I concluded Part 1 of my Canon Audio story by promising to begin Part 2 with a description of the third product launched by the fledgling UK specialist speaker company crossed with a Japanese camera multi-national. That third product was the S-30, a compact, stand/wall mount hi-fi speaker intended to offer most of the S-50s audio performance at a significantly reduced retail price. However, in spite of my promise to cover the S-30, I’ve decided it would be instructive first to write a little more about the problems of making Wide Imagine Stereo work in terms of the sound quality expectations of end users, hi-fi retailers and the all important specialist hi-fi press. So the S-30 description is hereby officially postponed to Part 3 – although to whet appetites, there’s a picture of the S-30 on the right.
To recap, if you didn’t read Part 1, or you’ve already forgotten; the premiss of Canon’s Wide Imaging Stereo concept was that an off-set broadly conical acoustic mirror could be employed to widen mid and high frequency dispersion, and that this would result in an enlarged stereo image ‘sweet spot’.
Broadly there was some merit in the WIS proposition, even if the wider stereo sweet spot was achieved at the cost of subjective stereo image focus. It wasn’t so much that the sweet spot was larger, more that the stereo image was more diffuse for nearfield listeners (technically, WIS brings the transition from the nearfield to the reverberant field closer to the speakers). However, there was (and still is) a valid argument I think that for many listeners, even card-carrying audiophiles, high performance speakers that don’t demand that you’re strapped into the hot-spot listening chair (and have no friends) are not such a bad idea. So the WIS element of the S-50 (and S-70, and later models) might well have found a reasonably sympathetic market. The snag was however that the off-centre mirror technique has some inherent acoustic characteristics that made it difficult to engineer speakers that engaged with listeners subjectively in the manner that they were familiar with.
The first of those characteristics was a result of the fact that despite drivers in WIS speakers being orientated to radiate primarily towards the acoustic mirror, significant direct radiation would reach the listener, particularly at low and mid frequencies. The direct radiation would also arrive at the listener first because its path length is shorter than mirror reflected radiation and so would disproportionately colour the subjective tonal balance. Voicing a WIS speaker was consequently a difficult balancing act between nearfield and reverberant field tonal balance. A tonal balance that worked in the nearfield would tend to sound wrong in the reverberant field (or vice/versa), and that was a problem because in a typical retail demonstration situation, first impressions are gained in the reverberant field while more concentrated listening tends to take place closer to the nearfield.
A second characteristic that caused much electro-acoustic head-scratching was that the acoustic load presented to the bass/mid driver by the mirror surface would tend to increase radiation efficiency over a narrow band centred around 700Hz (depending on the size of the mirror and its exact physical relationship to the driver); right in the critical “voice-band”. The resulting 6dB or more peak in the frequency response required equalisation via a passive network and that again brought voicing problems (along with the various distortions and colourations introduced by the necessary large value capacitors and inductors).
While these two issues were inherent to the acoustics of the WIS concept and simply had to be managed as well as possible, a third problem was unfortunately self inflicted in the early days of the development of the S-50. This was the decision to employ a, ‘parasitic tweeter’ equipped full range driver (a parasitic tweeter is alternatively known as a ‘whizzer cone’). Parasitic tweeters, which effectively work by adding a high frequency acoustic/mechanical resonance to a conventional cone bass/mid driver, are a rarity these days and are very much the technology of the 1960s. Their high frequency performance might as well be from decades ago too: typically unrefined, lacking in detail, limited in bandwidth and highly directional. Although the WIS mirror could effectively fix the directional nature of the parasitic tweeter, it could do little for the detail, refinement or limited bandwidth. The lack of a separate tweeter also limited the tonal voicing opportunities of the speaker as a whole. The decision to employ a parasitic tweeter full range driver, baked into the S-50 design very early on, was an incongruous one that I never really understood.
There was one final, and at first sight perhaps not obvious, disadvantage of the WIS concept: physical dimensions. The S-50, for example, was a relatively large stand-mount/bookshelf speaker, however a significant proportion of its physical and visual bulk was created by the acoustic mirror. And the space under the mirror didn’t contribute at all to the air volume loading the bass driver. This meant that the S-50 was limited in terms of sensitivity, low frequency bandwidth and maximum level. It had the bass performance and sensitivity of a very small speaker and the price of much larger one. In comparative retail demonstrations it would often be competing with significantly larger conventional speakers that would offer more extended bass and much higher sensitivity. It was often a one sided contest.
Despite the snags, the S-50 following launch performed surprisingly well in terms of sales (significantly thanks to Alan Boothroyd’s striking industrial design) and even now, pairs in good condition that turn up on eBay don’t hang around for very long. So the prospects for the S-30 seemed very promising – if it could be engineered to reach the manufacturing cost target of less than half that of the S-50. I’ll definitely keep my promise to cover that in Part 3.