LABORATORY TEST RESULTS

Newport Test Labs first measured the in-room frequency response of the Harbeth Monitor 30.1 using a pink noise test stimulus, averaging nine different sweeps measured at three metres, the result of which is shown in Graph 1. You can see the response is very flat across the midrange. From 220Hz up to the graph limit of 10kHz, it’s within ±1.25dB. The plus/minus decibel level variations are not uniform, however, with the response at +1.25dB at 220Hz then reducing gradually in level with increasing frequency to be –1.25dB at 2kHz, after which the response continues almost ruler-flat at this same level right out to the graph limit. So, overall, the output level drops 2.5dB over the decade from 220Hz to 2kHz.

The anechoic high-frequency response of the Harbeth Monitor 30.1 is shown in Graph 2, with the gating technique used by Newport Test Labs making it impossible to measure below 400Hz for this particular test. The graph shows both the frequency response without the grille fitted (the black trace) and with it fitted (the red trace). Perhaps the first thing to note is that Harbeth’s grille design is excellent, because unlike most grilles, it is truly acoustically transparent, having very little effect on the performance of the speaker. Because of this, I would recommend you leave the grilles in place when listening—even for critical listening sessions. You can see the response is very flat and linear out to 5kHz, where there’s a slight rise followed by a 2.5dB suck-out centred at 7.5kHz, followed by a rise to +2.5dB at 9kHz, after which the response gradually rolls off to –2.5dB at 17kHz, after which it abruptly picks up and extends out beyond 20kHz before rolling off to be around 20dB down at 35kHz.

I assume the abrupt transition at 17kHz is an artefact caused by the metal grille protecting the dome tweeter.

Although interesting from a technical view-point, it will have no bearing on the speaker’s sound as it happens at a frequency higher than most people are able to hear, and even those people who can hear 17kHz signals could not hear such a small difference in the level of a signal at this frequency. As for those variations below 10kHz that I drew to your attention, the human ear would not perceive those either, as they’re too narrow in bandwidth and too small in level to be audible. It’s only the extreme precision of the measurement technique used that makes them visible on the trace at all.

Newport Test Labs measured the low-frequency response using a near-field technique to simulate how the Harbeth Monitor 30.1 would respond in an anechoic chamber, the result of which is shown in Graph 3. You can see that the bass/midrange driver’s output starts its roll-off at 100Hz and then rolls off at 18dB per octave to a minima at 45Hz. Of course the output of the front-firing bass reflex port partially compensates for the losses, as you can see from the red trace. The port’s contribution is fairly limited, however, though it does deliver that contribution over a wide range of frequencies, being –3dB at 27Hz and 95Hz.

The impedance of the Harbeth Monitor 30.1 only drops below 6Ω at 160Hz (and then only to 5.8Ω), and is also mostly above 8Ω, so it’s higher than I usually see on most modern speaker designs with 8Ω nominal impedance ratings, but it means that the Monitor 30.1 is really a true 8Ω design. The kink in the impedance trace reveals the presence of a (presumably panel) resonance at this frequency, and the effect can aslo be seen both on the phase response and on the frequency response of the bass/midrange driver. The saddle between the two resonant peaks is at 45Hz, showing that you will get no effective output below this frequency. The rather high impedance of the system around 1.3kHz (it peaks at 27Ω) means some amplifiers might reduce their output in this region as a result, so amplifier matching will assume greater importance with this design than it might with some other speakers.

Graph 5 shows the overall response of the Harbeth Monitor 30.1, as measured by Newport Test Labs, and you can see that despite the overall linearity and the flatness of the midrange, there’s still a slight spectral tilt to the spectrum that sees the output higher at low-frequencies than it is at higher frequencies. Despite this, the overall response measured by Newport Test Labs for the Harbeth Monitor 30.1 was 48Hz to 30kHz ±4dB, which is self-evidently excellent.

Newport Test Labs measured the sensitivity of the Harbeth Monitor 30.1 as being 86.5dBSPL at one metre for a 2.83Veq input. This is a touch higher than Harbeth’s own specification of 85dBSPL, proving that Harbeth is obviously being conservative with its rating, but it is lower than the average for all speakers, so I’d suggest that you use a fairly high-powered amplifier, of at least 60 or 70-watts per channel to drive them unless you have a small room and listen at lower levels, in which case 40–50-watts would likely be sufficient.

Overall, the measurements of the Harbeth Monitor 30.1 show it to be a very well-designed loudspeaker, with a higher efficiency than I would have expected for its size and driver configuration. Steve Holding

Readers should note that the results mentioned in the report, tabulated in performance charts and/or displayed using graphs and/or photographs should be construed as applying only to the specific sample tested.