Uhm. I'm a bit confused. So stiff material isn't a good match to soft material. For example C95 to Esotar.
But how does this explain excellent reviews of Seas Thor/Odin and Linkwitz Orion for example? All use magnesium mid/midwoofers and soft dome tweeter.
Jussi
But how does this explain excellent reviews of Seas Thor/Odin and Linkwitz Orion for example? All use magnesium mid/midwoofers and soft dome tweeter.
Jussi
Jussi said:
Generally, yes.
As to the reviews.. most reviewers don't look at it, nor do most manufactureres. It doesn't mean that it isn't there though.
Furthermore, the lower the output impeadance of the paired amplifier, generally the less noticible it will be. EX. I believe S. Linkwitz (Orion's designer) uses amplifiers based on the LM3886 "opamp", which has a very low output impeadance. The chances are much greater that he won't percieve a tonal difference between the Millenium soft tweet and the mag. midrange. Now if he were to use a standard OTL amplifier (with an output impeadance around 14 ohms) - I'd think it much more likely that the differences in drivers would be noticed* (the millenium would likely sound "dull" in comparison to the mag. driver).
*and this is without reference to differences due to freq. response changes that would most likely be evident with a typical OTL and dissimlar driver impeadances vs. freq. (..which would also be audible and cause tonal balance changes).
So I assume high(ish) output impedance would suggest low or non feedback amplifier with relatively low dampingfactor. But why does this increase the difference? And is there something wrong with low output impedance amps in this mather? Is this difference necessary to hear?
And suggestions for Seas W18EX or C95 Accuton tweeters are also welcome.
Thanks!
Jussi
And suggestions for Seas W18EX or C95 Accuton tweeters are also welcome.
Thanks!
Jussi
Jussi said:
Generally, yes. As to why.. IMO (and others who recognize the difference) most mid to high freq. drivers are overdamped.. compound this with a traditional high damping factor and the overall sound is "wrong" (for lack of a better descriptor). This generality however does not neccesarily "extend" to the lower midrange and below. Now this doesn't mean that a low output impeadance amplifer is wrong per se, just that generally it is wrong for mid-to-high freq. drivers when the listener desires a realistic sound. I've shown this effect to "experienced" and "novice" listeners alike and its always audible. Of course the problem is that the lower freq.s can become under-damped.
If you've got the esotars I'd stick with them and get the midrange you like. And there are plenty of other options besides the Audax's I mentioned.. its just that I can't think of a better driver for the esotar's than Audax's in an MTM when considering ALL factors (including things like eff., impeadance, baffle step loss, off-axis response, ease of crossover design, cost, etc. - in addition to the tonal properties of the driver.)
If your willing to spend a LOT more you could consider the premium Skanning 4 inch flex unit.. but even it isn't as versital as the Audax driver (at least as far as the crossover is concerned). There is also the Fostex FW16HP-X which has not only bafflestep effectivly taken care of, but works well in a minimalist crossover because it was designed to be in a first order crossover without an inductor (i.e. run full-range with a high quality cap on the tweeter slightly above 3 kHz). It however is quite a bit more expensive than the Audax (though considerably less than the Skanning). Some of the SS revelators could also be considered (though all are more difficult with crossovers and again are expensive).
..well you have my ideas on the Seas Magnesium drivers and the Accuton's in general.
454Casull said:
Most definitly. Jussi, It could be that you won't hear a difference in drivers tonally, and/or you might not here a difference with respect to dampining factor. (or perhaps more precisely you might here so small a difference for either that it wouldn't concern you.) For instance while cyclo could hear the difference in drivers tonally, he didn't hear negative effects with high damping factor (..or perhaps he simply hasn't heard a low damping factor amp/speaker pairing with a ribbon b4).
The best thing is (if you can), to go listen to designs with disimilar driver materials - particularly those that have similar drivers to what your interested in. The Kharma Cermique 1 is I believe still in production - you might try to find this to listen to and then try a speaker from Dynaudio, and then perhaps a Sonus Faber design (and better still try them with amplifiers that have different output impeadances). Each speaker has quite different midranges while having moderatly similar tweeters (..though I think the Esotar is substantially better than any of them).
Despite this however, you'll still have to "factor" in things like baffle step and efficiency even considering a theoretical gain of 6db by going parallel with midranges.
Ok. Sounds pretty reasonable.
As for drivers I'm looking for a low distorsion combination. I know Esotar has very respectable THD ratings and in that aspect Excels or maybe Accutons might match it.
There are a great variation of different variables that all distribute to overall performance so comparing two different systems with different drivers doesn't tell that much. Defining one variable systems would have to be identical in all possible aspects and listen them in same acoustic environment using same electronic equipment.
This low damping electronics idea is intresting. You mean drivers suspension is normally overdamped and it needs low damping amp to give best possible performance? What's the problem with overdamping it some more with high damping amp?
Again in this aspect differences can be created from other factors. For example low damping valveamp and high damping solid state have very different behavior in several variables. Distorsion, compression, maybe even response... So examing this would require identical ampfiliers with only different damping factors.
Jussi
As for drivers I'm looking for a low distorsion combination. I know Esotar has very respectable THD ratings and in that aspect Excels or maybe Accutons might match it.
There are a great variation of different variables that all distribute to overall performance so comparing two different systems with different drivers doesn't tell that much. Defining one variable systems would have to be identical in all possible aspects and listen them in same acoustic environment using same electronic equipment.
This low damping electronics idea is intresting. You mean drivers suspension is normally overdamped and it needs low damping amp to give best possible performance? What's the problem with overdamping it some more with high damping amp?
Again in this aspect differences can be created from other factors. For example low damping valveamp and high damping solid state have very different behavior in several variables. Distorsion, compression, maybe even response... So examing this would require identical ampfiliers with only different damping factors.
Jussi
Those Audio Technology drivers also crossed my mind. I'd guess they as Dynaudio based drivers are a good match with Esotar?
Also thought of overall configuration. Maybe that MTM isn't even necessary. System is going to be 3-way so propably just single HQ midrange will do. C-Quenze 7" from Audio Technology?
Jussi
Also thought of overall configuration. Maybe that MTM isn't even necessary. System is going to be 3-way so propably just single HQ midrange will do. C-Quenze 7" from Audio Technology?
Jussi
Jussi said:
Ahhh, a possible change in design..
Perhaps the greatest limiting factor you have is Baffle Step compensation with regard to driver eff. IF you alter your design so that the dipole bass drivers extend up a bit higher in freq. without a substantial "slope" (i.e. make the design 2nd order or less for the lowpass on the woofers) then you "open up" your choices quite a bit. At that point one Skanning 4 inch flex unit for the midrange should work quite well with only minor "padding" of the tweeter.
I've looked over the flex units vs. the C-quenze drivers.. IMO IF your spending as much money on the C-quenze then why not spend more for the flex-units? I.E buy the best (that works with the design) and "don't look back" (..which I suspect was what motivated you to purchase the Esotars).
You might want to look at the new Dayton Reference subs:
http://www.partsexpress.com/pe/showdetl.cfm?&DID=7&Partnumber=295-464
Review here:
http://www.mfk-projects.com/woofers.htm
Note that the Review also said the driver was good higher in freq. (which would work well for the dipoles up to at least 300 Hz).
The mid has to be at least 7" in size. Propably a 8" model would work as well. Audio Technology carry both and both should be pretty good. Unfortunately there are not so much reliable measurements, distorsion graphs and so on from AT products. AT informs their drivers to be "very low distorsion" but no hard evidence to back it up.
The size is relevant since I'm going to use acoustic resistance cabinet for the midrange and use cardioid or hypercardioid radiation pattern for mid-treble section. Chech Finnish Gradient (www.gradient.fi) and Amphion (www.amphion.fi) for backround information.
So 7" is the minimum and I'd guess 8" the max. I need cone diameter to extend directivity to whole frequency range.
Do you have recommendation for Flex unit parametres for such task? I thought to use 120-150Hz 2nd order cross between separated dipole woofers and the top, depending on MT/MTM configuration, midrange size and overall performance.
I've ask some info on those Flex units and they could produce a nice 94dB efficient 8" custom for me. Unfortunately linear excursion only +-1,5mm and that was requested for 300Hz+ area midrange in my earlier designidea.
Two pairs of Excels cost about 600€ here, I could have those C95 in the same price. Custom Flex 8" mids cost more than that, custom 7" Flex a bit less...
AT looks nice, respectable ideas and response but still a bit like pig in a sack. Any solid info of them?
Thanks !
Jussi
The size is relevant since I'm going to use acoustic resistance cabinet for the midrange and use cardioid or hypercardioid radiation pattern for mid-treble section. Chech Finnish Gradient (www.gradient.fi) and Amphion (www.amphion.fi) for backround information.
So 7" is the minimum and I'd guess 8" the max. I need cone diameter to extend directivity to whole frequency range.
Do you have recommendation for Flex unit parametres for such task? I thought to use 120-150Hz 2nd order cross between separated dipole woofers and the top, depending on MT/MTM configuration, midrange size and overall performance.
I've ask some info on those Flex units and they could produce a nice 94dB efficient 8" custom for me. Unfortunately linear excursion only +-1,5mm and that was requested for 300Hz+ area midrange in my earlier designidea.
Two pairs of Excels cost about 600€ here, I could have those C95 in the same price. Custom Flex 8" mids cost more than that, custom 7" Flex a bit less...
AT looks nice, respectable ideas and response but still a bit like pig in a sack. Any solid info of them?
Thanks !
Jussi
Jussi said:
All of the above sounds (and is logical), despite this however I haven't found any of it to be terribly accurate. In other words even comparing different systems, while specifically listening for certain traits (like damping factor and tonality with driver material), usually yields audible similarities/dissimilarities that are fairly representive in general.
In fact I did this about 2 weeks ago (went around to the local HiFi shops in my area) and listened to the "latest and greatest" (..most of it wasn't so great). (Something I do about once every 9 months or so.) Sure there were MANY other differences, BUT dispite this damping factor (freq. dependent) and tonal character (from different driver material) was fairly identifiable almost imeadiatly (..because its something I've trained myself to listen for).
Now what I haven't found to be imeadiatly identifiable is harmonic distortion. I've tried b4, and this last time as well, but in no system could a point and say something like "this is higher in 2nd order distortion". One system may well have higher harmonic distortion than another, but it isn't easy to identify at the levels you commonly find in most systems. The ONLY time I've been able to identify an increase in harmonic distortion (with different systems) is with very low freq.s in subwoofers (or loudspeakers).. and it doesn't sound like a "flavor", rather it sounds louder and less clear when its higher in level.
Additionally..
One of the other things I listened for this time was dipole driver (vs. non-dipole) character depending on the freq.. In every case with widely varying loudspeakes, dipoles operating between 60 Hz and 300 Hz sounded MUCH more realistic because they didn't move instruments and voices (reproduced in this range) forward in the soundstage, ("bunched-up"), like every box loudspeaker I've listened to.
Jussi said:
Ok, your going to aperiodic vent to the side of the cabinet to control directivity lower in freq. for the midbass?
Why do you need or want to control directivity higher in freq. (near the top of the midbass's passband)? I've always found that the more omni-directional higher in freq. the more realistic the sound (all-else-equal, ..though it never is) - provided of course that the wavelength isn't in "correlation" with the room's walls.
Moreover, the radiation pattern of drivers like 7- 8 inches at the top of their passband display a very non-linear pattern when they start to become directive. (for lack of a better term.. its "sloppy".)
Finally, unless you load the Esotar with a waveguide it will be difficult to even approach the directivity of the midbass at the top of its passband. If that happens (no waveguide), the crossover becomes that much more difficult to get "right". (I believe SL slightly pads down the level of the tweeter below that of the dipole mid to achieve some measure of "continuity" - still a "tricky business" to get right.)
I can understand the desire to control radiation near the dipole midbass.. but I haven't found that it is warrented. I specifically use 2 10 inch drivers dipole up about 280 Hz (from about 65 Hz) with a monopole midrange and have never noticed a discontinuity in doing so (though I do use a 1st order low pass on the dipole portion). In fact I've also tried the same midrange in dipole configuration and the sound was noticably worse (less depth, less dynamic, tonally a bit "thin").
As to the Skanning drivers..
Any time you raise eff. you'll decrease bandwith and likely decrease excursion (for several reasons). Additionally, mass is usually whats lowered and because of it the VC becomes lite(er) and the freq. response usually increases near the top of its passband as a result. I don't think that any two excels (even in parallel with a voltage amp) will give you the same eff.. Moreover, double drivers operating in the same freq. range limit the top of your bandwidth (for a variety of factors), and neccesarily limit your choice of crossover slope. (..neither of which may be a bad thing.. just something to consider.)
After looking back at the Seas stuff again, IF you don't need to concern yourself with baffle step issues, the new Nextel 7 inch MIGHT be workable in a dual driver situation if you crossover the drivers fairly low (..and presumably use a waveguide for your esotar). They certainly are less expensive than the Skanning drivers (even the C-quenze versions), and presumably have similar distortion levels to their magnesium cousins while probably having significantly less disturbing effects in their "break-up" region.
Scott,
Just for my curiosity...
How do you define "natural sound"?
I think all these experiments with different variables should be done with blind listening. It's very easy to hear something in certain way if some variables like cone materials are known and seen in the listening situation.
So for example listening to metal cone soft dome combination with low damping and high damping electronic gear. It's easy to hear problems and discontinuous behavior while listener is aware of what's playd and with what gear.
But I'm not starting to argue about these things. Mostly just technical ******** and don't have that much to do with listening music. And this is also pretty off topic subject.
As for the cabinets and driver business...
I've got a MTM acoustic resistance prototype but I'm not sure am I going to use it or go simpler MT construction. But the main idea would be more or less constant directivity design. Having 7" mid and running it 2Khz+ sure it beams and energy response is dropping above 1Khz. But this is corrected with resistance cabinet structure. The idea is to have flat response from all angles. Only more or less cancelled below cones own directivity and therefore more overall directivity. Full monopole is one example of this but doesn't work that well in rooms since early reflections mess it all up.
Off axis behav is also one subject where soft cones and hard cones differ and cause differences in overall tonal balance if it isn't taken into count.
So the cone directivity isn't suppose to run any narrower than acoustic resistance beam at lower frequences. And Esotar is going to have a waveguide to match it to lower ranges.
As for custom drivers I don't see that many parametres I'd like to fix and use Flex units over C-quenze 7" if midrange is covering 150-2Khz+ range. With that earlier 300Hz+ midrange gives a bit different aspects and needs.
But I still don't know any facts of Audio Technology drivers. Or from Accutons either. Both are praised sure but why. Reading Stereophile archives and I can find couple Accuton designs. Sure, details details details, but why?
In this aspect I can count myself as lunatic. I want to know why some system sounds like that, I don't settle for information that it sounds like that and that's it.
AT doesn't have that much measurements of their drivers. Accuton has but only with one volume and at least responses are all smoothed. Expensive pigs in a bag I'd say.
And as for total pigs for example all praised 200 euros each Scan Speak new sliced paper 15W sucks in THD. One very reliable contact measured one of them and got ratings equal to 60 euro Vifa standard. And what I've heard from SS present design methods they don't even seem to try making maximum neutral drivers. Just ready to use instruments that have their own sound. It's just designers business to enhance this effect and have something. Mostly these are called "musical" systems. But it's impossible to make a neutral device from driver that has several % distorsion in reasonable levels.
They are all praised. I just ask what for? At least not for their neutral sonical presentation. Or at least in absolute scale.
I hope you now understand my need for reliable objective information and I like to take all possible variables into count and don't like guessing that much. Or naturally having as limited knowledgebase as whole humankind has all things we do are just more or less educated guesses. But my point would be having more educated guessing before firing to all directions and hope to hit something.
And while even hitting "something" that sounds "something", how to define are we closer to the truth or is this hit just a nice solution that pleasures human hearing?
But like said. Absolute neutrality, impossible to even define, but maybe a good target to shoot for technically. As for music listening most of it doesn't have anything to do with it. People just get caught into some ridiculous technical detail and make bull out of a fly.
But all this only as my very subjective opinion. I don't have power or almightly knowledge to make any deeper conclusions.
And as for my design which was the topic here...
AT drivers are a exotic guess, more or less pigs in a bag. Unfortunately they don't deliver with right to return them if they fail in measurements. And I sure don't want to invest 250 euro midwoofer that has performance of a 60 euro unit. Don't like that status value that high and personally I have pretty poor imagination to imagine them sound that much better...
W18EX is a strong candidate. Expensive but also tested in several places getting excellent results. Peerless has very good HDS serie which would reach the same with copper rings in their magnetsystem. Shame.
Accutons I don't know. Small shop just like AT. Make stuff by hand and with small tolerances price is astronomical. I just wonder why people still favor handmade things. With a robot you can install spider, upper suspension, cone and VC to chassis having below 0,1mm tolerances. Have to got very steady hand to achieve the same with human hands. And this in long period of time, year after year. And still machine stamps things in ratio of 10 people.
Jussi
Just for my curiosity...
How do you define "natural sound"?
I think all these experiments with different variables should be done with blind listening. It's very easy to hear something in certain way if some variables like cone materials are known and seen in the listening situation.
So for example listening to metal cone soft dome combination with low damping and high damping electronic gear. It's easy to hear problems and discontinuous behavior while listener is aware of what's playd and with what gear.
But I'm not starting to argue about these things. Mostly just technical ******** and don't have that much to do with listening music. And this is also pretty off topic subject.
As for the cabinets and driver business...
I've got a MTM acoustic resistance prototype but I'm not sure am I going to use it or go simpler MT construction. But the main idea would be more or less constant directivity design. Having 7" mid and running it 2Khz+ sure it beams and energy response is dropping above 1Khz. But this is corrected with resistance cabinet structure. The idea is to have flat response from all angles. Only more or less cancelled below cones own directivity and therefore more overall directivity. Full monopole is one example of this but doesn't work that well in rooms since early reflections mess it all up.
Off axis behav is also one subject where soft cones and hard cones differ and cause differences in overall tonal balance if it isn't taken into count.
So the cone directivity isn't suppose to run any narrower than acoustic resistance beam at lower frequences. And Esotar is going to have a waveguide to match it to lower ranges.
As for custom drivers I don't see that many parametres I'd like to fix and use Flex units over C-quenze 7" if midrange is covering 150-2Khz+ range. With that earlier 300Hz+ midrange gives a bit different aspects and needs.
But I still don't know any facts of Audio Technology drivers. Or from Accutons either. Both are praised sure but why. Reading Stereophile archives and I can find couple Accuton designs. Sure, details details details, but why?
In this aspect I can count myself as lunatic. I want to know why some system sounds like that, I don't settle for information that it sounds like that and that's it.
AT doesn't have that much measurements of their drivers. Accuton has but only with one volume and at least responses are all smoothed. Expensive pigs in a bag I'd say.
And as for total pigs for example all praised 200 euros each Scan Speak new sliced paper 15W sucks in THD. One very reliable contact measured one of them and got ratings equal to 60 euro Vifa standard. And what I've heard from SS present design methods they don't even seem to try making maximum neutral drivers. Just ready to use instruments that have their own sound. It's just designers business to enhance this effect and have something. Mostly these are called "musical" systems. But it's impossible to make a neutral device from driver that has several % distorsion in reasonable levels.
They are all praised. I just ask what for? At least not for their neutral sonical presentation. Or at least in absolute scale.
I hope you now understand my need for reliable objective information and I like to take all possible variables into count and don't like guessing that much. Or naturally having as limited knowledgebase as whole humankind has all things we do are just more or less educated guesses. But my point would be having more educated guessing before firing to all directions and hope to hit something.
And while even hitting "something" that sounds "something", how to define are we closer to the truth or is this hit just a nice solution that pleasures human hearing?
But like said. Absolute neutrality, impossible to even define, but maybe a good target to shoot for technically. As for music listening most of it doesn't have anything to do with it. People just get caught into some ridiculous technical detail and make bull out of a fly.
But all this only as my very subjective opinion. I don't have power or almightly knowledge to make any deeper conclusions.
And as for my design which was the topic here...
AT drivers are a exotic guess, more or less pigs in a bag. Unfortunately they don't deliver with right to return them if they fail in measurements. And I sure don't want to invest 250 euro midwoofer that has performance of a 60 euro unit. Don't like that status value that high and personally I have pretty poor imagination to imagine them sound that much better...
W18EX is a strong candidate. Expensive but also tested in several places getting excellent results. Peerless has very good HDS serie which would reach the same with copper rings in their magnetsystem. Shame.
Accutons I don't know. Small shop just like AT. Make stuff by hand and with small tolerances price is astronomical. I just wonder why people still favor handmade things. With a robot you can install spider, upper suspension, cone and VC to chassis having below 0,1mm tolerances. Have to got very steady hand to achieve the same with human hands. And this in long period of time, year after year. And still machine stamps things in ratio of 10 people.
Jussi
Ok. If I return to things I do know...
If I settle for MT structure I'll make a compromisse in distorsion and efficiency which ultimately also limits usable SPL range. What I'd gain is more compact overall construction and no MTM problems...
And as for Audio Technology C-quenze or similar driver MT leaves nice chances to try out different crossover frequences and maybe more importantly different slopes.
AT structure looks decent. Inductance is very low indicating decent midrange distorsion, basket is nice and open, enough xmax for my needs and nice effieciency for a 7" midwoofer. But still not that much 100% proof facts. AT is one manufacturer not so widely used and so even subjective information is hard to find. And not so many people say bad things about their own designs that even have considerable amounts of money invested in them.
I think I'll try to scetch that MT construction a bit and see if I could get a acoustically operational and also visually nice looking compromisse made.
For woofers I haven't decided yet. With MT tops it's like 2-ways with stereosubs. Only subs have a bit higher crossover frequency compared to what's used to. Single PA 18" woofer in dipole using triangle shape flat baffle is one possibilty, then lurking out few others..
Jussi
If I settle for MT structure I'll make a compromisse in distorsion and efficiency which ultimately also limits usable SPL range. What I'd gain is more compact overall construction and no MTM problems...
And as for Audio Technology C-quenze or similar driver MT leaves nice chances to try out different crossover frequences and maybe more importantly different slopes.
AT structure looks decent. Inductance is very low indicating decent midrange distorsion, basket is nice and open, enough xmax for my needs and nice effieciency for a 7" midwoofer. But still not that much 100% proof facts. AT is one manufacturer not so widely used and so even subjective information is hard to find. And not so many people say bad things about their own designs that even have considerable amounts of money invested in them.
I think I'll try to scetch that MT construction a bit and see if I could get a acoustically operational and also visually nice looking compromisse made.
For woofers I haven't decided yet. With MT tops it's like 2-ways with stereosubs. Only subs have a bit higher crossover frequency compared to what's used to. Single PA 18" woofer in dipole using triangle shape flat baffle is one possibilty, then lurking out few others..
Jussi
Sorry.. been sick as a "dog" recently.. no doubt a "holiday" virus.
I'm not sure about "natural sound".. but "realistic sound" for me is a combination of factors that includes things like proper image size and placement, image density/palpability, dimensionally accurate hall sound (i.e. soundstage), reasonably flat freq. response at the listening position (though room abberations can be problematic), good dynamic contrast ability (i.e. ability to change spl's quickly), excellent "timing" (or the ability to start and decay in a manner consistent with the music and its venue), no obvious tonal problems, etc.. By default then I'd say that "un-natural sound" is anything that obscures any aspect of realism (just mentioned).
Typically the "blind" thing is there by default (at least for speakers I haven't heard b4) when walking into a dealer's room. In other words I don't pull off the grill cloths/look for a makers symbol until after listening to a specific speaker. I can of course see what type of amplification is being use though (so it isn't "blind"). For instance one of the speakers I listened to a few weeks past was the new version of the Audio Artistry Vivaldi. I had no idea what it was.. and in fact I had no idea what any of the components were except for the BEL amplifiers. Nor had I ever heard the music that was playing through it b4. Despite this it was rather apparent that:
1. It was operating as a dipole in midbass (..realistic placement of bass instruments relative to other images)
2. It was operating as a dipole in the midrange (..good side wall reflection rejection while having poor front wall reflection rejection which created a delayed and accentuated tonal behaviour in the upper midrange from the in-phase reflection)
3. The upper midrange and treble were using drivers with a low propagtion velocity and a high internal loss (..over damped, "dull" transients, virtual lack of decay, etc.) Certainly the amplifer was effecting decay negativly, but at the same time it was probably offering more reastic bass and didn't seem to harm image size when used with such drivers (as compared to the Vandersteen 5's I heard it with where imaging became "pin-point")
It was also apparent that the digital "front-end" (in this store) was considerably better than in any other room (..latter I found out it was the Meitner stuff.. so no surpise there)
Despite all this I do agree that "blind" is better, but you'll rarely find an instance where it is achievable.. moreover it often isn't neccesary in a broad contex (i.e. achiveing general impressions).
In any event.. getting back to the design.
A bi-radial horn (i.e. constant directivity horn) essentially does what your looking for (flat on and off-axis response to a certian degree with little output beyond the specified degree.. say 90 degrees or +/- 45 degrees). Unfortunetly like many things - its a lot easier to talk about than actually acomplish.
The problem I had with the design as I (believe) you have specified is that the off-axis response of even an a-typical, (better than normal), aperiodic vent lending the midrange drivers's response to a cardoid shape probably will not extend high enough to intersect with the beaming portion of the driver's upper passband. In other words I'd think you'll end up with a reduced off-axis response in the midrange near the crossover with the bass drivers (i.e. 200-600 Hz at best) but THEN (as the freq. response rises) the response starts behaving much more as an omni until the top of its passband where it starts becoming more directive again. Essentially then the continuous directivity your trying to achieve will, "I" think, not work with a cardoid.. HOWEVER, it probably WILL work with a dipole - and furthermore the rear reverse phase could be attenuated with several layers of absorbing material a foot or so away from the rear of the driver. EVEN THEN, your tweeter at cross-point will behave in an omni-directional behaviour UNLESS you have a wave-guide/horn to control directivity of the tweeter. (.. which of course you've stated it will have.)
IF the horn/wave guide is big enough you could always extend the response of the tweeter down to where the midrange becomes directive in an aperiodic cardoid box with steep crossovers (a'la amphion). Then you wouldn't have to worry about the diretivity of the midrange's response at the top of its passband - it would be a moot point (..so would cone break-up and upperlimit distortion). In fact if it were me, (assuming I wanted something approaching a constant directivity design), this is exactly what I would do.. Instead of worrying about a midrange driver at this point, I'd start trying to model horns for the Esotar and see where that "got" me. (There is HornResponse and Martin Kings worksheets for horn modeling.)
Just a thought..
If your to the point where you have the Horn that will extend down to at least the esotar's fs then I'd start looking for the "midrange" driver. IF you want to match eff. of the esotar with midrange (around 92 db?) AND allow for compensation for baffle loss you'll be looking at a driver (or combination of drivers) that is around AT LEAST 96 db at 200 Hz. Alternativly IF your bass drivers can extend high enough in freq. with little distortion (and you have a steep crossover), you might just as well "ditch" the midrange altogether. For instance I don't see any reason why you couldn't have a 4th order LR for the tweeter around 600 Hz and run the bass drivers in dipole up to this 4th order low-pass (perhaps 4 of the PE Dayton woofers in the 10 inch version per side obviously "assisted" actively). If however you wanted to continue with the cardoid approach with this basic design then I'd suggest the Supravox 215 GMF for this limited freq. range (..it would be a LOT more dynamic than the dipole approach).
Finally, as to why? (with the differences in drivers)..
Well I've given you two angles, 1 objective, the other subjective.. and tried to "tie" the two together at least to some degree to give a correlative value.
(1. Driver diaphram properties - objective: propagation velocity * rigidity vs. internal loss)
(2. Driver diaphram properties - subjective: summed listening experiences with subjective responses, though in "non"-riggorus settings)
The "non-returnable" thing is definitily a problem.. perhaps then you should try out drivers you can return and pass on the AT's?
EDIT: (..where are my manners?)
Happy Holidays!
I'm not sure about "natural sound".. but "realistic sound" for me is a combination of factors that includes things like proper image size and placement, image density/palpability, dimensionally accurate hall sound (i.e. soundstage), reasonably flat freq. response at the listening position (though room abberations can be problematic), good dynamic contrast ability (i.e. ability to change spl's quickly), excellent "timing" (or the ability to start and decay in a manner consistent with the music and its venue), no obvious tonal problems, etc.. By default then I'd say that "un-natural sound" is anything that obscures any aspect of realism (just mentioned).
Typically the "blind" thing is there by default (at least for speakers I haven't heard b4) when walking into a dealer's room. In other words I don't pull off the grill cloths/look for a makers symbol until after listening to a specific speaker. I can of course see what type of amplification is being use though (so it isn't "blind"). For instance one of the speakers I listened to a few weeks past was the new version of the Audio Artistry Vivaldi. I had no idea what it was.. and in fact I had no idea what any of the components were except for the BEL amplifiers. Nor had I ever heard the music that was playing through it b4. Despite this it was rather apparent that:
1. It was operating as a dipole in midbass (..realistic placement of bass instruments relative to other images)
2. It was operating as a dipole in the midrange (..good side wall reflection rejection while having poor front wall reflection rejection which created a delayed and accentuated tonal behaviour in the upper midrange from the in-phase reflection)
3. The upper midrange and treble were using drivers with a low propagtion velocity and a high internal loss (..over damped, "dull" transients, virtual lack of decay, etc.) Certainly the amplifer was effecting decay negativly, but at the same time it was probably offering more reastic bass and didn't seem to harm image size when used with such drivers (as compared to the Vandersteen 5's I heard it with where imaging became "pin-point")
It was also apparent that the digital "front-end" (in this store) was considerably better than in any other room (..latter I found out it was the Meitner stuff.. so no surpise there)
Despite all this I do agree that "blind" is better, but you'll rarely find an instance where it is achievable.. moreover it often isn't neccesary in a broad contex (i.e. achiveing general impressions).
In any event.. getting back to the design.
A bi-radial horn (i.e. constant directivity horn) essentially does what your looking for (flat on and off-axis response to a certian degree with little output beyond the specified degree.. say 90 degrees or +/- 45 degrees). Unfortunetly like many things - its a lot easier to talk about than actually acomplish.
The problem I had with the design as I (believe) you have specified is that the off-axis response of even an a-typical, (better than normal), aperiodic vent lending the midrange drivers's response to a cardoid shape probably will not extend high enough to intersect with the beaming portion of the driver's upper passband. In other words I'd think you'll end up with a reduced off-axis response in the midrange near the crossover with the bass drivers (i.e. 200-600 Hz at best) but THEN (as the freq. response rises) the response starts behaving much more as an omni until the top of its passband where it starts becoming more directive again. Essentially then the continuous directivity your trying to achieve will, "I" think, not work with a cardoid.. HOWEVER, it probably WILL work with a dipole - and furthermore the rear reverse phase could be attenuated with several layers of absorbing material a foot or so away from the rear of the driver. EVEN THEN, your tweeter at cross-point will behave in an omni-directional behaviour UNLESS you have a wave-guide/horn to control directivity of the tweeter. (.. which of course you've stated it will have.)
IF the horn/wave guide is big enough you could always extend the response of the tweeter down to where the midrange becomes directive in an aperiodic cardoid box with steep crossovers (a'la amphion). Then you wouldn't have to worry about the diretivity of the midrange's response at the top of its passband - it would be a moot point (..so would cone break-up and upperlimit distortion). In fact if it were me, (assuming I wanted something approaching a constant directivity design), this is exactly what I would do.. Instead of worrying about a midrange driver at this point, I'd start trying to model horns for the Esotar and see where that "got" me. (There is HornResponse and Martin Kings worksheets for horn modeling.)
Just a thought..
If your to the point where you have the Horn that will extend down to at least the esotar's fs then I'd start looking for the "midrange" driver. IF you want to match eff. of the esotar with midrange (around 92 db?) AND allow for compensation for baffle loss you'll be looking at a driver (or combination of drivers) that is around AT LEAST 96 db at 200 Hz. Alternativly IF your bass drivers can extend high enough in freq. with little distortion (and you have a steep crossover), you might just as well "ditch" the midrange altogether. For instance I don't see any reason why you couldn't have a 4th order LR for the tweeter around 600 Hz and run the bass drivers in dipole up to this 4th order low-pass (perhaps 4 of the PE Dayton woofers in the 10 inch version per side obviously "assisted" actively). If however you wanted to continue with the cardoid approach with this basic design then I'd suggest the Supravox 215 GMF for this limited freq. range (..it would be a LOT more dynamic than the dipole approach).
Finally, as to why? (with the differences in drivers)..
Well I've given you two angles, 1 objective, the other subjective.. and tried to "tie" the two together at least to some degree to give a correlative value.
(1. Driver diaphram properties - objective: propagation velocity * rigidity vs. internal loss)
(2. Driver diaphram properties - subjective: summed listening experiences with subjective responses, though in "non"-riggorus settings)
The "non-returnable" thing is definitily a problem.. perhaps then you should try out drivers you can return and pass on the AT's?
EDIT: (..where are my manners?)
Happy Holidays!
Scott,
Acoustic resistance cabinets work well while implementation is made well. They don't have discontinuous behavior at any frequence and if compared to dipole system they don't have a change of radiation pattern at some frequence. At least most dipole mid/woofer systems have standard front radiating tweeter and with or without waveguide this causes difference in radiation pattern while lower range drops off and tweeter takes over.
With a normal waveguide I think Esotar is good down to 1800Hz or so with steep crossover. I don't want to strain tweeters so 3-way it is. And it's also a difficult to achieve decent bass while woofers would run up to 600Hz or so. In this aspect maybe a 8" midrange running 120-1800Hz for example in MT config fits the bill pretty well. Below 120Hz it's easy to use heavy construction dipole woofers and place them along sidewalls.
As for cardioid/hypercardioid systems they are pretty room immune systems. It's pretty same where you put them. They play rather decently even in corner positioning but of cource have their slight advantage over free positioning just like everything else. But this leaves top 2-way setups (both MTM or MT) easy to place in standard listening triangle and then have separated dipole woofers that can be turned and moved a bit to get best possible bass definition.
I don't know any distributor or manufacturer that would lend a pair to normal streetjockey and have them back after testing. With company brand behind me I could do it but I don't have a company to use for such task...
Without financial limitations I could order all, measure and test them and throw into woods the crappy ones but unfortunately I can't do this. 🙂
Jussi
Acoustic resistance cabinets work well while implementation is made well. They don't have discontinuous behavior at any frequence and if compared to dipole system they don't have a change of radiation pattern at some frequence. At least most dipole mid/woofer systems have standard front radiating tweeter and with or without waveguide this causes difference in radiation pattern while lower range drops off and tweeter takes over.
With a normal waveguide I think Esotar is good down to 1800Hz or so with steep crossover. I don't want to strain tweeters so 3-way it is. And it's also a difficult to achieve decent bass while woofers would run up to 600Hz or so. In this aspect maybe a 8" midrange running 120-1800Hz for example in MT config fits the bill pretty well. Below 120Hz it's easy to use heavy construction dipole woofers and place them along sidewalls.
As for cardioid/hypercardioid systems they are pretty room immune systems. It's pretty same where you put them. They play rather decently even in corner positioning but of cource have their slight advantage over free positioning just like everything else. But this leaves top 2-way setups (both MTM or MT) easy to place in standard listening triangle and then have separated dipole woofers that can be turned and moved a bit to get best possible bass definition.
I don't know any distributor or manufacturer that would lend a pair to normal streetjockey and have them back after testing. With company brand behind me I could do it but I don't have a company to use for such task...
Without financial limitations I could order all, measure and test them and throw into woods the crappy ones but unfortunately I can't do this. 🙂
Jussi
Jussi said:
OK.. so no problems with directivity vs. freq., however - if thats the case then why consider a driver that will become more directive near 1-3 kHz? In otherwords shouldn't a 5 inch driver in a box with an aperiodic vent display the same directivity as an 8 inch driver in a similar configuration? (..or am I missing something again?)
Well with the low of 1.5 kHz (and mid driver need), your back to the eff. problem unless you pad down the tweeter. Are you willing to sacrifice eff.? If not you will greatly limit your choices, and perhaps make the task of choosing the appropiate driver a bit easier. I wish I could help you more and say that the custom AT's will be better than Seas.. but I really don't know. What I can say though is that it has been my experience that the higher the eff. of the driver the greater the subjective quality of "dynamics". So IMO, all else isn't equal - I'd choose the single driver higher eff. design over the 2 driver paralleled design (particularly for such a limited bandwidth).
I don't know about "abroad" purchasing.. but in the States we typically have a 30 day return without question (so long as the returned item is in essentially the same condition as when it arrived). In otherwords buy a driver you think will be good and if it doesn't sound good or measure up then send the thing back for a full refund (minus the shipping costs and a re-stocking fee.. something to always check on). I wasn't suggesting the purchase everything and "chuck" method..
😉 Well, good luck with it (..it looks like I'll have to research more into cardiod implementation.. because from what your saying I'm more than a little off in my understanding)
But its all good, I love learning about various loudspeaker design methods.😀arggg.. 20 bucks I think I'll have to pass on this bit of education..
http://www.aes.org/e-lib/browse.cfm?elib=5231
http://www.aes.org/e-lib/browse.cfm?elib=5231
Resistance cabinet operates well with certain range. Range is about 2-3 octaves wide depending on several variables. And using 5" midrange is fine but rises up bottom end extension. Problem isn't even directivity but drivers efficiency runs out while it needs more and more EQ. With larger driver resistance cabinets construction can be a little different and made to operate at lower frequences while larger cone takes over lower than smaller so resistance doesn't have to reach that high top end.
Differences in efficiency isn't a large problem. I'm using full active construction and each drivers full efficiency is used. But still I don't like 80dB or so efficiency. It kills some dynamics like you said and limits usable SPL range before driver turns from dynamic driver to a lightbulb. Normal home hifidrivers can handle about 30-40W continuous power before running into thermical compression. And that 100-150W limit manufacturers claim is the point where coils wire burns out and driver is ruined. But they sure show some stress problems way before that. Dynamic punches can of cource be more powerful.
Shame we don't have such 30 day return policy. Distributors just deliver, get the money and that's it. Of cource they cover flaws in specific pieces if it's caused by error in manufacturing but I don't think its valid if drivers whole production range is "flaw" in some aspect.
I don't know about AT. They claim to have "very low distorsion" and I don't know clear reasons why they couldn't be. I should propably just ask them, can I return my investment if they don't fit into certain window in THD measurement.
Jussi
Differences in efficiency isn't a large problem. I'm using full active construction and each drivers full efficiency is used. But still I don't like 80dB or so efficiency. It kills some dynamics like you said and limits usable SPL range before driver turns from dynamic driver to a lightbulb. Normal home hifidrivers can handle about 30-40W continuous power before running into thermical compression. And that 100-150W limit manufacturers claim is the point where coils wire burns out and driver is ruined. But they sure show some stress problems way before that. Dynamic punches can of cource be more powerful.
Shame we don't have such 30 day return policy. Distributors just deliver, get the money and that's it. Of cource they cover flaws in specific pieces if it's caused by error in manufacturing but I don't think its valid if drivers whole production range is "flaw" in some aspect.
I don't know about AT. They claim to have "very low distorsion" and I don't know clear reasons why they couldn't be. I should propably just ask them, can I return my investment if they don't fit into certain window in THD measurement.
Jussi
- Status
- Not open for further replies.