bursts of 60 average watts at the lowest frequency, so the measurements were terminated at this level, even though the second and third harmonic distortion levels were still not excessive.
The NS-1000 is quite good at handling transients at high level. But a frequency dependence, which was puzzling from a technical viewpoint, was evident in the test which we call the crescendo handling test. The 110-Hz and 262-Hz inner musical voices were absolutely unmodified by the presence of broadband random noise at 20 dB higher average level than those tones. This indicates that wandering in the stereo sound image should not be a problem for those tones when other loud signals are present. But 440 Hz was a "worst case" frequency in the crescendo handling test. The 440-Hz tone was acoustically reduced a nearly uniform 0.3 dB when incoherent noise signals were present over the full dynamic range up to noise peaks of 800 watts. This frequency dependence is a bit unusual, although the 0.3 dB effect is most probably low enough not to be audibly troublesome.
The acoustic transfer linearity of the NX-1000 has a slight downward trend with increasing drive level. A one-decibel increase in drive level produces slightly less than one-decibel increase in sound level. The effect was checked at 60 Hz, Middle C (262 Hz), and A4 (440 Hz) and found to be a decrease of about a hundredth decibel of sound per decibel of electrical drive. This will impart a very small, probably insignificant, amount of signal compression to program material.
What this means is that the sound does not get as loud as it should for increases in drive level.
(left) Fig. 13--Energy-time response with grille in place. (right)
Fig. 14--Energy-time response with the grille removed.
The cross-modulation of A4 (440 Hz) when a tone of E, (41.2 Hz) is present at equal level is shown in Fig. 12. At moderate listening levels, this type of distortion is quite low in the NS 1000 and is composed of both amplitude modulation and phase modulation of the higher tone by the presence of the lower tone. At 10 average watts, the amount of phase modulation is 3 degrees peak to peak and the amplitude modulation is of the order of 2 per cent peak to peak. Even up to the highest tested power of 100 watts, the character of IM remained the same. Net displacement of the average position of the woofer was less than one degree at 440 Hz when 41.2 Hz was present. In other words, the higher frequency tone did not have its arrival time altered by the presence of the lower tone. Many loudspeakers do this; the Yamaha does not. While the harmonic distortion measurement shows that at high sound levels the woofer can create subtle changes of timbre on low bass tones, the IM test shows that the same tones at high level do not substantially intrude on upper register instruments to modify orchestral tutti.
The energy-time curve, which is a measure of transient performance, is shown in Figs. 13 and 14. This is the envelope of the impulse response measured one meter on axis. The equalizer controls on the NS-1000 are set to their Normal position for both tests. Figure 13 is the energy-time response when the grille is in place, and Fig. 14 is the same test with the grille removed. The better response is clearly obtained when the grille assembly is removed. As discussed in the phase test, the in-phase contribution of the tweeter melds smoothly with the quadrature phase contribution of the midrange driver to produce a smooth wavelet for the first few tenths of millisecond. The first sound from the tweeter begins to arrive at the microphone at 2.85 milliseconds on this scale. The first midrange sound starts at around 2.93 milliseconds. Neither of these facts can be discerned from Figs. 13 or 14 because the amplitudes of these two sounds combine so smoothly. When the grille is removed, there is some internal reverberation in the 3.1 to 3.2 millisecond interval and then the sound energy falls off smoothly to a level some 35 dB down from the peak at 3 milliseconds. The effect of the grille is to create the equivalent of a "slap" echo at a 0.15 millisecond period decaying at a rate of around 30 dB per millisecond. In addition there is a subsidiary reflection component around 3.8 milliseconds caused by the presence of the grille frame. I realize that Yamaha has gone to great pains to create an attractive grille assembly, but it is my opinion that the sound will be more accurate with this particular grille removed.
Listening Test
In accordance with Yamaha's recommendations, the speakers were positioned flat against a hard wall and placed on cinder blocks. This raised the speakers approximately 17 cm off the floor.
In order to get at the mid and high frequency equalizer controls, it is necessary to remove the grille assembly. Removal is quite simple, but both enclosures reacted to my attempts at reinstallation by popping out on the side opposite to the one I was pressing into place. Yamaha's explicit instructions to the contrary, this is a two-hand job.
My initial impression was that the NS-1000 had a bass dominance on percussion but sounded quite natural on pipe organ. After considerable experimentation, I found that a preamplifier tone-control equalization which dropped the 50-Hz level by 2 dB relative to 1 kHz was about right to my ears for bass response. When listening to vocals I sensed a need for a slight increase in level in the upper midrange. Rotating the midrange control on the NS-1000 halfway between Normal and +3 filled the need.
Stereo lateralization is good on the NS-1000. More significantly, the illusion of depth is quite excellent. In order to achieve a good stereo illusion, it is not necessary to sit in one spot; the illusion remains good over a large range of listening positions, though this freedom of position is marred by vertical beaming in these speakers. You can move from side to side and the image remains fixed, but stand up and the image shifts and slightly blurs. The part of the spectrum which is most affected by this vertical change in listening position is that from the higher midrange and on up.
Although these speakers can handle quite a large dynamic range and produce lease-breaking sound levels, I sensed a top end crunch on some clean material which was reproduced at very high level. There is a bite to the top end of this speaker which, to my ears, produces a spectral dominance at around 9 kHz. When driven very hard on sharp percussion, this bite seems to have difficulty relative to the rest of the reproduced spectrum, and appears to overload. Of course, this is at sound levels well above that which most users might demand from their speakers, so that it will rarely, if ever, be encountered.
The Yamaha is an accurate sounding system when properly placed in a room. And although it does benefit from some minor spectrum adjustments in the driving signal, this system requires nothing more than a clean signal to deliver a good, clean sound.
--Richard C. Heyser
(Source: Audio magazine, Jan. 1979)
Also see: Stereophile 1976 review.
Yamaha NS-2000 Loudspeaker System (Auricle, Sept. 1984)
= = = =