subwoofers and panels don't mix

I may sound like a broken record here: Room analysis, PEq, felexible x-over. I've used a Velodyne SMS-1 sub controller to seamlessly integrate a pair of Velodyne SPLR 800 subs with Maggie SMGs. It took a lot of time and a lot of tweaking, but I can't hear the crossover at all. Let me emphasize, I've always loved the IDEA of hybrid speakers, but, to my ear, even the best (IMHO - Eminent Tech) have had issues with the planar/dynamic x-over.

Using the SMS room analyzer, you can easily see how dipoles differ from forward firing speakers throughout the bass. The nature of the response irregularities in panels makes integration a real challenge - but it can be done.

IMHO, the key to good sub/speaker is smooth on-axis FR around the x-over point, so I follow the following procedure:

1) Use the room analyzer to find a smoothish chunk of bass and start with a frequency in the center of this area as your x-over point. Be sure that this "smooth area" is reasonably balanced w/the rest of the spectrum. A smooth plateau at + 15db doesn't work very well. You may have to move the sub(s) around the room a bit, 'til you find a good result.

2) Flip polarity (this will usually kill the primary suckout quite effectively) to see whether + or - works better.

3) EQ around the x-over point for further smoothing.

4) EQ below the x-over for best balance of smooth vs extended deep bass response.

IME, this procedure will get excellent results from subwoofers with a dipole, unidirectional, or omni speakers (I've tried all three).

Good Luck

Marty

BTW - I really doubt that your choice of subwoofer model is critical - looking "fast vs. slow" isn't likely to help. IME, integration trumps sub performance. I chose the Velos because of size. There are several other models at/near their price that return far better specs (group delay for speed and distortion for clean output) and will very likely outperform them. Yet, I still get great results despite the mediocre subs I use.

Two more group delay observations:

First is inconsistency in test results:

If you cross check AVTalk and Hometheatershack you will see differences in the test results of the same model subwoofer (these may look smallish at a glace) which prove greater than the delta between differing makes & models. In fact, HTshack tested 2 samples of the SVS 12NSD and got significantly different GD results.

Second is Shadorne's illustration of GD's impact at 20hz:

While the GD test results really diverge at this low frequency, they tend to converge significantly as pitch rises. I wonder if the lack of really low frequency program content tends to obscure the audibility of a sub's GD capability. At least, the program material I used in my listening tests.

Marty

Lightminer,

Are you using your system for HT as well as music? Are you a pipe organ guy? If not, why worry about ultimate LF below 25hz. In my collection of >5,000 records and cds, I can only think of a few that go below 30hz. I only mention this because there are tradeoffs in choosing subs and the first one I'll give away is extension.

BTW, I can't imagine that you'll miss the difference in bass performance between EQ'd subs w/ your 3.6s and full range 3.6s. Both will "go there", but they'll sound quite different in terms of "impact". I personally have little issue with timbre (my subs are actively crossed at 74hz). I assume that this is because timbre becomes less meaningful as you descend in frequency. At 25hz, you feel it more than you hear it. I may be less sensitive to timbre issues higher in hz near the x-over than some other folks, but I don't hear the timbre change im my set up; even in the lower registers of electric bass or piano where fundamentals may fall in this critical area. However, this is a classic YMMV.

AVTalk.com and Hometheatershack.com are the two subwoofer distortion & group delay databases I know of (Thanks Shadorne and Bob Reynolds). None of the data you seek is on either site, but there are several REL model test results vailable. General rule for REL = great group delay #s, ugly distortion #s. SVS subs seem to offer a nearly unmatched balance of distortion & group delay performance (if you value both equally), and they do so at a comparatively low price.

Marty

Shadorne,

I understand your point entirely. Now mine:

I recently heard the JL 113 Fathom (low group delay) set up with the
Maggie 1.X (don't recall the current designation) at a local dealer. After an
hour, I had the following observations:

1) The bass sounded no tighter, quicker, or less smeared than my (high
group delay) SPLRs.

2) Mid-bass and mid-range didn't sound more articulate.

3) Discontinuity between planars and subs was evident to me (I attribute this
to careless set-up).

4) The system was more dynamic sounding than mine.

5) This was a different room/different source & amplification/ different set-
up regimen than I use at home, therefore an imperfect test.

However, I concluded that the lower group delay wasn't OBVIOUSLY audible
to me in this set-up, as I expected it might be. Neither was the improved
distortion performance (The JLs produce far better distortion specs than my
SPLRs). My conclusion is that careful set-up (probably including EQ) likely
trumps specific sub performance. IOW, I suspect that the specs we use are
measuring stuff I can't hear (or, at least, hear very well). YMMV.

Marty

BTW, once I figure out which subs make the most sense (I wish I had data on
more models) , I'll probably replace the SPLRs (if only on principle alone). I
want lower GD in my system, even if the effect isn't dramatic. Same for lower
distortion. To be clear, I'm not saying better subs don't sound better. I'm just
saying that IME better set-up (including EQ if that -as is likely- proves
necessary) is more obviously beneficial to my ear.

Further, for the OP, I suspect that great subs won't fix his problem. Good
subs, properly set up, just might.

Finally, thanks again to you, Bob and Drew for pointing the way toward betetr
bass performance in my system. Your advice has been invaluable.

Agreed.

I'm just finding it difficult to reconcile the published numbers I see with what I actually hear. I'll eventually order a pair of SVS 13" Ultras and use them in sealed mode - pretty low distortion from a pair and GD stays below 30 at 20hz. The cost of a pair is still below that of one JL113, so my sense is that this is a good value play. At least as far as upgraded subs can represent good value for me. I guess I'll have a better idea when I get 'em in place.

Marty

Lightminer,

For your goals, I suspect that a pair of SVS Ultras in sealed (or maybe 16hz tuned) mode would be hard to beat. Remember that the additional sub provides 3db (more like 5db in my room) output, so read the distortion/clean output graphs at lower output to maintain "apples to apples" comparisons.

Also, bear in mind that the output at, say, 50hz from your Maggies will NOT sound the same as the 50hz output from a sub. I can't say for sure which you'll prefer, but I'm confident that you'll hear a difference. If you EQ the subs, I'm nearly 100% sure you'll hear quite a significant difference.

Good Luck

Marty

Somewhat further OT here:

NHT's lifestyle "woofer" system is being closed out at Audio Advisor (I get their mailers) @ $300 per. This isn't really a "subwoofer", but it puts two 10" drivers in a very slim sealed box (app 6" deep) with IIRC a 200 watt amplifier. Max extension is something like 38hz @ unspecified spl/unspecified THD.

Observation: Since the narrow cabinet puts the front of the woofer 6" from the wall, cancellation effects should be minimal.

Observation 2: Max excursion of the drivers is unspecified, but can (presumably) be limited by employing multiple units in the room.

Question: If you scatter 4 - 6 of these units in your room ($1200 to $1800) how do you suppose performance would compare with conventional subs in the price range?

I'm not in the market, but the question crossed my mind and I wonder what you guys think.

Marty