Sidebar 4: Martin's Measurements
On axis, the primary frequency response measured respectably flat out to 12kHz with a minor, narrow prominence in the 14–18kHz range that was only present directly on axis. Output continued beyond audibility before finally settling at –10dB at a nicely extended 33kHz. Notably absent was the resonant peak often seen at or above 20kHz; its absence here is no doubt due to this tweeter's pleated-ribbon design (footnote 1). Whether coincidental or not, the high treble sounded sweet and never drew attention to itself.
Tested on axis, pair matching was excellent, the differences holding to ±1.5dB limits from 100Hz to 20kHz. I estimated sensitivity at an industry-typical 88.5dB/2.83V/m. At normal loudness, the two ports (which share the same volume) shared a tuning frequency of about 33Hz. Blocking one port will lower the system tuning frequency meaningfully, to about 28Hz, and taper the alignment if this is useful to deal with increasing room gain at low frequencies. The listening panel and I preferred single-port operation, so we blocked the upper port. Blocking both ports provided a near sealed-box condition, which could be useful in an awkward room or when placed near the front wall. Checked for spurious higher-frequency output, the upper port read a fine –40dB, and in most installations this port will be blocked. The lower port was quieter, at –53dB, and judged inaudible by the panel.
The optimal listening axis was found to lie at a glancing 5° off-axis laterally, halfway between the midpoints of the tweeter and the midrange driver. On this axis from 100Hz to 15kHz, we observed close, ±2dB limits, which suggests excellent driver integration. Good focus was noted here. There is a small prominence at 16kHz, which I think is likely inaudible. Even at 30° off-axis laterally, output was judged very uniform at ±2.5dB from 200Hz to 15kHz, which implies excellent imaging and natural in-room timbre.
I heard the effect of installing the grilles as a slight dulling with some loss of focus and dynamic expression. In terms of measured behavior, the grilles introduced ±3dB response variations in the range of 5–8kHz and a 2dB loss above 12kHz. Most owners will leave the off. In my large room, the bass was extended to 36Hz at –3dB and a low 25Hz at –6dB, though this was more informative of a good phase response than physical bass pressure (footnote 2).
When standing, a discriminating listener would notice a dip in the vicinity of the mid-treble crossover frequency; a seated listener would hear no such effect, as the output was well integrated on and below the tweeter axis. For a person of my moderate stature, this tall speaker sounded best tilted down a few degrees, aimed at my seated ear height; if you sit too far below the axis, the speaker will be less expressive.
Averaged over the room space, I found very uniform frequency response, particularly in the subjectively dominant range from 100Hz to 5kHz; as previously noted, the drivers were well integrated. In room, the Gold 300 6G displayed the expected, gradual directivity-driven decline in the octave up to the 20kHz test limit.
I had heard on the grapevine that these speakers were powerful. They are! They took a 100W sinewave at 35Hz before distortion became audible. They also survived a 20V RMS (50W) short-term signal at 30Hz with only mildly audible distortion; this distortion would likely be inaudible on program music due to aural masking. Further tests found low distortion throughout the audible range, typically measuring 0.1% to 0.15% at realistic listening powers.
Midband harmonic distortion was commendably low with a fairly loud 1V input at 350Hz, a little above middle C. The only harmonics seen were low in level and subjectively innocuous: –59dB (0.12%) of second-harmonic distortion and –62dB (0.08%) of third-harmonic distortion. No other harmonics were above –70dB. This is commendably low-distortion performance, especially in the range where it matters most. Such low distortion may be why I experienced so little listening fatigue.—Martin Colloms
Footnote 1: JA's measurements indicated that the resonance lay above 30kHz. Footnote 2: Loudspeaker bass is a high-pass filter whose order and damping will control bass quality. What you get in a given room will depend on the room acoustic and where object source and subject listener are each placed.
Footnote 1: JA's measurements indicated that the resonance lay above 30kHz. Footnote 2: Loudspeaker bass is a high-pass filter whose order and damping will control bass quality. What you get in a given room will depend on the room acoustic and where object source and subject listener are each placed.
