Full-Range, Multi-Way, etc.: offshoot thread response

[WmAx]

This thread is an offshoot concerning full range and multi-way systems, as well as other various loudspeaker issues from the last couple of pages of the following thread:

http://forums.audioholics.com/forums/showthread.php?t=18072

-Chris

 

[WmAx]

The point is that I was surprised by just how much time a (passive) crossover designer needs to spend working his way around issues, that are not really to do with the goal of creating a suitable filter, but are simply there to side-step strange driver behaviour.

There's an interesting tutorial by the author of lspCAD on how 'easy' it is to do a digital crossover design using a Behringer DCX2496. He then does a passive design for the same system... it's enough to put you off passives for life .

Indeed, a DSP crossover is much easier to use for a crossover design.

Considering the price of some of the higher end fullrange driver systems (comparable to a very good bass/mid and treble pair) I certainly accept that the compromise of fullrange (with no crossover) or multiple drivers (with crossover) favours the latter.

This statement is too ambigous, as in it is not certain which compromises to which you refer.

Am I on the right lines though, that a single point source would theoretically be better - assuming you could match the frequency response, distortion etc. of a comparable two/three way configuration?

Yes, a point source is theoretically better(from a power response point of view) if it can match the other characteristics that are critical. However, it is rare that one can match the other critical characteristics.

-Chris

 

[WmAx]

As for passive xovers, they can in fact exibit some 'nasty' behavior in certain circumstances. Active xovers can be superior if done properly. Exceptional passive xovers are not inexpensive to produce either.

Yes, for a superb example of an extremely complex crossover in order to allow superb signal performance, refer to Dunlavy Labs speaker systems. The passive crossovers used in these were very complex/expensive.

I'm not sure comparing a 'typical' full range to a 'proper' multi-way is really apples to apples. How about comparing a typical full range to a typical multi-driver or a 'proper' full range to a 'proper' multi-driver? While multi-driver systems do have some advantages in terms of feeding drivers the ranges that they are most linear and being able to contour the frequency response as part of the deal - there is always a tradeoff.

A typical single driver full range hi-fi speaker, per my definition, would be defined as a Fostex or Lowther. Very typical. Poor performance in all objective regards as compared to even a low cost well designed multi-way system. Some people may like the resonant, non-linear, and highly colored sound of a typical single driver full range system; but it's technically poor.

A properly designed full range system will always be in phase. A properly designed full range system can exhibit much better imaging (not always but in theory it can all else being equal).

Properly designed full range; that is a rare thing. For the ideal full range driver, which can exist(well, mostly full range), very expensive special application drivers need to be used in order to achieve good objective performance. Manger drivers are an example of a specialized full range driver that can accomplish this feat -- at least from about 100Hz-2000Hz. Some electrostatic panels can perform very good, but not usually as well as very well designed dynamic drivers. No single driver full range system can truly be full range from the low bass frequencies without making severe compromises in low end dynamic range and bandwidth. Having the best phase response is overblown, as it's not shown in perceptual tests to be important as compared to the parameters that must be compromised to allow this improved phase response.

Also, there is a limit to what you can do with multi-driver systems. Is a 3 way better than a 2 way? 4 better than 3? 5 better than 4? etc.? At what point are you monkeying around SO much that you've caused issues in the time domain, increased price drastically due to necessary xover complexity, etc? K.I.S.S. principle?

The number of drivers is dictated by the engineering objectives and the drivers themselves. There is no 'best', since one has not defined what is supposed to be best, but only 'best for the intended target objectives'.

Why do 'monitors' have the ability to 'disappear' into the room? Part of it is small cabinet size. I firmly believe that part of it is also significantly simpler xovers.

The narrow cabinet helping the system appears to be a myth. There are too many variables that must be isolated; but are not considered when people jump to this conclusion. I don't know of any reason why a smaller cabinet width in itself should help anything except for the delay time of the diffractive edge artifacts. Ideally the cabinet should be using a large radius on the edges, or use of a material that will absorb high frequencies, or use a tweeter mounted in a very small enclosure on top of the cabinet, to prevent most of the diffractive effects in the high frequency band from occuring in the first place. More likely the imaging associated with small monitors is due, is that it is more difficult for a small monitor to screw up midrange to treble power response linearity as the midrange transitions to the tweeter, because the midrange is a small diameter, thus has inherantly superior polar response at higher frequencies as compared to a larger midrange. This will lend itself to a more coherant stability throughout the crossover transition off axis. Consider that usually the speakers are optimized for 0 degree measured performance, but are listened to at a 15-30 degree angle, the power response in this region will be very important in this case.

-Chris

 

[bpape]

Many of the things you've put forth are your opinion. Smaller cabinets IMO absolutey contribute to the disappearing act. Phase coherence not being important IMO is just wrong. If that's the case then why are certain xover types eschewed purely because of their phase issues by well respected designers? Purely hype? I think not but you're welcome to your opinion.

Have you ever heard a Lowther used in a bi-amped system with active xovers to a good woofer? Tri-amped with a 3 way ACTIVE xover with woofer and a padded Esotar tweeter? It is a very compelling sound. Yes, it has it's limitations but it is capable of some of the most realistic reproduction I've heard outside of the things I described earlier.


While I agree that doing a proper full range is very difficult, we were after all discussing theory here. In theory, a single driver single point source is the optimal goal (short of directly driving air with no 'driver' - see Hill Plasma tweeter). To the item on the Dunlaveys, that's exactly my point. How many 'inexpensive well designed' speakers are out there that have really accurate, well done passive xovers? Not many. Dunlaveys hardly qualify for inexpensive - sorry.

As for Rob's quesiton about hearing full reproduction from a small monitor - I'll point to the Totem Mani-II as example that this can in fact be done. A simple 2 way with a modified Isobarik dual woofer configuration that images like mad, has excellent full range performance (-3db @ 32Hz I believe), etc. Now, it's tough to drive, but that's a whole 'nother issue

 

[Rip Van Woofer]

Good idea to start another thread for this. Very mature and gentlemanly behavior. Gold star on your foreheads!

Now, as for phase coherence: It's not quite an "IMO" question. The subject has been very scientifically investigated by some of the sharpest brains in audio from both the practitioner and academic domains: Vanderkooy, Lipschitz, Toole, Linkwitz, and others. And the upshot is that any effects are at best extremely subtle to nonexistent in typical music program material, and only reliably audible with certain very specific test tones on some types of speakers. Indeed, quite shockingly large phase errors (up to 1000 degrees out of phase!) are difficult or impossible to hear reliably in controlled tests using music instead of test tones -- <b>IF</b> they don't also cause frequency respose aberrations as well (which they often do). Then you're actually hearing the latter.

Getting good time domain coherence between drivers (often confused with, but not the same thing as phase coherence) can be a good thing, though, because otherwise there can be comb-filtering (I think) in the crossover region which will mess up the frequency response. But be very clear here: just as with phase errors it's the frequency response aberrations we hear, not "smear" or some such due to time displacement. And its easier to get time alignment with active crossovers either with digital delay or analog allpass filters.

So why do some reputable designers still care about phase coherence? Partly because even engineers and scientists are not immune to honest, deeply held bias. Partly because the results of the above researches allow for some uncertainty (which, by the way, is the nature of science). And maybe in some (not all!) cases because there's money to be made by catering to some audiophile prejudices and to use coherence as a way to differentiate your speaker in the marketplace. "Branding", if you will. Remember, audio is a business not just an academic or engineering discipline.

As for the theoretical virtues of a point source. Yes, it is "perfect". Unless you happen to be partial to a theoretical line source, which is also "perfect". All of which is beside the point because there is no such thing as a perfect realization of a theoretical ideal.

To summarize: Phase coherence and time alignment are underlying mechanisms that the conscientious speaker designer will consider (among many others) as means to insure fewer frequency response errors, but they are only reliably audible with the vast majority of music and cinema programs to the extent that they do or do not cause freqency response errors. They have not been shown to be the cause of peculiar, characteristic defects in and of themselves.

 

[WmAx]

Many of the things you've put forth are your opinion.

I qualify every statement within context clearly. The opinions I offer are clearly presented as such, but in this discussion, I have primarily questioned assertions, and provided replies based on up to date perceptual research in the field.

Smaller cabinets IMO absolutey contribute to the disappearing act.

Who has demonstrated this in a credible study? It's just speculation.

Phase coherence not being important IMO is just wrong.

Rip Van Woofer has already addressed this, no need for me to repeat it.

Have you ever heard a Lowther used in a bi-amped system with active xovers to a good woofer?

The Lowther is a poor full range driver in all important aspects as defined by credible perceptual research. It might be a pleasant coloration to some people, but it's hardly an example of a tecnically good driver if used as a full range driver.

While I agree that doing a proper full range is very difficult, we were after all discussing theory here. In theory, a single driver single point source is the optimal goal (short of directly driving air with no 'driver' - see Hill Plasma tweeter).

And it can't be done with a full range driver without severe compromises in critical parameters; certain exceptions exist for a substantial part of the normal compromises, such as the Manger driver(which is by no means a normal dynamic driver), but these exceptions are very limited.

To the item on the Dunlaveys, that's exactly my point. How many 'inexpensive well designed' speakers are out there that have really accurate, well done passive xovers? Not many. Dunlaveys hardly qualify for inexpensive - sorry.

Neither can one produce a well performing full range driver system for low cost. There is the cabinet properties/quality to be concerned with that contribute to performance, and a very good performing full range driver(like the Manger) costs a substantial amount of money. Even as a full range system, a passive network is needed to adjust the response finely to meet an intended curve target. Leaving the response unaltered is unlikely to be a good choice.

-Chris

 

[sploo]

This statement is too ambigous, as in it is not certain which compromises to which you refer.

The compromise to which I was referring was the choice between using a full-range driver (and not having the design time/complexity/cost of a crossover) vs. a 2/3 way system (more drivers, a crossover, but probably better sound).

Unfortunately, the most expensive Fostex full range units would lead to the cost of a pair of speakers being comparable to a bass/mid and treble pair of good drivers from Scan-Speak/Seas Excel etc. (in UK prices anyway).

I think it's pretty clear the two way system would be better in that instance.

However, for about a third of the cost, you could put together a pair of Fostex FE206 drivers. I'd be interested to hear the results (compared to a two way system + crossover at a similar price). Any thoughts on how they would fare?

On a, not entirely unrelated, subject: If I were to instead go for the two/three way setup, but used a Behringer DCX2496 as an active crossover, what sort of amps should I buy?

I know the answer depends on the drivers being used, but I get the impression that audiophile snakeoil (and inflated prices) also afflicts amps. Surely you just need two/three reasonably clean, suitably powered, stereo amp 'modules'?

 

[Rip Van Woofer]

I know the answer depends on the drivers being used, but I get the impression that audiophile snakeoil (and inflated prices) also afflicts amps. Surely you just need two/three reasonably clean, suitably powered, stereo amp 'modules'?

Yep and yep! Nice thing about active x-overs is that you have the amps directly connected to their respective drivers. Unless the drivers are weird (i.e., crazy low impedance or severely reactive) it makes for an easier load than the driver-plus-passive crossover, so a merely competent amp will do the job just fine. You can even use different wattage amps (less powerful for the tweeters and mids, more powerful for the woofers) as long as they have the same voltage gain. Sometimes you have to dig a bit for that spec.

(Do units like the Behringer allow you to "pad" the drivers for amps with different gains? That would be handy! I think some pro amps have variable gain, too.)

That's why I felt OK about going the DIY route for my active system's amps.

 

[WmAx]

The compromise to which I was referring was the choice between using a full-range driver (and not having the design time/complexity/cost of a crossover) vs. a 2/3 way system (more drivers, a crossover, but probably better sound).

Unfortunately, the most expensive Fostex full range units would lead to the cost of a pair of speakers being comparable to a bass/mid and treble pair of good drivers from Scan-Speak/Seas Excel etc. (in UK prices anyway).

And the full range drivers specified would not begin to approach the the critical parameters in performance of the multi-way.

However, for about a third of the cost, you could put together a pair of Fostex FE206 drivers. I'd be interested to hear the results (compared to a two way system + crossover at a similar price). Any thoughts on how they would fare?

.

Here is the thing: you can purchase superb quality midrange/bass drivers, such as Dayton branded reference drivers that are equivalent in objective technical performance to the big-name hi-end drivers, but for 25% of the price. For tweeters, you can also choose dayton, or you can choose superb quality devies like BG, which are low cost. You could never approach the objective performance possible with these low cost drivers in a multi-way by using any current full-range drivers; you would have to resort to using something like a Manger driver, which is very high in cost, to match the overall performance.

On a, not entirely unrelated, subject: If I were to instead go for the two/three way setup, but used a Behringer DCX2496 as an active crossover, what sort of amps should I buy?

I know the answer depends on the drivers being used, but I get the impression that audiophile snakeoil (and inflated prices) also afflicts amps. Surely you just need two/three reasonably clean, suitably powered, stereo amp 'modules'?

The amplifiers should be picked based on the specific drivers that you use. But generally, 25 watts for the tweeter, 50 for the midrange, 100 for a reasonably efficient woofer is a good proportion. The ideal specifications will change depending on your specific circumstances.

-Chris

 

[WmAx]

Yep and yep! Nice thing about active x-overs is that you have the amps directly connected to their respective drivers. Unless the drivers are weird (i.e., crazy low impedance or severely reactive) it makes for an easier load than the driver-plus-passive crossover, so a merely competent amp will do the job just fine. You can even use different wattage amps (less powerful for the tweeters and mids, more powerful for the woofers) as long as they have the same voltage gain. Sometimes you have to dig a bit for that spec.

Matched gain? Just use some potentiometers on the amplifier inputs to match up the levels if the active crossover used lacks such adjustments.

(Do units like the Behringer allow you to "pad" the drivers for amps with different gains? That would be handy! I think some pro amps have variable gain, too.)

The Behringer DCX has an analog gain stage with adjustable gain of +/- 15dB on each output channel. But if feeding a consumer line level amplifier, you still have to use potentiometers between the DCX outputs and the amplifier inputs, because the voltage is so high that you have to use up all of your adjustment gain by setting outputs to minimum setting to get good SNR performance and not clip the amplifier inputs(on some amps).

-Chris

 

[bpape]

WmAx,

You've done a wonderful job of avoiding my questions and simply restating your opinion or shifting the discussion. My statement was that while qualifying it as your opinion, you go on to present that opinion as fact unless someone can disprove it with scientific studies. I'll bet you were on the debate team in High School I guess I could do the same thing. So I guess that small cabinets are easier to get to disappear. Anybody got a scientific study that PROVES in a double blind fashion that they don't

1. Have YOU heard a Lowther/Fostex in an actively bi/tri-amped system? I have. I wasn't a big theoretical fan of the them either - until I heard them.

2. Again, the small cabinet helping to make the speakers disappear not being significant is simply a restatement of your opinion. I'll restate mine and say that I disagree - as do many, many, many others. We'll just agree to disagree and move on.

3. Rip stated clearly that part of what phase issues in xovers can cause is frequency related and ARE audible. While the phase itself may not be audible, other side effects are - or can cause additional xover complexity to compensate.

4. Agreed that a full range is very difficult to accomplish, but that doesn't make the theoretical single point source the optimal goal.

Please state the 'credible perceptual research' and the areas specifically the parameters where it is very poor. You throw these things out there but are very vague about them. Specifically what are the issues? Where in the performance range do they exist?

 

[sploo]

Yep and yep! Nice thing about active x-overs is that you have the amps directly connected to their respective drivers...

Excellent. That matches with what I was thinking (and the positive reasons for going active).

That's why I felt OK about going the DIY route for my active system's amps.

Is this one of Siegfried Linkwitz's creations? (I seem to remember you mentioning this in another thread).

And the full range drivers specified would not begin to approach the the critical parameters in performance of the multi-way.

I presume you mean distortion, off axis response etc., or are there other factors of which I'm not aware?

You could never approach the objective performance possible with these low cost drivers in a multi-way by using any current full-range drivers; you would have to resort to using something like a Manger driver, which is very high in cost, to match the overall performance.

So what we're basically saying here is that, unless there's a very specific reason for choosing a fullrange driver, you're far better off with a multiway system.

Do you have any experience with the coaxial units by Seas? They're a combination of a treble and woofer in a single cone, so presumably they aren't as heavily compromised as a fullrange unit?

Accepting that a crossover is going to be required... Assuming you get the basics right in a passive crossover design, is it possible for a novice to design a reasonable sounding crossover? In other words, given the same drivers and cabinet, would a skilled crossover designer be able to create a massively better sounding system than someone just following the guidelines of a book, and maybe some simple crossover design software?

Alternatively, thinking in active terms...

The amplifiers should be picked based on the specific drivers that you use. But generally, 25 watts for the tweeter, 50 for the midrange, 100 for a reasonably efficient woofer is a good proportion. The ideal specifications will change depending on your specific circumstances.

...

The Behringer DCX has an analog gain stage with adjustable gain of +/- 15dB on each output channel. But if feeding a consumer line level amplifier, you still have to use potentiometers between the DCX outputs and the amplifier inputs, because the voltage is so high that you have to use up all of your adjustment gain by setting outputs to minimum setting to get good SNR performance and not clip the amplifier inputs(on some amps).

OK. That all sounds fine.

So, if you're buying drivers of reasonable expense, then it's probably worth the extra cost over a passive xover of using a DCX (plus some pots) and a couple of amps.

I get the impression you're indicating that because the DCX will probably need pots anyway, then exactly matching amps aren't critical as you could balance them with a bit of tweaking?

So... a question to both of you (and anyone else that might know) could you recommend some decent (sensibly priced) amps that would be suitable for driving a two or three way stereo system following an active xover (i.e. four or six channels driven at say 40-100w per channel)?

Thanks.

 

[WmAx]

WmAx,

You've done a wonderful job of avoiding my questions and simply restating your opinion or shifting the discussion. My statement was that while qualifying it as your opinion, you go on to present that opinion as fact unless someone can disprove it with scientific studies.

Incorrect. I stated that nothing of credibility of which I am aware shows that they will make the substantial differences claimed. What I did was to offer plausible reasons(based on specific circumstances of what will for certain affect imaging consistancy) as to why they may be erroneously concluded to offer said performance. I do no such thing as offer my opinion as fact. Go back and read my 1st reply in the thread and the posts in the linking thread that are addressed to you.

1. Have YOU heard a Lowther/Fostex in an actively bi/tri-amped system? I have. I wasn't a big theoretical fan of the them either - until I heard them.

I do not offer my personal impressions or what I 'hear' in a factual debate. It does not matter if I am deaf. But I am curious : How was this Lowther that you refer to used as a full range, if it is in tri-amped configuration?

Again, the small cabinet helping to make the speakers disappear not being significant is simply a restatement of your opinion. I'll restate mine and say that I disagree - as do many, many, many others. We'll just agree to disagree and move on.

If a small cabinet width as found on very small bookshelf monitors has a substantial improvement effect on imaging in itself, then by this logic, only speakers with such a small cabinet baffle must have excellent rated imaging. No speakers with average or large size baffles can have excellent rated imaging based on this simplisitc assumption. There are several variables relevant to the situation, and so far as I can tell, they are being ignored by the vast majority of people, and a conclusion is being prematurely concluded.

3. Rip stated clearly that part of what phase issues in xovers can cause is frequency related and ARE audible. While the phase itself may not be audible, other side effects are - or can cause additional xover complexity to compensate.

Rip Van Woofer gave a limited amount of information. And what was not further stated, but what needs to be stated now, is that a multiple source system trying to adhere to the best on axis phase response(1st order) is going to cause far more problems in frequency response off axis as compared to a steeper slope rate. His blurb about frequency response difference was a qualifier. The higher order system, using the same drivers and layout, will have superior frequency response linearity off axis, and equal response on axis, as compared to the 1st order system. Additionally, a properly engineered 1st order system will result in increased cost, because the drivers will have to operate in a significantly wider bandwidth, and thus be required to have greater power handling as compared to a higher order system. The cost of getting better on axis phase response(which has little to no audible effect on it's own) is to eschew off axis response(which is a critical parameter for audibility) and to require drivers with a wider usable bandwidth and higher power handling. No cost is saved. No complexity is saved(because it requires as complex or a more complex crossover to tame drivers to operate optimally in this wide bandidth). Important parameters are ignored to focus on a parmater that has not been demonstrated to have a signifncant effect on audibility.

Please state the 'credible perceptual research' and the areas specifically the parameters where it is very poor.

I assumed that you already knew which parameters I was referring to; that's why I did not specify them earlier. Here are the critical areas as demonstrated in [1][2][3]credited perceptual research, where a Lowther or Fostex or other popular full rand driver falls flat on it's face:

-polar response(due to the effective radiation area, upper midrange and treble bands beam severely, causing a very different, and poor response off axis)

-frequency response(response is not smooth)

-resonances(frequency response is not linear, maninfesting many severe resonances especially into upper midrange and into treble)

-frequency response bandwidth(the low frequency respnse is not actually full range with these small drivers, they can not respond to the lowest frequencies effectively. Another driver must handle bass if true full range is desired.)

-non-linear distortion(the speakers can not remain linear at the excursion/incursion required for true full range use at lower frequency SPL at moderate to high SPLs, thus resulting in high levels of THD)

Additionally, a parameter I am not aware of a specific perceptual paper, but is a very obvious parameter which must be addressed:

-phase modulated distortion(a.ka. Doppler Distortion)(the speakers can not really be used as full range, they must have another driver function as the bass driver, otherwise at moderate to high SPL, Doppler Distortion will manifest in the upper midrange and treble bands due to excessive cone movement while reproducing frequencies with very short wavelenghts)


-Chris

[1]Loudspeaker Measurements and Their Relationship to Listener Preferences: Part 2
Floyd E. Toole
JAES, May, 1986, Vol. 34, pages 227-235

[2]"The Modification of Timbre by Resonances: Perception and Measurement", Floyd Toole, Sean Olive, JAES, Vol. 36, No. 3, 1988, March, pages 122-141

[3]Just Detectable Distortion Levels
James Moire, F.I.E.E.
Wireless World, Feb. 1981, Pages 32-34 and 38

 

[WmAx]

I presume you mean distortion, off axis response etc., or are there other factors of which I'm not aware?

Please refer to my post to bgape, where I outline the problems, along with the relevant perceptual study references.

So what we're basically saying here is that, unless there's a very specific reason for choosing a fullrange driver, you're far better off with a multiway system.

Correct.

Do you have any experience with the coaxial units by Seas? They're a combination of a treble and woofer in a single cone, so presumably they aren't as heavily compromised as a fullrange unit?

I have not used, nor am I immediately aware of the specific measurements of the Seas co-axial units, but such a design does offer potentially signficant improvement over a standard single driver full range unit in regards to polar response and resonances behaviour.

Accepting that a crossover is going to be required... Assuming you get the basics right in a passive crossover design, is it possible for a novice to design a reasonable sounding crossover?

What is the definition of 'reasonable'?

In other words, given the same drivers and cabinet, would a skilled crossover designer be able to create a massively better sounding system than someone just following the guidelines of a book, and maybe some simple crossover design software?

This is about probability, and can not be answered with absolutes. But with high probability, the skilled crossover designer will have a far improved result as compared to the beginner.

I get the impression you're indicating that because the DCX will probably need pots anyway, then exactly matching amps aren't critical as you could balance them with a bit of tweaking?

Right, there is no good reason to match amplifier exactly.

-Chris

 

[Rip Van Woofer]

WmAx accurately characterized my post. His points about the drawbacks of first-order crossovers also accords with what I know. They can work well IF the designer knows what he is doing. Thiel speakers come to mind. But pursuing phase coherence as an end in itself, or as the primary determinant of a speaker's sound, is unsupported by the available evidence.

Coaxial drivers too depend on how well they are implemented. There have been both very good and very bad ones. The very concept carries its own engineering difficulties and problems -- just like every other speaker design. Designing and producing a good one is not a trivial exercise. Fact is, there are very few truly successful (in both market and engineering terms) coaxial driver systems on the market today. That says something, IMO.

Perhaps the best coaxial speaker ever made was the short-lived Win SM-10. The desginer, a physicist named Sao Zhao Win (sp?), created a driver that was both coaxial and co-planar; the woofer and tweeter were both flat and in the same plane. Some high-tech materials science and gnarly math made it possible AFAIK. Pictures I have seen make it look like a high-tech Frisbee bolted onto a box. I have not seen or heard one, but I'd love to someday.

As for big vs. little cabinets "disappearing" all I can say is that my speakers are fairly big, and they "disappear" quite nicely. Purely subjective opinion, of course.

 

[sploo]

Please refer to my post to bgape, where I outline the problems, along with the relevant perceptual study references.

Read and absorbed (and, I think, understood). Thanks.

I have not used, nor am I immediately aware of the specific measurements of the Seas co-axial units, but such a design does offer potentially signficant improvement over a standard single driver full range unit in regards to polar response and resonances behaviour.

For an project I was thinking of doing, having a single driver would be an advantage. However, I'm beginning to wonder if the compromises required in these drivers means that spending the equivalent money for a bass/mid and treble pair would probably yield a superiour speaker.

I'd love to just go and buy (and test) some of these drivers, but I can't really afford to spend money on things I'm not going to use.

What is the definition of 'reasonable'?

42?

No?

OK, fair point.

This is about probability, and can not be answered with absolutes. But with high probability, the skilled crossover designer will have a far improved result as compared to the beginner.

I can quite believe.

I don't doubt that I could create a crossover, I'm just beginning to wonder if the time/effort, and cost (due to buying lots of different resistors/caps/coils), means that an active solution would be much better.

I'll come clean here; I design software, and at home I'm happier with a router and a bandsaw than a soldering iron. I've had some ideas for creating 'interesting' shaped loudspeaker cabinets, so I want to concentrate on designing/making them, and I want the electrical/electronic bits to 'just work'.

I have considered using existing designs - plenty available on the net, but I'm aware that crossovers are sometimes tailored to the box (baffle step etc.) and I'm unsure about the etiquette of producing a speaker using someone else's driver configuration/crossover design.

I guess for the moment, the best thing I could do is to buy some good value drivers, an active crossover (e.g. DCX2496), a couple of stereo amps, and have a play...

PS I've found a Behringer SUPER-X PRO CX2310 (about a third of the price of the DCX2496). Am I right in assuming the DCX is a DSP based digital xover, and the CX is an analog unit?

 

[bpape]

WmAx

You're using false logic. To state that small speakers are easier to get to disappear leads to only small face speakers being able to image is not a valid argument. Sorry. Just because women overall are more sensitive does not mean that only women can be sensitive - see the flaw?

Your opinion DOES matter in terms of what you hear when you put it forth as the way it is unless proved otherwise. So now that we know you have not hear them, how can you make such statements? Simply by reading others' writings? By looking only at measured specs?

The Lowthers were obviously not full range when bi-amped. The point is that you poo-poo'd them as being horrible, techically inferior drivers. My point was that when used in their optimal range and NOT being restricted by passive xovers, they can do a good job. This is back to the original point of them being primarily a point source (other than the bottom end which is primarly omnidirctional and non-localizable if set up and x'd over correctly) and avoiding the pitfalls of a complex passive xover.

Yes. 1st order xovers DO require that someone know what they're doing - as do most thing in the audio world. Poor or mediocre design no matter the philosophy lead to mediocre products. Good design, no matter the medium, lead to good products - that's irrelevant to the discussion.

The point you originally made was to compare a 'typical' full range system to a well designed multi-way system. I'll ask again, how about comparing a well designed simple xover system to a well designed complex multi-way system? All else being the same, the simple one will perform better FOR THE DOLLAR than the complex multi-way simply due to economies of the components.

I don't want to be difficult - but it just rubs me the wrong way when someone makes categorical statements that something is inherently inferior when they've never even heard what they're bashing. For the 3rd time, I'll ask specifically which parameters and measurements you're referring to when you say the Lowther/Fostex/etc. fail in every significant technical measure.

Everyone has their own opinions. Everyone has their own tastes. I personally HAVE an expensive, well designed, multi-driver system that I like very much. But, that doesn't blind me to other things and their strengths simply because someone else says so or because I read some techincal paper that shows some measurements to be less than desirable in some aspects.

I DO care what my ear says - after all, that's what audio is all about, isn't it?

 

[WmAx]

You're using false logic. To state that small speakers are easier to get to disappear leads to only small face speakers being able to image is not a valid argument. Sorry. Just because women overall are more sensitive does not mean that only women can be sensitive - see the flaw?

If your contention that small baffle width substantially helps imaging is true, then all normal width speakers would benefit substantially from a small face rather than an average width face, if they do not have special means to eliminate diffraction. Thus, based on this contention, a small face will always have better imaging(and you seem to contend that it's a substantial effect), thus average face width size speakers without special diffraction control means would never be able to have such highly rated imaging, since a small face one woud always be 'better', based upon your contention.

If average width speakers with no elaborate method of diffraction reduction can image in a way rated of comparable excellence to small width speakers, then the variables involved are not so simple as your contention would imply.

The point is that your contention is not supported by any real coorelated studies of which I am aware. It is your unsupported opinion(and your positive claim) that small baffles have a significant effect upon imaging. I am pointing out that is an unverified speculation at this point.

In order to have a substantial effect in removing diffractive effects by way of reducing baffle width as a lone variable, one needs a baffle that is small in relation to the wavlengths relevant to frequencies primarily responsible for imaging/localization cues, so that they can not reflect off of the surface(the driver is part of the surface area in this respect). To improve the diffractive effects of frequencies that are actually responsible for most of the imaging component(treble) by means of a small width, the tweeter needs to essentially have no baffle surrounding itself; B&W Nautilus speakers have a tweeter that takes this approach. Even a small speaker baffle of 6" is large when considering the wavelengths emitted from the tweeter. Note: I make no claims of audibility one way or the other, I am merely pointing out the surrounding facts/conditions.

Your opinion DOES matter in terms of what you hear when you put it forth as the way it is unless proved otherwise. So now that we know you have not hear them, how can you make such statements? Simply by reading others' writings? By looking only at measured specs?

If this thread was about my opinions, then they would be relevant. However, this thread is not about my opinions. I limit factual response based upon the established scientific works

I'll ask again, how about comparing a well designed simple xover system to a well designed complex multi-way system? All else being the same, the simple one will perform better FOR THE DOLLAR than the complex multi-way simply due to economies of the components.

I see no definate way to draw absolutes on the value you refer to without a defined meaning/weighting of that value(which is not provided), along with a specific case to analyze. Therefor I can not answer your question sufficiently.

I don't want to be difficult - but it just rubs me the wrong way when someone makes categorical statements that something is inherently inferior when they've never even heard what they're bashing.

Becuase they are inherantly inferior. See next line.

For the 3rd time, I'll ask specifically which parameters and measurements you're referring to when you say the Lowther/Fostex/etc. fail in every significant technical measure.

I already outlined this in the last reply. I have reviewed 3rd party measurements of Lowther drivers, and of several Fostex, along with many other full range drivers over the years, and they are inherantly poor in all of the parameters I listed previously when used as full range drivers. Not a suprise, since well understood physical limitations cause this.

-Chris

 

[WmAx]

For an project I was thinking of doing, having a single driver would be an advantage. However, I'm beginning to wonder if the compromises required in these drivers means that spending the equivalent money for a bass/mid and treble pair would probably yield a superiour speaker.

I will remind you of one thing: you do not personally need to be concerned with off axis performance. This is because (if my memory is correct) your listening room has virtually no ambient ability, with absorbitin on the majority of surfaces, and no 1st reflections are existant. Your room resembles the acoustics of an anechoic chamber, essentially. If I am confusing you with someone else, please correct me.

I have considered using existing designs - plenty available on the net, but I'm aware that crossovers are sometimes tailored to the box (baffle step etc.) and I'm unsure about the etiquette of producing a speaker using someone else's driver configuration/crossover design.

So long as you stay within the approximate width of the original plans, and ensure that the tweeter/mid are place in the same approximate positions relative the original top boundaries, and in the same distance/spacing as the original plans, it should remain reasonably close.

PS I've found a Behringer SUPER-X PRO CX2310 (about a third of the price of the DCX2496). Am I right in assuming the DCX is a DSP based digital xover, and the CX is an analog unit?

The CX is an analog unit, and it lacks the ability to customize the target curves and slop rates. It is not suited as well as the DCX to crossover design.

-Chris

 

[bpape]

WmAx.

Sorry for missing your responses earlier - was skimming.

Your logic is still wrong on the baffles. You CANNOT make the logical jump from narrow baffles helping to all narrow baffled speakers being better than wider ones or certainly not to only narrow baffled ones being ABLE to image. There are many other things that contribute to good imaging. Narrow baffles help but can't overcome other major technical blunders. Obviously you're not seeing the logic flaw (my college professor for logic would have a cow over this one). This is one of the classic logic flaws taught to show what not to do. I guess we'll have to move on.

I do find it interesting that many of the things you claim make the Lowther/Fostex drivers problematic and inferior would be eliminated if they were used in a multi-way system. If you measured other drivers that are normally used in a multi-way system I suspect you'd find many of those same issues - not all but many.

Also, thank you for helping to make the point about 1st order xovers and mutiple drivers causing issues. That's EXACTLY my point. In theory, a 1st order xover is superior. However, the complexity in a multi-way system almost necessitates a MUCH more complex (read more phase issues and more expense and more changes to muck it up or use cheap components) xovers.

Again, don't get me wrong. Properly implemented, with assumptions that you'll spend the extra money for the speaker and the extra money for the beefier amp to deal with any odd loads due to strange xover/driver interactions, a multi-way system can sound very very good.

I'll also still say that if you use an actively bi-amped system with a good full range driver and relieve the bottom end with a capable driver, you can also get excellent performance. Are some of the issues you brought up valid? Absolutely. Some of the full range drivers can be a bit beamy - but NOTHING like the old Acoustats or Stax's that were also excellent speakers but defnitely for one person at a time only - and these don't have the vertical dispersion issue either.