Here come some long-delayed replies to several points raised by Prof on 12/9 in his response to my original post to this thread. First, a correction:
I wrote on 12/9:
The [Thiel] SI-1 uses first-order crossovers; the [JL Audio] CR-1 uses 4th-order crossovers.
This is what I was told during my research before bidding on the SI-1, but it’s not true, as seen on p 10 of this document:
https://www.manualslib.com/manual/328036/Thiel-Smartsub-Ss2.html. (For more complete information, read the whole Integration section, starting on p 8.)
Thiel argues that standard LP and HP filters can not produce linear frequency response in the net acoustic output, even with some extreme tweaks. See pp 8-9 for examples in which optimal independent selections of standard LP and HP filters combined with optimized level and phase adjustments fail to deliver linear acoustic output. The standard LP and HP filters used in most active crossovers, on the other hand, are designed to deliver matched (mirror-image) voltage profiles, not matched acoustic output. For example, the JL Audio CR-1 uses 4th-order Linkwitz-Reilly filters that guarantee matched voltage profiles.
Thiel’s goal for the SI-1 in crossover mode (augment mode is another story) was to match the total response of the main speakers to the total response produced by a generic sealed sub driven by an ideal (according to Jim Thiel) 4th-order LP filter with Q = 0.5 damping. Evidently, “total response” refers to the net acoustic output determined by the voltage profiles of the HP and LP crossover filters in combination with the distinct acoustic properties of the sub(s) and main speakers. To accomplish this goal, he uses a nonstandard, highly customized HP voltage profile to make the total response of the main speakers match the shape of the target sub response (see Fig. 11 on p 10 in which the red curve shows the customized HP voltage profile for reflex mains that are typical of the Thiel line). Once this is accomplished, “straightforward” phase alignment and level adjustments should complete the integration (this is IMO; it’s not stated explicitly by Thiel).
BTW, the first-order crossovers in Thiel’s main speaker line are designed for first-order acoustic response, not first-order voltage profiles. From Thiel sales literature for the CS3.7: “The Crossover is a true first order *acoustic type* that provides the utmost in spatial and depth imaging performance as well as overall realism. This is the only type of crossover that provides complete accuracy of amplitude, phase, time, and energy, and therefore the only type that does not distort the musical waveform.” The design of first-order voltage crossovers is relatively simple; the design of first-order acoustic type crossovers is significantly more difficult and expensive. Presumably, this is one reason why very few speaker designers undertake phase and time-coherent designs.
I wrote:
Note: The 4th-order crossovers in the JLAs are not compatible with the 1st-order crossovers in Thiel speakers.
Prof’s challenge to this comment is well taken. First, the words “not compatible” are too strong. I thought I changed incompatible to “inconsistent” before posting, but clearly I did not. Better yet, I should have written “inconsistent with Thiel’s emphasis on phase and time coherence in speaker design (through the use of first-order acoustic type crossovers). This is accurate, but one might still infer from my comment that the Thiel SI-1 can deliver phase both and time coherence. I did think this was true at the time of my post, based on the misinformation about the SI-1 using first-order filters and my assumption that Jim Thiel would only bother with first-order filters in pursuit of his signature phase and time coherent designs. However, according to my correction above, it appears the main advantage of the SI-1 is more accurate frequency response relative to active crossovers that use voltage-type matching. Is this distinction important in practice? It’s not clear. Jim Thiel presumably thought so, but Soundoctor makes some good arguments that suggest owners of active crossovers with voltage-type matching shouldn’t lose any sleep over it; see here
http://www.soundoctor.com/whitepapers/subs.htm.
This still leaves a few more questions/comments from Prof that I should answer. I’ll address these, as best I can, in another post; this one is already pretty long.
