Top notch speakers with their own sub

Drew,

Nulls can be fixed with EQ - with some limitations. I have Eq'd to +12db with a pair of Velodyne SPLR 8 subs (1000wpc internal amps IIRC) and achieved excellent results. To muster +12db, I doubled down on the SPLR's internal EQ and added extra maximum EQ in my SMS sub controller. This approach IS DEFINITELY NOT recommended by Velodyne, but I did it anyway. Make no mistake, it worked....in my application.

I suspect that the real reason this approach can work well in music only systems is because my max EQ occured below 35hz where the pair of 8" drivers in the subs conked out. In actual use (playing music) there's little sustained signal there. OTOH, the EQ'd subs performed like a champ running test sweeps! I won't disagree with you that this approach lacks "elegance" and, in fact, could prove risky if amplification demands exceed the power available.

The real point I was making Re: DEQX powered speakers, however, is that straight EQ (including bass nulls) is the theory behind the design. I agree that you have correctly pointed out one significant limitation of the "brute force" approach.

Marty

Marty,

It may work on certain soft music but an extra 12 db boost below 35 Hz is a gargantuan and scary extra amount for a poor subwoofer to put out. If you do this then I'd be extremely careful not to blow them up through excessive excursion or from burning up the voice coils.

Most 8 inch subwoofers are unlikely to play cleanly at 20 Hz or much below 35 Hz with anything more than 90 sb SPL output (which is barely audible). Below 35 Hz is really really tough territory for ANY subwoofer.

Another issure is ringing.....filters that boost will ring like a bell - as you exceed 6 db boost with a narrow high Q filter then the ringing will become progressively audible. Therefore it is best to use notch filters (sharp cuts) rather than any boost.

Shadorne,

I can't recommend the approach (you state several good reasons not to), but I will say it worked really well for me. I filtered below 25hz, so the max boost was applied only to the narrow range between 25hz and 35hz. Test sweeps were run at app 95db and were quite loud - they exhibited no ill behavior that I could detect. Music reproduction was always pretty damn impressive to my ear, though I can't swear that there was ever much signal in the highly boosted frequency range under discussion here.

Marty

Equalizing peaks and nulls becomes more practical with a distributed multisub system. Before equalization the peaks are less peaky, the nulls are less nully, and there is less spatial variation (meaning that the response doesn't change as much from one location within the room to another). That being said, I'd still be hesitant about putting a lot of boost into a null unless you had plenty of headroom available - which apparently was the case in Marty's situation.

Hey Duke (or anyone else who cares to comment) - Is a flat response always necessarily an ideal target? I can understand addressing major dropouts at the sweet spot, certainly. When we go attend a concert the folks on the mixing board aren't getting a flat response from the room, are they? So if we're just talking about bringing the intent of the sound engineer / producer's work into our listening room with the least coloration/distortion, I expect that would lean towards the flat response, but could that also, for any reason, tend to sound any more or less engaging than a response that is more typical of most rooms (ie peaky)? Just curious what your take on that might be. Certainly if the Planetarium system speaks to this end it is a very convincing voice for that approach. I just wondered if that was the ultimate goal, or is that only a shade of something broader in what you are trying to achieve? What occurred to me as extraordinary about what you've achieved, at least in the room I heard, was a very natural and seamless integration of the low-end that did not call attention to itself as such, it was simply there at one end of the spectrum, yet was unmistakable as such. What I've objected to in some other approaches is that you become very aware of the low end to where it becomes distracting. I don't know whether this is due to overemphasis, room nodes, or some other imbalance. So is a flat response always the goal? Or?