sorry what I meant was the region of 4-10khz is where a peak or rise can cause 'shout' in the vocals, which I hate.
The Satori speakers, they look nice altho quite expensive and there's a weird dip around 1khz. The Peerless Vifa NE range is possibly better value and similar in looks, and I hear the sound is v good too, across the range.
The Satori speakers, they look nice altho quite expensive and there's a weird dip around 1khz. The Peerless Vifa NE range is possibly better value and similar in looks, and I hear the sound is v good too, across the range.
Last edited:
old Vifa mid woofer are used in the very musical OBX living voice with their "elastic box" (thin wood not thick medium wood).
Mister Linkwitz measured a lot of distorsion with some Vifa's but a lot of good speakers are made wit it ! An another forgotten non expensive brand. Their Satory clone look great ! Basic Sb acoustics are known to sing good too... All these are not simple... it's like the graal in a Monthy Python movie ! Which are making the noise : the little faster white rabbit or the heavy aluminium Knight ?
Mister Linkwitz measured a lot of distorsion with some Vifa's but a lot of good speakers are made wit it ! An another forgotten non expensive brand. Their Satory clone look great ! Basic Sb acoustics are known to sing good too... All these are not simple... it's like the graal in a Monthy Python movie ! Which are making the noise : the little faster white rabbit or the heavy aluminium Knight ?
BBB I don't think it's motor related distortion but rather cone breakup modes that induce such issue we are describing. Tall order harmonics are produced by the motor (in general) and do not have such an effect on even order harmonics. Not from any measurement I've made or seen at least
Maybe it wasn't clear what I mean.
Imagine a device which generates 2% F2 at 1 kHz. Now apply an EQ after this device, with a peak of 10 dB at 2 kHz. The F2 distortion means that the first device will generate a tone at 2*1 kHz = 2 kHz. The filter amplifies this tone, and therefore you have higher distortion.
It is the same with speakers. The motor is the non-linear device, and the cone is the filter.
Just want to add that a person would be better off avoiding reviews of speaker systems or speaker drivers by any kind of magazines or anyone trying to sell you anything. Reviewers are only hired and paid when they write favorable reviews. It's usually a dollar driven review.
I consider the frequency response graph one of the most important things to evaluate drivers with. Any abrupt variations are to be avoided either electronically (crossover freq choice) or altogether. Many drivers of any brand have horrible frequency response curves. End user reviews would be better if most end users weren't corrupted by reading these dollar driven reviews and beautiful graphics in many ads or reviews.
Another area that should get more attention is dispersion. At higher frequencies the drivers get beamy (less diffraction), and you don't want to have an abrupt change in dispersion at a high crossover frequency (roughly 800HZ and above), or the room power response becomes a problem, and the "sweet spot" gets small.
Having said that, I'm tempted to argue that reduced dispersion could increase imaging clarity if it were only in the upper midrange (800HZ- 6kHZ), because it would reduce the effects of room acoustic reflections in the frequency range where we determine image location apparently exclusively by amplitude comparisons, left to right. You don't want room acoustics messing with that. What I see in many speaker systems is a woofer going up to about 2-3kHZ, then an abrupt change in dispersion from narrow to wide where a small diameter tweeter takes over. Not good.
Right now I'm thinking a great speaker would be the Peerless Nomes 6.5 inch woofer going up to 1kHZ, where a Peerless TG9FD10-08 (or TC9FD-18-08 if you need the cone to be black) 3 inch would take over from 1kHZ - 7kHZ (would require a sub-enclosure), and then above that a Fountek 1.5 inch ribbon tweeter (Fountek ribbons are rated very high by Zaph). These drivers used in this way would only need 1 pole crossovers (higher order would be better of coarse). And this puts the invariably damaging crossover points where the ear-brain mechanism is least likely to see serious damage to the imaging capabilities.
Imaging below 800HZ is all about timing comparisons left to right. Imaging from about 2kHZ to 6kHZ is all about amplitude comparisons left to right. Above 6kHZ the top dog scientists have pointed out that the ear-brain mechanism hears that high treble in a diffuse way, and is more tuning in on height than left to right. It becomes a product of how 6kHZ and above interacts with the shape of the "pinea" or ear on a given head (and they're all a little different).
Below 80 HZ one might want to employ a separate woofer. Personally I think distributed woofers are the way to go. Typical listening room acoustics often mess up the frequency response at the listening position pretty bad below about 250HZ. It gets to where the wavelengths fit into the dimensions of the room, creating severe peaks and dips for any given location in the room. Linkwitz tries to minimize this problem by using a dipole woofer that will have minimal output at it's sides, thereby stimulating fewer "room modes". Roger Russel, formerly of McIntosh, presently produces a vertical line array speaker system that distributes the bass evenly floor to ceiling, and in two locations horizontally (using 25 of the 3 inch Peerless TG9FD10 drivers mentioned above per side), and claims to have no boominess in that frequency range at all. I think it was Infinity that has a separate woofer tower about 5 feet tall with maybe four 8 inch woofers in it, to reduce this room acoustics problem, but then you have four tower cabinets to get two channels. The latest top of the line McIntosh tower system has 6-12 inch woofers aligned vertically on each side, with the higher band drivers right in front of the woofers, thereby avoiding the need for four tower cabinets to get two channels of sound.
Sorry to be a bit off subject here. Yes metal cones... I know they sound very good in a Hartke brand bass guitar amp. Especially when more than one string is plucked at a time.
I consider the frequency response graph one of the most important things to evaluate drivers with. Any abrupt variations are to be avoided either electronically (crossover freq choice) or altogether. Many drivers of any brand have horrible frequency response curves. End user reviews would be better if most end users weren't corrupted by reading these dollar driven reviews and beautiful graphics in many ads or reviews.
Another area that should get more attention is dispersion. At higher frequencies the drivers get beamy (less diffraction), and you don't want to have an abrupt change in dispersion at a high crossover frequency (roughly 800HZ and above), or the room power response becomes a problem, and the "sweet spot" gets small.
Having said that, I'm tempted to argue that reduced dispersion could increase imaging clarity if it were only in the upper midrange (800HZ- 6kHZ), because it would reduce the effects of room acoustic reflections in the frequency range where we determine image location apparently exclusively by amplitude comparisons, left to right. You don't want room acoustics messing with that. What I see in many speaker systems is a woofer going up to about 2-3kHZ, then an abrupt change in dispersion from narrow to wide where a small diameter tweeter takes over. Not good.
Right now I'm thinking a great speaker would be the Peerless Nomes 6.5 inch woofer going up to 1kHZ, where a Peerless TG9FD10-08 (or TC9FD-18-08 if you need the cone to be black) 3 inch would take over from 1kHZ - 7kHZ (would require a sub-enclosure), and then above that a Fountek 1.5 inch ribbon tweeter (Fountek ribbons are rated very high by Zaph). These drivers used in this way would only need 1 pole crossovers (higher order would be better of coarse). And this puts the invariably damaging crossover points where the ear-brain mechanism is least likely to see serious damage to the imaging capabilities.
Imaging below 800HZ is all about timing comparisons left to right. Imaging from about 2kHZ to 6kHZ is all about amplitude comparisons left to right. Above 6kHZ the top dog scientists have pointed out that the ear-brain mechanism hears that high treble in a diffuse way, and is more tuning in on height than left to right. It becomes a product of how 6kHZ and above interacts with the shape of the "pinea" or ear on a given head (and they're all a little different).
Below 80 HZ one might want to employ a separate woofer. Personally I think distributed woofers are the way to go. Typical listening room acoustics often mess up the frequency response at the listening position pretty bad below about 250HZ. It gets to where the wavelengths fit into the dimensions of the room, creating severe peaks and dips for any given location in the room. Linkwitz tries to minimize this problem by using a dipole woofer that will have minimal output at it's sides, thereby stimulating fewer "room modes". Roger Russel, formerly of McIntosh, presently produces a vertical line array speaker system that distributes the bass evenly floor to ceiling, and in two locations horizontally (using 25 of the 3 inch Peerless TG9FD10 drivers mentioned above per side), and claims to have no boominess in that frequency range at all. I think it was Infinity that has a separate woofer tower about 5 feet tall with maybe four 8 inch woofers in it, to reduce this room acoustics problem, but then you have four tower cabinets to get two channels. The latest top of the line McIntosh tower system has 6-12 inch woofers aligned vertically on each side, with the higher band drivers right in front of the woofers, thereby avoiding the need for four tower cabinets to get two channels of sound.
Sorry to be a bit off subject here. Yes metal cones... I know they sound very good in a Hartke brand bass guitar amp. Especially when more than one string is plucked at a time.
Last edited:
EJ Jordan ? So JXR 6 HD the best speaker in the world ? flat curve and wide band, 4" but metal cone...
Bob, what about the two way approach of StigErik with B&G Planar , cut at 300 and big multiple 18" for bass ?
Is a flat frequency curve between 100 Hz et 20 Khz the way to go ?
About review speaker, where are we looking for ? Zaph ? Diya fellow ?
Bob, what about the two way approach of StigErik with B&G Planar , cut at 300 and big multiple 18" for bass ?
Is a flat frequency curve between 100 Hz et 20 Khz the way to go ?
About review speaker, where are we looking for ? Zaph ? Diya fellow ?
The Nomex material seems to offer the benefits of paper cones (less severe break up and high sensitivity) without the 'paper' sound, would that be right?
EDIT the Jordans are v good indeed, incredibly detailed -the only thing I wd say is they can sound slightly cold and clinical in the wrong system. I prefer the similar Bandor 50's.
EDIT the Jordans are v good indeed, incredibly detailed -the only thing I wd say is they can sound slightly cold and clinical in the wrong system. I prefer the similar Bandor 50's.
Last edited:
The B&G planars are very good, but way out of my price range. Zaph tested the B&G ribbon tweeter against a Fountek ribbon and it sounded to me like it's a toss up which one is better above 4kHZ. Fountek is about $100 and B&G is about $600 if I remember correctly. He tested the five inch Fountek. Personally I'd go with the 1.5 inch since I would use it only above 6kHZ, where maximum dispersion on both axis seems ideal. The 1.5 inch Fountek is $55 here in the U.S.
I'm not familiar with the StigErik.
On the 18 inch woofer, I don't have room for that where I live. I much prefer the sound of smaller woofers (5 - 12 inch) with the approach being distributed emission vertically. I'm a huge fan of closed box woofers with active EQ making them flat to whatever frequency I want (20 - 30HZ usually). The bigger the cone, the more likely the cone itself will have other various resonances. If I was building a system for outdoor rock concerts I might consider an 18 inch, otherwise no.
I generally always design for flat frequency response initially, and then use my 4 section Baxandall tone circuit to make that actually sound good. Recordings often have bad EQ, listening room acoustics will mess up EQ, there's dispersion issues, and then there's the Fletcher-Munson ear curves. Flat response speakers are very rarely good sounding at the listening position. People who don't have good tone controls are misled. People who spend a lot on speaker wire are misled as well. The connectors can be a source of distortion (I use gold alloy plated banana connectors soldered to the wires), but any 16 AWG AC line cord from a hardware store is fine for up to 20 feet with an 8 ohm speaker in my opinion. If the system is 4 ohms, go with 14AWG. You'll be very hard pressed to tell the difference. We put a 1 ohm resistor in series with a speaker once to see if it made a difference. We lost approx. 10% of the power across it (with an 8 ohm speaker) but it actually sounded better. More relaxed sounding...
The reason I cite Zaph is because he isn't trying to sell anything. He's a hard core hobbyist who just wants to share his findings. He seems to be very sharp and very experienced. Linkwitz is another very sharp guy, but he is trying to sell his systems so I take what he says with a grain of salt. Every experiment (project) is riddled with variables that are rarely if ever addressed fully and properly. The trick is to extract what's real from what's more often a poorly founded opinion, or what's just plain dollar driven sales talk.
I've never heard the Jordans. I would caution however that many metal diaphram drivers sound spectacular, but can be fatiguing to listen to over time.
I'm not familiar with the StigErik.
On the 18 inch woofer, I don't have room for that where I live. I much prefer the sound of smaller woofers (5 - 12 inch) with the approach being distributed emission vertically. I'm a huge fan of closed box woofers with active EQ making them flat to whatever frequency I want (20 - 30HZ usually). The bigger the cone, the more likely the cone itself will have other various resonances. If I was building a system for outdoor rock concerts I might consider an 18 inch, otherwise no.
I generally always design for flat frequency response initially, and then use my 4 section Baxandall tone circuit to make that actually sound good. Recordings often have bad EQ, listening room acoustics will mess up EQ, there's dispersion issues, and then there's the Fletcher-Munson ear curves. Flat response speakers are very rarely good sounding at the listening position. People who don't have good tone controls are misled. People who spend a lot on speaker wire are misled as well. The connectors can be a source of distortion (I use gold alloy plated banana connectors soldered to the wires), but any 16 AWG AC line cord from a hardware store is fine for up to 20 feet with an 8 ohm speaker in my opinion. If the system is 4 ohms, go with 14AWG. You'll be very hard pressed to tell the difference. We put a 1 ohm resistor in series with a speaker once to see if it made a difference. We lost approx. 10% of the power across it (with an 8 ohm speaker) but it actually sounded better. More relaxed sounding...
The reason I cite Zaph is because he isn't trying to sell anything. He's a hard core hobbyist who just wants to share his findings. He seems to be very sharp and very experienced. Linkwitz is another very sharp guy, but he is trying to sell his systems so I take what he says with a grain of salt. Every experiment (project) is riddled with variables that are rarely if ever addressed fully and properly. The trick is to extract what's real from what's more often a poorly founded opinion, or what's just plain dollar driven sales talk.
I've never heard the Jordans. I would caution however that many metal diaphram drivers sound spectacular, but can be fatiguing to listen to over time.
Last edited:
Bob, I'll be using these Peerless drivers
Peerless SDS-160F25PR01-08 6-1/2" Paper Cone Woofer Speaker 264-1146
But the 4 used for the MTM section are modified, Tpole with faraday shielding to reduce the tall order harmonics.
Behave quite nicely (would say better than HDS Nomex) despite old school motors meaning the moving bits were designed very well. While I don't have distortion numbers on them (deleted errr) but will redo with before and after results. I have a dozen to play with. In listening tests with non modified drivers active with some Airborne rt4001 amt's, X'd @2.2k LR4 sound wonderful. No stress in the vocals, natural sounding, crisp and clear without any disconnect.
Peerless SDS-160F25PR01-08 6-1/2" Paper Cone Woofer Speaker 264-1146
But the 4 used for the MTM section are modified, Tpole with faraday shielding to reduce the tall order harmonics.
Behave quite nicely (would say better than HDS Nomex) despite old school motors meaning the moving bits were designed very well. While I don't have distortion numbers on them (deleted errr) but will redo with before and after results. I have a dozen to play with. In listening tests with non modified drivers active with some Airborne rt4001 amt's, X'd @2.2k LR4 sound wonderful. No stress in the vocals, natural sounding, crisp and clear without any disconnect.
Thank you for the inputs Bob,
A ribbon could be hard to match with a cone mid... have you an idea here ?
I was talking about the 600 $ B&G because it can be crossed between 250... and more but ned a trap notch at 5000 hz. So just 2 units and no box for the ribbon... the total owner cost is reduced.
About the price, two ways : if you buy many speakers to find the good one the price is the same with buying the good one first ! But too much opinions even with the better mentioned: SB, Peereless, ScanSpeak...
heeek The SB satory midrange is 300$ if I remember... all these are expensive for beginners who fails in design!
The another ways is that speakers are your hobby and many here buy a lot. i can't but would like find a good three ways design equal to a 5000 to 10000 $ commercial brand ! But If I spend equal money and 5 years by trying to do all designs it's good for the hobby ( and I like that as everyone here without commercial interests). Is zaph earning money when cited in part express kits near the ones of S. Linkwitz ? Or is it to less money to worth with ?
I said that because I spent the last 16 years to listen to a too brighty "High End" speakers and when I switch with an old Kef (the one with MTM semi Apolitto filter)....it was like the release of my ears !
I like the closed box for bass : I have two speakers with symetrical closed box.
A ribbon could be hard to match with a cone mid... have you an idea here ?
I was talking about the 600 $ B&G because it can be crossed between 250... and more but ned a trap notch at 5000 hz. So just 2 units and no box for the ribbon... the total owner cost is reduced.
About the price, two ways : if you buy many speakers to find the good one the price is the same with buying the good one first ! But too much opinions even with the better mentioned: SB, Peereless, ScanSpeak...
heeek The SB satory midrange is 300$ if I remember... all these are expensive for beginners who fails in design!
The another ways is that speakers are your hobby and many here buy a lot. i can't but would like find a good three ways design equal to a 5000 to 10000 $ commercial brand ! But If I spend equal money and 5 years by trying to do all designs it's good for the hobby ( and I like that as everyone here without commercial interests). Is zaph earning money when cited in part express kits near the ones of S. Linkwitz ? Or is it to less money to worth with ?
I said that because I spent the last 16 years to listen to a too brighty "High End" speakers and when I switch with an old Kef (the one with MTM semi Apolitto filter)....it was like the release of my ears !
I like the closed box for bass : I have two speakers with symetrical closed box.
Last edited:
First of all, the glass fiber cone on the 3 inch Peerless TG9FD10-08 driver ($22 at Madisound) is allegedly pretty fast according to Roger Russel. Also, having the crossover point at 6 or 7kHZ has the feature of being at the transition of where the ear-brain mechanism switches from trying to locate a sound on the left/right axis based on amplitude comparisons, to where it doesn't do that anymore, so I'm guessing that is a frequency where you would have less damage done by a driver change.
Also, the thing about cone drivers is that the cone acts like a waveguide at the highest frequencies, and virtually all of the acoustic energy radiates from the coil itself and any dust cap, so it's not really a 3 inch directional dispersion at the frequency where the ribbon would take over. More like one inch in this case.
From what I've heard and read, it's a rare ribbon that sounds good in the lower frequencies. Below 1kHZ you might start needing quite a bit of EQ to compensate for baffle step loss, and if the diaphram is small it could run out of dynamic range sooner than expected. It may be more about frequency modulation distortion than power rating in the lower hundreds of HZ.
I'm using that 3 inch Peerless TG9FD10-08 driver I mentioned above in my soundbar project (five of them actually), taking it down to 150HZ (4 pole active crossover). I found that I needed quite a bit of EQ to get it to do it's job (about 15 dB up at 150HZ relative to 2kHZ, to measure acoustically flat at the listening chair). A physically small ribbon might need something similar. Any baffle step EQ done passively (after the poweramp) will reduce the system efficiency significantly. Sometimes considerably. A system that would have been 90dB SPL might then be 80dB SPL (if 10 dB of boost at 100HZ relative to 1kHZ for example). It's like a 10 times loss in poweramp wattage.
In the ribbon tweeter catagory, zaph said not to use any of them below 4kHZ, (he tested about 8 different ones) but the ones I've heard do sound lush and crystal clear at the high end, so I figure bringing one in at 7kHZ should work well. Ribbons can't go as low in freq., and they are usually very directional on one axis, which I don't like above 6kHZ. So I go with the shortest good one I can afford (Fountek 1.5 inch).
One of the things I love about those 3 inch Peerless drivers is that they are so flat and usable from lower hundreds of HZ all the way up to 15kHZ, so used as a midrange you never have to use them right at the edge of their usable range, where various distortions get worse (ringing at the low end and slewing at the high end). Plus the half roll suspension isn't huge, so very little cavity effect, as opposed to for example the Dayton 3 inch drivers with the huge Xmax. The Dayton doesn't measure nearly as flat - I think that's why.
Use of too much Xmax implies significant frequency modulation distortion, so maybe a 3 inch should only be taken down just so far in frequency - maybe an Xmax beyond about 2mm is not really what you want to be using with a 3 inch driver. Below about 400HZ, I think it's wise to be using at least the cone surface area of a 6.5 inch driver (or two five inch drivers). This of coarse depends on how loud you want it to go and how big your listening room is.
There's one scenario where the hobbyist tries to save money by buying cheap drivers that he eventually decides aren't good enough, and after several tries he has spent enough to have bought the really good one. There's also the scenario where the hobbyist buys what he believes are top notch drivers, only to find that they are no better than what he could have gotten for about 1/7th the cost. And that these very expensive drivers actually have some surprising shortcomings in their performance. I've done both, and witnessed others doing the same.
There's also a third scenario where the guy eventually figures out that most of the drivers he picked, cheap and costly, were actually quite good when used right, and he just didn't do the rest of the project as right as he thought. Baffle step loss can be a huge issue. Even above 1kHZ. Published graphs don't take this into account at all. Driver are measured mouted on a large wall in an anechoic chamber. Not necessarily real world.
Where a driver rolls off mechanically can be where it puts out a significantly distorted waveform, made most visable (perhaps) by use of a tone burst in a bandwidth limited envelop.
Every speaker project I've ever done (probably about 35 over the years) has been a learning experience. I hope by sharing this I can help others shorten their learning experience and do good work for less $$. Sorry to be so long winded here.
Last edited:
You're welcome Bob, we are not on tweeter and free to share... sure it's help.Don't be worry to be too long because here you have very informative input and I can be worry myself to be too long because of my lake of skill to write english (thank you to support that). I don't think it's Off topic here because behind we have in mind the sibilance of aluminium cone.
What about the bump of sensibility we haerd near 3500/4000 Hz (seen at 90 db: a good level most of the time in our rooms, the fletcher curve change with db) seen in the Fletcher curve (ears here do like a band pass between 2500 and 5500? Is it like the 6000/7000 Hz which is the re beginning of flat level but just after the decrease of the 3500/4000 bump ?
Agree about purchasing philosophy, a simple flower given at the right moment to the right lady may your life sing (and so do I... probably about 35 over the years too... sorry french humour I Joke !)
Thanks again
Last edited:
Interesting.
I wonder what besides the cone color is different compared to the one SB sells to Harman for the Revel Performa3 line.
An externally hosted image should be here but it was not working when we last tested it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.