Prof wrote:
If using a first order crossover with the sub were about maintaining time/phase coherence, I’m not going to have that anyway. The subs will have some delay, and will be placed behind the speakers, hence time delay. I could ideally get them phase coherent with the Thiels with the phase controls, but they’d still be at least a cycle behind in terms of time coherence (unless I go whole hog and want to digitize the entire signal to allow the mains to be delayed for time coherence with the subs - which I’m not yet willing to do.
Prof is right. Even if he places his subs at the same distance from the listening position as his main speakers, the group delay of his JLA subs means that the best he can achieve is phase, but not time, coherence. This is true for both the Thiel SI-1 and the JL Audio CR-1. I explained how I came to post fake news about the SI-1 on this point in my previous post. Apologies for that.
Prof wrote: And, btw, isn't that what the Thiel subwoofer integrator does? Digitize the entire signal to mains and subs?).
Nope. See p 3 of the SmartSub Owner Information manual,
https://www.manualslib.com/manual/328036/Thiel-Smartsub-Ss2.html:
All signal processing [in the SI-1 Integrator] is done with analog circuitry. Digital circuitry is used for user interface, calculation and circuit control functions.
How does the SI-1 accomplish level adjustment and phase alignment, accounting for both SmartSub group delay and variable subwoofer placement, and why do SmartSubs lack a phase control? ... We know from the documentation that the SI-1 includes continuous phase adjustment in its bag of tricks for “Smart” integration. Since Thiel surely knew the group delay of its own SmartSubs, computing the phase adjustment to compensate for group delay at any crossover frequency is simple enough. On the other hand, I suspect that the SmartSub circuitry does not account for variable subwoofer placement. Instead, this issue is probably addressed by the instructions on p 9 of the SmartSub Technical Information manual,
http://audio.manualsonline.com/manuals/mfg/thiel_audio_products/sw1_ss2_ss3_ss4.html:
The subwoofer(s) preferably should be approximately the same distance from the main listening area as the main speakers.
To the extent that this instruction is followed, the SI-1’s phase adjustment for SmartSub group delay is sufficient for total phase alignment. The user must enter the main speaker sensitivity into the SI-1, and Thiel knows the SmartSubs’ sensitivity. With this information, automatic level matching is straightforward.
OK, then, how can we implement the “straightforward phase alignment and level adjustments” mentioned in my previous post using the SI-1 with non-Thiel subs or if the subs are not placed at the same distance from the listener as the main speakers? Although in this case phase alignment and level adjustment are no longer smart, they can be implemented simultaneously using Method B, as described toward the bottom of p 4 of the instructions that accompany the SoundDoctor test CD:
http://www.soundoctor.com/testcd/Soundoctor_Test_CD_v2-7-2.pdf. In this approach we use the level, polarity and phase controls on the subwoofer, just as we would with a more conventional crossover with voltage-type filters, such as the CR-1. Note, however, that the phase adjustment on non-Thiel subs will be the amount needed beyond the default phase adjustment that the SI-1 computes for SmartSub group delay. Fortunately, we don’t need to know the SmartSub group delay to implement Method B, because I don’t know the value of the SmartSub delay.
My initial attempts at implementing Method B indicate that the group delay for SmartSubs and my JL Audio f112s are surprisingly close (within 5 degrees at 80 Hz). I should do some more careful work to confirm this. If anyone knows the group delay spec for SmartSubs, I’d be very interested to know
In the end, what did I get with the SI-1? Definitely a good price (thanks for not bidding you guys). Definitely a high-quality crossover with a nice remote. Possibly, improved frequency response with the SI-1’s acoustic type filters.*
BTW, I’m very pleased with the results I’m getting so far. I’m definitely hearing the improvements promised for switching to two subs, for using an active crossover, and for addressing phase alignment.
* I think my use of the DARO (automatic DSP room correction) that comes with the JL Audio Fathom line and the fact that my whole system, including acoustic treatments, is designed to optimize the sound at the sweet spot (vs. throughout the room) undercuts to some extent the Sound Doctor’s argument against worrying about frequency response. Jim Thiel believed the difference between acoustic-type and voltage-type filters matters in the cross-overs within our speakers, so why not also for sub integration? … I have no plans to test this conjecture by installing a more conventional crossover for comparison. So with no evidence to the contrary, I’ll continue to tell myself that the frequency response matters, too, just to keep myself happy.
But, as I said in my previous post, no one using a CR-1 or other crossover should lose any sleep over this IMO.