Dynaudio Special Twenty-Five loudspeaker Measurements
I estimated the Special Twenty-Five's voltage sensitivity at 89.3dB(B)/2.83V/m, 1dB higher than the specified 88dB. The speaker is specified as a 4 ohm load, but as its impedance plot (fig.1) reveals, it drops to this figure for only a small range of frequencies in the lower midrange. The impedance actually stays above 6 ohms for most of the audioband, which, in conjunction with the generally modest electrical phase angle and the higher-than-average sensitivity, suggests that the Special Twenty-Five is a fairly easy amplifier load.
Fig.1 Dynaudio Special Twenty-Five, reflex electrical impedance (solid) and phase (dashed). (2 ohms/vertical div.)
This plot was taken with the reflex port open—the saddle centered at 39Hz in the magnitude plot in this graph indicates the port tuning frequency. Inserting the supplied foam plug results in a single impedance peak in the bass at 63.5Hz (fig.2), implying modest bass extension in this mode.
Fig.2 Dynaudio Special Twenty-Five, sealed-box electrical impedance (solid) and phase (dashed). (2 ohms/vertical div.)
Although it cannot be seen at the scale this graph is printed, there is a very slight wrinkle between 400Hz and 500Hz in both the amplitude and phase plots. Investigating the cabinet's vibrational behavior with a simple accelerometer revealed a resonance mode present at 473Hz on the side panel (fig.3). Though this mode is quite high in level, it is also high enough in frequency and decays sufficiently quickly that it should have a minimal effect on the speaker's sound quality. A mode at 240Hz could be found on the top panel, but this was well down in level.
Fig.3 Dynaudio Special Twenty-Five, cumulative spectral-decay plot calculated from the output of an accelerometer fastened to the center of the cabinet's side panel (MLS driving voltage to speaker, 7.55V; measurement bandwidth, 2kHz).
Turning to the frequency domain, the rather complicated plots in fig.4 show: the overall response on the tweeter axis, averaged over a 30º horizontal angle (black trace), spliced to the complex sum of the nearfield woofer and port outputs (also black); the nearfield woofer response with the port open (blue); the nearfield port response (green); and the nearfield woofer response with the port blocked by foam (red).
Fig.4 Dynaudio Special Twenty-Five, anechoic response on tweeter axis at 50", averaged across 30º horizontal window and corrected for microphone response, with the complex sum of the nearfield woofer (blue) and port (green) responses, taking into account acoustic phase and distance from the nominal farfield point, plotted below 300Hz. Red trace shows the nearfield woofer response with the port blocked by the foam plug.
Comparing the red and black traces, it can be seen that blocking the port drops the speaker's midbass by just above 6dB at 42Hz, the lowest note of the four-string double bass and bass guitar. This will compensate quite effectively for the boom that would otherwise result if the speaker has to be used very close to the wall behind it. The minimum-motion notch in the woofer's output when reflex-loaded occurs at 39Hz, the frequency predicted by the impedance plot, and the port's output peaks between 30 and 50Hz. A notch at 300Hz in the port response indicates some kind of anti-resonance at that frequency, and while there are a couple of peaks apparent at higher frequencies, these are both well down in level.
Moving higher in frequency, the Special Twenty-Five's overall response looks impressively flat, with little evidence of the 2.4kHz crossover point. However, a slight excess of energy can be seen in the upper midrange and the tweeter appears to be balanced around 2dB too high in level. Both of these correlate with the slightly bright tonal balance I noted in my auditioning. Interestingly, while my measured response correlates in some respects with that published by Ultimate AV in their October 2004 issue, that magazine's samples had a slightly less hot treble. However, in his review notes, Joel Brinkley commented on a balance that was too forward in the upper midrange, which echoes my own feeling.
The Dynaudio's lateral dispersion is shown in fig.5. There is only a slight sign of the 8" woofer suffering from the expected limited dispersion at the top of its passband, and though there is an excess of off-axis energy apparent just above the crossover point, this will compensate for a small suckout at the same frequency in the Twenty-Five's on-axis output. The tweeter's response drops rapidly to the sides above 10kHz, which will perhaps make the speaker sound somewhat lifeless in very large rooms. More likely, however, it will leave the mid-treble a little emphasized. In the vertical plane (fig.6), a big suckout rapidly develops in the crossover region as the microphone moves above the tweeter axis. Despite its bulk, the Dynaudio needs fairly high stands to ensure that the listener's ears remain on or just below the tweeter axis.
Fig.5 Dynaudio Special Twenty-Five, lateral response family at 50", normalized to response on tweeter axis, from back to front: differences in response 90–5º off-axis, reference response, differences in response 5–90º off-axis.
Fig.6 Dynaudio Special Twenty-Five, vertical response family at 50", normalized to response on tweeter axis, from back to front: differences in response 45–5º above axis, reference response, differences in response 5–45º below axis.
Fig.7 shows the Dynaudio Special Twenty-Five's spatially averaged response at the listening position in my room, with the speakers in the preliminary, more free-space locations. Though there is a creditable amount of bass evident, down to the 25Hz 1/3-octave band, the 50Hz and 63Hz bands are a little suppressed and the entire LF region is shelved-down slightly compared with the upper midrange and treble. The overall response trend above the region dominated by room acoustics (below 400Hz or so) is smooth, but tilted-up between 1kHz and 10kHz, rising by 3dB in that region.
Fig.7 Dynaudio Special Twenty-Five, spatially averaged, 1/3-octave response in JA's listening room.
In the time domain, the Twenty-Five's step response (fig.8) reveals that, despite the speaker's use of first-order crossover filters, it is not time-coincident, the tweeter's positive-going step arriving at the microphone about 200µs before the woofer's. However, the speaker is time-coherent, in that the overshoot of the tweeter's step smoothly blends with the rising edge of the woofer's step, which correlates with the excellent integration between the drive-units on the tweeter axis in the frequency domain. The cumulative spectral-decay plot on the tweeter axis is superb (fig.9), with a very clean initial decay.—John Atkinson
Fig.8 Dynaudio Special Twenty-Five, step response on tweeter axis at 50" (5ms time window, 30kHz bandwidth).
Fig.9 Dynaudio Special Twenty-Five, cumulative spectral-decay plot at 50" (0.15ms risetime).