The Unseen Variable Page 3

Test 1, WATT V: The Watt V is nominally a 5-6 ohm loudspeaker. The calibrated loss in loudness due to the simulated source resistance was therefore a substantial 3.6dB. (Note that with an even-lower-impedance speaker, the output impedance of a tube SE amplifier will result in an even greater loss in loudness.) No sense could be made of the comparison, so I raised the volume by 4dB to match the subjective loudnesses.

What a fascinating result! The resistor feed dramatically shifted the tonal balance of the Wilson WATT, with the general view that it now sounded more like a free-field-optimized design. In particular, the bass was louder and better balanced with the series resistor, revealing pretty good power and extension. The midrange sounded open, fast, and articulate, the treble airy and sparkling. The balance was a little brighter than before, but not to the point of upsetting the overall presentation. If anything, stereo perspectives were judged to be superior, with increased spaciousness and a more holographic projection of midrange depth.

Forgive me if I point out at this juncture that some of these characteristics are often promoted by the manufacturers of costly SE tubed amplifiers, and yet here they were achieved by the simple use of a resistor costing just 50 cents!

It must also be noted that, while the above does describe a genuine quality improvement in the case of the WATT V "misused" in a free-field application, there were none of the engineering elements present that could be associated with better sound—such as a better amplifier.

Test 2, Tempo: While the calculated loss was 3dB, given its nominal 8 ohm load, the subjective loudness of the Tempo was not greatly impaired by the 3.3 ohm series resistor. If anything, certain frequency regions, such as the midbass, were slightly louder. Overall, without any attempt at level matching, the bass loudness was similar but less even, with a loss of extension. Interestingly, the bass did not sound significantly boomier or less well-controlled. On a more subtle level, the listeners perceived losses in bass rhythm and impact.

Perceptible shifts in tonality were apparent—there seemed to be more presence and edge definition, yet these were achieved without hardness or glare. The sound appeared to have opened out somewhat. And, again, a feeling of enhanced focus was heard in the midrange, though the impression of depth was now thought to be impaired.

Remember that the complexity of the human hearing process will mean that some of these observed effects result from subtle loudness changes alone. For example, one effect of a slight drop in level is to increase the sense of distance from the perceived soundsources.

For the next set of comparisons, the replay level with the series resistor was increased by 2.5dB to restore the subjective impact and loudness of the program material. Now those previously misleading interpretations of fidelity caused by level reductions in certain frequency regions were blown away. This speaker, with its relatively neutral balance, was recognized as showing clear signs of a warped personality with the series resistor!

The Tempo sounded less neutral with the series resistor. Its vocal balance, for example, was lightened and higher pitched. Cellos were emasculated, sounding more like violas than cellos, while the treble was projected. This can be attractive on some dull-sounding material, but causes undesirable emphasis on others. The Tempo's bass was now unquestionably inferior to that with normal drive. Less even and well-controlled, the low frequencies were less a part of the whole sound. By whatever psychoacoustic association was at work, the bass appeared to lose drive and impact, percussive power and rhythm.

On the plus side, gains were perceived in the midrange focus, depth, and perspective—these are the sorts of changes a speaker designer can introduce by judicious control of frequency response within a specified tolerance (see the sidebar). Often a price has to be paid for such indiscretions, different kinds of music revealing them as false tonal emphases or colorations.