Well, I'm browsing over Audioholic's crossover page:

http://www.audioholics.com/education/loudspeaker-basics/filter-crossover-types-for-loudspeakers

And all of this is blowing my mind. I've been running under the assumption that there is no "best" crossover configuration, and it depends on what kind of equipment you are running...but 4th-order L-R looks like it has no drawbacks whatsoever, other than more complex parts.

So would this crossover be optimal for any and all equipment?

I've been running under the assumption that there is no "best" crossover configuration, and it depends on what kind of equipment you are running I think your assumption is pretty good. ;)

I agree, your initial assumption is the right one. For example, you might choose a lower order network if you've got wide-band drivers so that they blend better. The problem with the generalization is that often it depends on something like 'all things being equal' which in practice, invariably aren't.

Passive crossovers are all tied for worst. LR4 is hardly more complex than LR2 in an active circuit, so it depends on the rolloffs of the drivers.

Well I had this rather interesting conversation with the folks that installed the new JBL THX custom array at the famous Empire Leicester square where they have installed the dbx 4800 and a mixture of (Bessle and Linkwitz-Riley) for the loudspeaker array.

Passive crossovers are all tied for worst. Wow, that's quite a statement. Well-done passives are better than poorly done actives and vice versa. Actives aren't without problems. Stock analog crossovers like the pro units are very limited in what they can do. DSP crossovers add an extra AD/DA step and work best if the volume control is after the crossover to keep from dropping digital resolution. Custom analogs are a pain to build for a one-off -- designing and etching the circuit boards etc. And all actives require multiple amps so those with receivers can forget it.

LR4 is hardly more complex than LR2 in an active circuit, so it depends on the rolloffs of the drivers In a passive, an LR4 can often be built with fewer components than an LR2 once you add in the driver rolloffs. LR2, B3, whatever, is the desired acoustical response so you can't just slap on a textbook electrical filter and expect it to work well with either actives or passives.

Behringer DCX2496 may lose resolution, but given that it's 24/96, it makes no audible difference.

Behringer DCX2496 may lose resolution, but given that it's 24/96, it makes no audible difference. Again, opinions vary on the sound of the DCX for whatever reason. People who have tried replacing the analog active crossovers in their Linkwitz Orions with the DCX have generally thought it didn't sound as good. Close but not quite there. There's a large group of DCX users that has grown into a cottage industry to modify the DCX to "improve" the sound of the box including better clock, better output stages, etc. Maybe they're wasting their time and maybe they aren't. ;)

What does Linkwitz say about this?

What does Linkwitz say about this? http://www.linkwitzlab.com/orion-faq.htm#Q14

(snip)...... I doubt that ultimately you will obtain the same dynamic signal range, low distortion and noise floor as with the ASP, especially if you are new to digital signal processing and have to work around the limitations of these commercial units..... (snip)

Not exactly a ringing endorsement, eh?

The reason pro speaker designers make the big bucks :) is because they chose the right cross-over for the system they are designing.
There is a lot more to the Orion cross-over than a LP & HP filter.

The reason pro speaker designers make the big bucks :) is because they chose the right cross-over for the system they are designing.
There is a lot more to the Orion cross-over than a LP & HP filter. Yup, that's for sure! But even when a couple of Orion owners measured the output of SL's ASP and duplicated it in the DCX with a combination of filters, they didn't think the DCX sounded quite as good. I think it's the old 'do no harm' thing. There's less to go wrong with SL's analog implementation. He avoids the AD/DA thing and I'm sure he uses better analog components than Behringer.

Of course, you can make the argument that passive crossovers have even less that can go wrong -- no opamps that can clip, etc. -- as long as you're able to get the transfer function you need out of the passive crossover. In the case of the Orion, a passive crossover would be impractical because of the radically different voltage gain needed for the different drivers. But, if he chose more efficient woofers, it could be done. See JonMarsh's Arvo and Isiris dipole projects.

They need to ABX it.

They need to ABX it. Oh, please. I see you call yourself a UcD lover. If all electronics sound the same, why not just buy some Behringer amps and save yourself the time and extra expense of building UcD amps? ;) I know for a fact that Behringer uses really cheap electrolytic DC blocking caps on the DCX which have easily measured distortion. Put 'em on a scope and what should be a straight line is an oval because of hysteresis -- they're not the same in the + and - directions. And then there's the background hiss that caused a lot of them to be sent back for service. The DCX is a bargain for the price and it's pretty damn good but it's not SOTA. More 'respected' brands such as BSS can cost 10x for similar functionality.

I like UcD for the efficiency, the simplicity and the price.

Not exactly a ringing endorsement, eh?

With up to 24dB of gain to offset woofer Q and dipole roll-off the Orions are somewhat of a special case. People have also had better luck running a multi-channel (VCA or volume control on a chip) attenuator on the digital cross-over outputs thus allowing for a full-scale digital input with better resolution.

Yup, that's for sure! But even when a couple of Orion owners measured the output of SL's ASP and duplicated it in the DCX with a combination of filters, they didn't think the DCX sounded quite as good. I think it's the old 'do no harm' thing. There's less to go wrong with SL's analog implementation. He avoids the AD/DA thing and I'm sure he uses better analog components than Behringer. snip
One channel of the ASP board has 25 blocks (op-amps) it might take a lot of work to get the same transfer function out of the DCX as the ASP.