One obvious disadvantage of most full range drivers is their lumpy FR. An extreme example would be Lowther drivers. One could reduce these FR anomalies with say notch filters. On the face of it, there would seem to be an obvious benefit.
Yet I have seen positive reviews of speakers using full range drivers, which have obvious FR unevenness. And when a FR correcting filter is added, the reviewer has opined that something has been lost, and the speaker is simply not as pleasing to listen to. Usually the losses are expressed in terms like " immediacy" or more abstractly "naturalness" " musicality". They prefer the speaker unfiltered, warts and all.
What are the views of members here?
Yet I have seen positive reviews of speakers using full range drivers, which have obvious FR unevenness. And when a FR correcting filter is added, the reviewer has opined that something has been lost, and the speaker is simply not as pleasing to listen to. Usually the losses are expressed in terms like " immediacy" or more abstractly "naturalness" " musicality". They prefer the speaker unfiltered, warts and all.
What are the views of members here?
"Reviewers" are rarely worth paying attention to.
That said, I have always found Lowthers to be colored since the first time I heard them (thank you, Nelson!) but they can be fun in small doses. However, I heard a setup by Lowther USA a few months ago that was far and away the best I've heard them sound- though limited in the type of music they can do well with. I have no idea if they were filtered or not, cherry-picked or not, or how the frequency response was different than their usual upper-midrange resonant "crispy" quality.
Member
Joined 2009
Paid Member
My experience with ragged treble was the now obsolete Fostex FE127. It was intolerable to my ears. I have some hyperacusis that sensitizes me and perhaps it is a common afliction that results in a difference of opinion between different listeners. The FE127 was very popular. But it had some nasty resonances up hi. I tried filters - they just `sucked the life out of the driver`.
I also tried the EL-70, it`s a bit ragged but has a falling treble response - it doesn`t trigger off any objections by my ears but the treble doesn`t sound quite right all the same.
I also recently tried the Mark Audio CHN70. It sounds smooth to my ears. It also sounds a bit boring and I find this size of driver too small to give me the balanced sound I prefer. The EL-70 being the only exception I`m aware of where bass output is good for this size of driver. An EL-70 replacement with better treble would set the standard for FR drivers no doubt in my mind.
I built speakers with Mark Audio 10.3 drivers. These are simply amazing. They have a smooth clear and clean sound with good dispersion and stunning imaging. No filters required or desired. But they need a good `box`to pull bass out of them.
My main use driver is the Audio Nirvana 15. It has a superb clean detailed sound even though the response shows a bit of uneven treble. I have not used any filters with it. I have compared it with Mark Audio 10.3 and it compares well - the AN15 has more bass, a fuller more balanced sound overall but I have to accept that it doesn`t match the 10.3 in terms of treble coherence.
I also tried the EL-70, it`s a bit ragged but has a falling treble response - it doesn`t trigger off any objections by my ears but the treble doesn`t sound quite right all the same.
I also recently tried the Mark Audio CHN70. It sounds smooth to my ears. It also sounds a bit boring and I find this size of driver too small to give me the balanced sound I prefer. The EL-70 being the only exception I`m aware of where bass output is good for this size of driver. An EL-70 replacement with better treble would set the standard for FR drivers no doubt in my mind.
I built speakers with Mark Audio 10.3 drivers. These are simply amazing. They have a smooth clear and clean sound with good dispersion and stunning imaging. No filters required or desired. But they need a good `box`to pull bass out of them.
My main use driver is the Audio Nirvana 15. It has a superb clean detailed sound even though the response shows a bit of uneven treble. I have not used any filters with it. I have compared it with Mark Audio 10.3 and it compares well - the AN15 has more bass, a fuller more balanced sound overall but I have to accept that it doesn`t match the 10.3 in terms of treble coherence.
Last edited:
One of John DeVore's more interesting ideas is never to use resistors in a crossover. This is easier if you have control of manufacture and choice of drivers, so no level matching is needed, of course. 
https://www.youtube.com/watch?v=xyjF1j3xlWY
It is also one of those principles that purely reactive filters really only affect things around the crossover point. The rest of the range is all-pass.
Trouble with all filters and attenuators and notches is that they muck up the amplifier source impedance that the driver sees. As John says, even if you take impedance high, you are still giving the amp more circuit to drive. This results in a loss of immediacy. So less is often more.
https://www.youtube.com/watch?v=xyjF1j3xlWY
It is also one of those principles that purely reactive filters really only affect things around the crossover point. The rest of the range is all-pass.
Trouble with all filters and attenuators and notches is that they muck up the amplifier source impedance that the driver sees. As John says, even if you take impedance high, you are still giving the amp more circuit to drive. This results in a loss of immediacy. So less is often more.
Last edited:
For some of the hard cones I've worked with, the had big peak over a narrow radiation angle in front, but off axis it either wasn't there or much much lower in level. These peaks also usually occur in the upper break up frequencies where the driver has suppressed output even moderately (30+ deg ) off axis.
To my ears the result was that when listening close and on axis, such as a desktop application, the notch helps. When listening further back and especially in a room that has allot of contribution, making the peak flat on axis did dull the sound
That's the big challenge with using one driver over a very extended range. It's not always easy to find one that responds to eq on axis without making a bit of a mess of the off axis
To my ears the result was that when listening close and on axis, such as a desktop application, the notch helps. When listening further back and especially in a room that has allot of contribution, making the peak flat on axis did dull the sound
That's the big challenge with using one driver over a very extended range. It's not always easy to find one that responds to eq on axis without making a bit of a mess of the off axis
Regarding the first statement, darnit, I resemble that remark!
But yeah, lowthers, used well, can be very good at specific things. Nearfield, they gain even more flexibility and can be quite lovely, particularly with subs and/or supertweeters.
It's not for me, but some of the best small-scale music sound I've heard has been from lowthers.
I had the same experience with Feastrex D5nf. It had several 20dB swings through the middle decade of FR. On some recordings it sounded incredibly good, but on most it was not an accurate tone. I tried on a few occasions using different methods to apply EQ to flatten the FR: DEQX, manual broadcast analog PEQ box, Rane DSP, but each time the EQ stole the life out of the music and it sounded better without. I noticed that with EQ it sounded similar to passive hifi speakers with steep crossover filters and lots of EQ applied. Dead, boring, sounded like a normal speaker. I heard the same effect when on other multiway active projects I changed electronic crossover slopes from low order to higher order. I also heard speakers with quasi linear phase like Danley, 1st order passives like Gemme Katana and 2nd order passive 2ways like my current Kairos, and I loved the openness, the natural tone, the relaxed openness they offer, just like FR drivers. I came to believe anecdotally that flattish phase response is important to me, and maybe adding all that electronic correction added phase distortion that ruined the Feastrex "thing." I'm not sure the cause, but that's my hunch at the moment.
Au contraire... I also remember that even the FIR linear phase filters of DEQX and three DEQX experts collaborating for a couple hours could not cure the problems of the Feastrex, so I don't know for sure. Maybe the huge FR swings were too much even for DEQX. Maybe DEQX "aint all that." I'd like to try FR again, thankfully now I know the importance of measurements and can be an informed shopper. There are many very good choices now. A dash of EQ is fine, but too much kills the golden goose.
Au contraire... I also remember that even the FIR linear phase filters of DEQX and three DEQX experts collaborating for a couple hours could not cure the problems of the Feastrex, so I don't know for sure. Maybe the huge FR swings were too much even for DEQX. Maybe DEQX "aint all that." I'd like to try FR again, thankfully now I know the importance of measurements and can be an informed shopper. There are many very good choices now. A dash of EQ is fine, but too much kills the golden goose.
I have had Lowthers, which I put in cabinets with the Lowthers pointing upwards. If a deflector (8"x 6" tile) was positioned at 45 degrees with the back edge just inside the further edge of the cone, the peak at 2-2.5k was almost eliminated. I imagine this is to do with direct/ indirect path length partial cancellation. To my ears it got rid of the peak without losing the life of the sound.
Hi,
It sounds interesting and similar to Audiovector Auditorium (for the upward unit). What if there's no deflector/reflector nearby and just firing up?
And I can't help thinking of the Jordan Delay Line.
Thanks for sharing
So, supposedly the 2-2.5k peak is the most severe breakup so it's splashing everywhere. And how the top HF is reflected from ceiling depends on the room acoustics when there's no obstacle nearby.
I guess, by the dimension of the deflector related to the dispersion, there's only a (small) part of the 2.5k peak reflected to the listener, while most top HF is sent forward because it's beaming into a small angle which is pretty much covered by the deflector.
It seems there's a lot of fun stuff to play with the size/angle/position of the deflector/reflector. Maybe your 6x8 tile at 45 degree is already very near the optimal situation!
So, supposedly the 2-2.5k peak is the most severe breakup so it's splashing everywhere. And how the top HF is reflected from ceiling depends on the room acoustics when there's no obstacle nearby.
I guess, by the dimension of the deflector related to the dispersion, there's only a (small) part of the 2.5k peak reflected to the listener, while most top HF is sent forward because it's beaming into a small angle which is pretty much covered by the deflector.
It seems there's a lot of fun stuff to play with the size/angle/position of the deflector/reflector.
Maybe your 6x8 tile at 45 degree is already very near the optimal situation!- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.