Loudspeakers

From the disposition of its drivers, the intended listening axis for the SCM100SE appears to be on the axis of the midrange cone, which is at typical seated ear height. Our initial frequency response measurements were made at this microphone height and above and below it, which confirmed that the midrange axis gave the flattest response. Measured at 1m on this axis, the responses of the two speakers [see Graph 1, below] show a decline in output below 300Hz which should be ignored as this is an artefact of the measurement time window used. Otherwise there is clear evidence of a 'BBC dip' through the presence band but it only amounts to about 1.5dB and overall the response is very well controlled indeed, the errors being only ±2.1dB and ±2.3dB respectively – excellent results for a loudspeaker lacking DSP equalisation.

Pair matching over the same 300Hz-20kHz frequency range is good too at ±0.7dB and would have been ±0.5dB but for a very narrowband disparity at 5kHz. Diffraction-corrected nearfield measurement showed the bass extension to be 34Hz (–6dB re. 200Hz), so there's no doubt that with a little LF boundary gain the SCM100SE will provide useful output to the very lowest audible frequencies. Furthermore, bass output begins to decline gently from 70Hz which should make room integration easier. ATC uses an underhung (short coil, long gap) motor geometry in both the bass driver and midrange driver. so it's no surprise that they both achieve low distortion at 90dB SPL, recording 0.2% at 100Hz and <0.1% at 1kHz, respectively.

The exact figure for the midrange unit was 0.048% with second, third and fifth harmonics predominating. But from the cumulative spectral decay waterfall [Graph 2] it appears that the soft dome midrange unit is not entirely blameless with regard to breakup resonance, with a clear mode at around 5kHz and hints of other, poorly resolved modes at lower frequencies. KH