Active Speakers Don't Sound Better

@erik_squires wrote:

I just wanted to settle a debate that has often raged in A’gon about active vs. passive speakers with my own first hand experience. I’ve recently had the chance to complete a 3-way active center channel to match my 2-way passive speakers.

I can absolutely say that the active nature of the speaker did not make it sound better. Or worse. It has merged perfectly with my side speakers.

Good thing then you're not in a position to settle anything - for anybody else than yourself, that is, and in a very limited, local context. A context btw. I'd urge you to challenge and broaden, and from the quoted part below it would seem there's an opportunity for you to do so:

I can also state that as a builder it was such a positive experience that I may very well be done with making passive speakers from now on.

That's a good outset, at least.

In your response to poster @roxy54:

I’ll argue that you heard ATC speakers that were better active than passive... and that you are extrapolating a universal truth from that which did not hold up for me.

It was his subjective opinion of a particular active speaker, as he clearly pointed out. He wasn't extrapolating any "universal truth."

Let me clarify my original point somewhat. I’m not backpedaling, but clarifying. Based on my experience, the mere conversion of a speaker from passive to active, or design of a speaker as active does not automatically make it better sounding in a meaningful way.

That is, if I were to take an existing speaker design, map the voltage transfer functions from the passive crossover to the active crossover with precision I’d end up with an equally good or bad sounding speaker.

No, it doesn't automatically make it better sounding converting speakers to active, and that's just it: it depends, but the potential is there for sure in a variety of contexts and to a bunch of peoples ears. 

And here’s the issue. The features available in DSP crossovers are vast and tempting. Even with the exact same box and drivers you almost never design the same crossover from one to the other. The economics of part costs and engineering effort needed upend what a good engineer will see as possible and you almost never get to hear a true apples to apples comparisons.

The point is, I can believe you heard two similar ATC speakers. I also believe the reproduction from the active speaker won you over. What I don’t know is all the differences that wend under the hood. Crossover points, slopes, time alignment, driver equalization, etc could all be different and so for me this is no longer a fair comparison.

It's not about fairness, but what the filter choices of active vs. passive offer respectively. You think ATC makes watered down versions of their passive variants? I can assure you they do not, and what their active iteration offers by comparison that the passive ditto doesn't comes down to the nature of active configuration and its inherent possibilities; if active comes out on top here, it's because this specific design route facilitates it. In any case I would say ATC is as good a ground for comparison here than any.

Speaking of fairness: do you think it's fair to base your findings of active speakers/configuration, as a conclusive statement no less, from class D plate amps and a DSP of unknown and questionable overall quality? From the get-go of your main speaker active endeavor it's been clear you've held the staunch decision to limit your context to these components only, which is fair enough, but going on then to make the bold statement that active speakers don't sound better is just seeing the train unrail. 

I will repeat though that using DSP/active configuration gives me a lot more features available to incorporate in my design, and I can be very happy with the results but also not able to say "active is always going to sound better" because I don’t believe that to be true.

Who makes this claim, that active is always going to sound better? 

I'd be happy to see you exploring active configuration down the road in a variety of permutations that isn't only about plate amps with integrated DSP's, and then let's see where you're standing in all this. 

@o_holter wrote:

I have not tried top of the line active speakers. Only mid level and below. But my experience is this: take the ’passive’ speaker from a pair of active speakers. Instead of the output from the amp in the active speaker, give it the output of a good amp (in my case, the amp in my main rig).

Result - to my ears, it usually sounds much better! So I am sad, reconnecting it to the so-so amp in the other speaker.

Comment?

What’re the technicalities of your experience/experiment here? I mean, what do you do filter-wise with the active speakers once you’ve stripped them to a passive state - fit them with passive crossovers between the amp and drivers instead, or use the existing DSP and then combine it with your own external amp + 2 more amp channels if it’s a 2-way design?

Let’s say it’s a 2-way speaker you’d want to keep actively configured, then you would need to add another of your preferred stereo amp (one for each driver section), somehow use the existing DSP and your once active speaker with built-in amps and DSP has now been retrofitted by-passing their internal amps and re-configured actively with amps of your choosing. Is that the way of it?

@erik_squires wrote:

@o_holter

You make exactly the right argument for an external, active crossover. If you want to roll your own amps you can’t do this with a fully active speaker design.

Wrong. A fully active speaker design can be done fitting it all into the speaker as well as having the amps and DSP external to it. Semantically you wouldn’t call the latter an active speaker per se because its amp and DSP components are external to it, but insofar the crossover function via the DSP is done prior to the amplification on signal level (and they’re no passive crossover parts between the amps and drivers) and the respective amp channel are connected directly to their driver sections, it’s a fully actively configured speaker design. Period, end of f*cking story.

Fully actively one can use any external amp one sees fit, as long as you have enough of them to cover the respective driver sections. You would of course need an external DSP as well (again: prior to amplification on signal level, and sans passive XO parts for it to be called fully active), the real challenge being setting filter values - certainly if they’re done on your own. Or, a manufacturer could set the filter values just as they would, basically, a passive crossover - like ATC (electronic, analogue XO), Bryston, Sanders Sound, JBL etc. - and provide them as part of the design one way of the other. The ATC SCM300 Pro ASL’s, and others, have power amps and electronic XO/DSP external to the speakers, but that doesn’t make them any less actively configured.

@o_holter wrote:

my experiments have been very simple. No special filtering or DSP. Mainly with desktop or small monitors. I just unplug the speaker without the amp from the speaker with the amp. This is a speaker-level connection. Then I plug the speaker without amp to my main amp. Often, it sounds better. I make no claim that this happens with more costly speakers. But it makes me think that ’affordable’ active speakers will often sound better with a better amp. The amp really matters. Of course this is the case with passive speakers too. And I think that the quality of the amp is more important than if it is placed in a speaker cabinet or outside it.

Thanks for clarifying. You bring up a great point, and I fully agree; the quality of the amp is more important than whether it, or rather they are placed internally or externally to a speaker. In either case active config. will better harness the potential of a given amp and make for a more efficient use of its power and overall quality, instead of seeing those wattages more or less drained and wasted in passive crossovers, which further leads to a compromised amp to driver interfacing and all that entails.

My experiments can be ’shot down’ since in a sense they are grossly unfair. The amps I’ve used for comparing are much more costly than the speakers. You cant get the sound from the Atma-sphere MA-1 or the Krell FPB600 from inside a compact active speaker. Not that I know of. So my only point, in describing the experiments, is to draw attention to the quality of the amp in the active speakers - I think this is often overlooked.

The important takeaway is the core issue you’re trying to address with your example here. Yes, those amps are very different animals compared to whatever amps are placed inside a cheap active speaker, but you could take much cheaper external amps and still get a basic idea of the importance of their quality here, and the difference they would make.

An argument for active speakers is that the amp and speakers can be more closely matched and tuned to each other. Yet I have not been gripped by this, with the low cost active speakers I have tried. Instead, the big amps just made the speaker sound better. Interestingly, this main effect was the same even with two quite different amps (tube, solid state). My guess is that ’matching’ in affordable active speakers is only approximate, "good enough", so and so many watts drive them to required volume. The amp and the matching are hopefully much better with mid to top level active speakers - I have not tried.

Matching amps to drivers actively has been hotly debated around here (not least involving business developers of active speakers), with my main point being that the most important aspects with active config. are a) getting rid of the passive crossover between the amp and drivers, b) having frequency band independently functioning amp-to-driver sections, c) freely seeking out the external quality amps and additional gear one prefers, and d) having basically a carte blanche repertoire of speakers, irrespective of size or principle to go by - if one so chooses.

Impedance matching, current or voltage drive, tailoring damping factor, power matching, etc. can have their degrees of influence, but the problem is working with compromised amp sections (as well as DSP/DAC’s) within a tight budget that have to be mounted inside speakers, and so what’s attempted to be gained initially is hampered by overall component quality and design/construction eventually. Not to mention that active speakers are oftentimes physically hampered size-wise to cater to interior decoration demands and the misplaced, general notion that active speakers have to be plug-and-play, convenient solutions that fit nicely on the shelves and pleases the spouse - when active as a system could be much more than that and is really only limited by the one implementing it.

Listening to a pair of outboard actively configured ATC SCM300ASL Pro’s - which represent a more old school, analogue-only, meat and potatoes, no frills, excellent component quality and class A/B topology approach - is being confronted with a pair of world class speakers that to my ears puts to shame many high-end, passively configured speakers of higher cost, and that’s not even including the astronomically priced amps that are typically needed with such heavy-load speakers to bring them to life.

Brad --

Ad 1. Definitely agree, although there are different ways to approach phase implementation. Usually we have a limited number of phase bands to go by that’re set and fixed around each driver section, meaning only one chosen value per band. Linear phase filters (FIR-filtration in the DSP domain) on the other hand have over 60,000 phase points over the frequency region and offer some unique possibilities here, although it’s also more processing heavy which in turn can have its drawbacks with an audible ghosting effect.

In any case phase correction has degrees of importance, not least with larger, horn-loaded speakers, and it’s one of the reasons why active config. pairs so well with this speaker segment. Another reason is being given the opportunity to use very steep filter slopes to more effectively and sensibly use horns within their bandwidth range, and thereby avoid off-band irregularities. This is also why many haven’t heard what horn-based speakers can really do when properly implemented actively. Passive filters simply fall short here.

Ad 2. I can also set gain in 0.25dB increments with my Xilica DSP, of course with each driver section, which is indeed audible and to the point even that we’d prefer having 0.1dB gain steps.

Ad 3. Precision with active filtering is a big plus, yes, and also comes in handy with notch placements, not to mention their q-value and gain factor. Another advantage with horn-based speakers.

Ad 4. The limiting factor here is that imposed by the drivers and the passive crossovers themselves; the latter when thermally challenged will lead to fluctuations in filter values, and this further exacerbates this issue of (the nature of) the lower precision found in passive filters. Active filters on line level will remain rock solid and totally impervious no matter the load. Where drivers are concerned the use of limiters aren’t needed when power handling is prodigious, aided not least by higher efficiency.

Ad 5. I’d question the significance of speaker cable runs no more than ~10 feet per channel with proper gauge, certainly as the only medium between the amps and drivers. Crossover coils, another matter, not least in conjunction with steep phase angles created by passive crossovers and its components on the whole. The purer impedance load with active is a vital factor in its advantage.

Ad 6. You can hardly over power the system. If a high power solution sounds great, it sounds great. I’ve used 30W class A power and 600W class A/B ditto with 111dB horns, and the latter, high power solution didn’t fall short - on the contrary. The "right size" amp, from my chair, is really about having (more than) enough power, and where plentiful - depending on the design - isn’t a disadvantage. With that out of the way it’s really about finding the right sounding amp, and using the same topology/design top to bottom is paramount to my ears - even into the subs region. Outboard active gives you more opportunities here.

Another, very important aspect with active configuration is having amp-driver independent bands, as well as using the different amps in limited frequency ranges. In a 3-way system like my own the top band, 600W amp is only fed with a ~620Hz on up signal (with 6th order filters); it cruises along with its direct-connected 111dB horn/compression driver section, and the distortion is at an absolute minimum - even at blasting levels. On the end of the scale a similar sub amp, also 600W, can blast along as much as it wants to (which likely never amounts to more than 10-20 watts, at most), and it won’t have any effect whatsoever on the 2 other amps with their driver sections used above. In a passive setup the typically single amp covers the entire frequency and driver range; what it does down low affects everything above.

Ad 7. Agree on the cosmetics part, but practically speaking many if not most active speakers suffer from overall amp quality compared to outboard solutions. Not everyone is at ATC level here. Yes, used actively the amps can more effectively reach their fuller performance envelope, but that doesn’t negate the impact of absolute amp quality.

celtic66 wrote:

Shouldn't some of you just get a room?  Reminds me of pointless religious diatribe. While you guys were engaging in pseudo-intellectual combat, I was enjoying my system which at this point just sings and celebrating enormous talent.

Why don't you chime in on the subject, or be more specific about what it is you're trying to address? 

I'd say there are many threads to which your criticism could be leveled, perhaps more rightfully so, and do you engage in a similar fashion with them too? Why this thread? 

celtic66 wrote:

Rinse and repeat.  Cheers

It's a discussion, man, with disagreements. Enjoy your system, like we do ours. 

lonemountain wrote:

Speaker systems (with a passive crossover) can sound different when they get hot.

And that includes the passive crossover parts as well; when they get hot, filter values can fluctuate.