Tweaking crossovers by changing values really requires you to understand what each component is doing and why. First order crossovers are the easiest. As poles increase, it becomes increasingly more complicated. Not only do you have shape of slope to worry about, but you have to contend with phase, baffle and relative positioning of drivers. Computer modelling programs are a big help in this area.

I want to address your comment about established company spending a lot of money on XO design and whether or not it's really that easy. The short answer is yes, it is that easy. Considering that there are many in the DIY community armed with nothing more than a laptop, microphone, a corner in their basement, and several hundred dollars of software, that are able to design and build perfectly flat speakers, it's not really that difficult if you really want to do it. The big difference is that established companies have to design these things so that can be built consistently under price points.

Gar.

I'd be willing to bet that if you listened to several generations of Wilson Audio WAMMs side-by-side the sound character would be pretty much the same.

As far as pros having an advantage over DIY? I'm willing to bet a dollar to a dime that their advantage (excluding obvious charlatans like Audionote selling 2-way $28,000 speakers that they proudly declare sound like a drumhead when tapped on)is that the pros are more likely to be clear-eyed (eared?)about their results, less self-delusional because they're not working in isolation, open-loop, no feedback.

99 and 44/100% of any loudsepaker is the driver and very, very few manufacturers of dynamic loudspeakers build or even specify drivers; some big Chinese company produces a huge range of variations and they pick one. If you turn to the Japanese companies (TAD, GOTO, ALE) I highly doubt (but don't know for a fact) that they wouldn't consider any outside input beyond "thank you for the honor of allowing me to purchase your products".

So you're left with a company buying drivers and building a box to hold them. How hard is that really? The guy who designed and built the driver had an ideal enclosure in mind when he started. Ask him. If you don't want to ask him as T-S - they know to a very high degree of certainty.

And the crossover? I'm convinced that all the "mystique" about crossovers is largely manufactured - that even a textbook crossover will work just fine with well-behaved quality drivers. It's undoubtedly a challenge to knock off the peaks and dips in the response of some "Made in China" POS but that wasn't your point. And even if you engineer a brilliant response shaping network, how meaningful is it when applied to a line of drivers with poor quality control and large sample-to-sample variations?

As far as cost goes, look at the prices of Westlakes or Wilsons or those of similar stature - a speaker with perhaps $5K of drivers and materials at retail sells for $30,000 - there's a hell of a lot of $ for the DIYer to play with.

And what I think most important is that the DIYer doesn't need an anechoic chamber - he's building a custom speaker for a custom space, his livingroom - all that really matters is the in-room response he gets, not how his speakers measure in some absolute sense.

And another point - people who build and sell "statement" speakers with passive crossovers are just full of shit - there's no justification for a $30K pair of speakers with a passive crossover these days, especially as cheap as DSP has become. But again, it's part of the manufactured mistique - you use passive crossovers and your customer is much more likely to waste thousands on the speaker wires and amp-trading that your retailer relies on between big speaker sales.

And yet ANOTHER point - most loudspeakers above a certain quality level are a distinction without a difference, a matter of personal taste and peference (what the hell is a $50,000 Stereophile Class A loudspeaker with LIMITED FREQUENCY EXTENSION anyway???)

Much if not most of the consumer loudspeaker market is an exercise in illusion and intimidation - mass hysteria or a sort, assuming you're vulnerable to it.

You wrote :

"And another point - people who build and sell "statement" speakers with passive crossovers are just full of shit - there's no justification for a $30K pair of speakers with a passive crossover these days, especially as cheap as DSP has become."

BAM! I feel like somebody just hit me square in the forehead with a 2x4 to get my attention. I had never considered that thought before but it makes absolute perfect sense. Spend $30K and you get to listen to your high priced amps and preamps filtered by a bunch of Solen caps, sand cast resistors, and coils of magnet wire. But they are safely hidden in a beautiful styled exotic cabinet that makes a statement to anybody coming into your home. That was probably one of the most enlightening statements I have read on a forum in a long time.

Excellent,

Martin

First, Benson was probably writing as an academic for an academic audience which really means that the elegance and complexity of the derivations is elevated. People who write in these circles are all extremely talented and the style tends to be as important as the content. You see similar styles in articles in most professional technical society Journals. It satisfies a certain community and can unintentionally exclude the average person trying to learn the sunject. The author is demonstrating his or her skill and completeness in mathematical derivation and documentation and this is required to remain and florish in academia. Anybody pursuing a technical Phd is probably well schooled in this style of writing.

As an aside, early in my career I wrote an internal company paper on bending of bolted joints and closures for nuclear pressure vessels. I tried to explain it so any other engineer could understand and use the simple equations/methods I had derived. A month or two after the paper was distributed, I got a copy of an article put together by a Phd based on my work. Everything was a variable including the shape of the bolts. He had derived the equations to include square bolts! Totally impractical and crazy.

I am not trying to imply that Benson's work is either impracticle or crazy, it is really a very nice book but aimed at a limited audience. I have the same book on my shelf and have read it but not used it very much.

The second thing to remember is that people trained in this type of work can read Benson's text and not struggle too much. Somebody with an electrial engineering, mechanical engineering (like me), or maybe a physics education has seen this type of math modeling of systems in school. I make my living analyzing structures and hardware mathematically and have worked on several different products using the same set of skills. When you look at a closed or ported box system, you are really working with the simplest of these type od math models. One (the driver) or two (the driver and the air in the port) mass systems with a couple of springs (driver suspension and air volume in the box) are as simple as it gets in math modeling. These math models can be derived and solved by hand, then they can be easily prgramed to make nice response plots (think Boxplot, WinISD and my MathCad worksheets). This is why they are "not rocket science". Now you can add more and more complexity to the models and the predictions that will make the solution more accurate and at the same time more complex. But basically, closed and ported speaker systems are fairly simple systems compared to the work most analytical engineers/physicists do at their day jobs.

Hope that helps,

Martin