Didn't say that.
That is what i was convinced about my implementation as well. And so many audiophiles, that heard my system.
That is your personal opinion, and i respect it. I shared it actually, for many years. I will consider however only relevant opinion , of who has made the same experience as i have, and has made direct listening comparisons with Beyma TPL150.
Sure Angelo, whatever. I've listened to horns, I've listened to direct radiators, I've listened to big and small planars. All have their strengths and weaknesses, but well done can (and should) sound remarkable similar. What you then like or hate about them becomes a matter of taste.
It certainly IS a matter of personal opinion - and taste. No speaker is perfect, I just happen to prefer what good horns do well over what other types of speakers do well. Those are my priorities. Considering only the opinions of those who have heard a certain Beyma mid-tweeter is your privilege, but rather limited in scope, don't you think?
It certainly IS a matter of personal opinion - and taste. No speaker is perfect, I just happen to prefer what good horns do well over what other types of speakers do well. Those are my priorities. Considering only the opinions of those who have heard a certain Beyma mid-tweeter is your privilege, but rather limited in scope, don't you think?
Angelo - wasn't your horn something of a random mixture of different flare types? From the drawings I saw, there also appeared to be huge mismatch in flare rate at the throat. I've heard, for example, the same driver on 2 slightly different flare rate horns (JMLC) with vastly different levels of colouration. I'm convinced the throat matching is critical. And obviously not crossing over too low... which gets all the more tempting the bigger the horn. If you didn't address these aspects, I'm not surprised at the outcome.
Speaking of the JMLC - I recently had a brief listen to an E-JMLC 600. No horn colouration that I could detect whatsoever. Something of a first to my ears.
How much of that is due to the elliptical profile vs (probably fortuitous) flare match I don't know. Certainly the round JMLCs I've heard can come very close, and I suspect with proper attention to the thoat angle matching would be just as good.
I'd love to see an E-JMLC calculation option is hornresp some day hint hint
Speaking of the JMLC - I recently had a brief listen to an E-JMLC 600. No horn colouration that I could detect whatsoever. Something of a first to my ears.
How much of that is due to the elliptical profile vs (probably fortuitous) flare match I don't know. Certainly the round JMLCs I've heard can come very close, and I suspect with proper attention to the thoat angle matching would be just as good.
I'd love to see an E-JMLC calculation option is hornresp some day
hint hint
Angelo - wasn't your horn something of a random mixture of different flare types? From the drawings I saw, there also appeared to be huge mismatch in flare rate at the throat. I've heard, for example, the same driver on 2 slightly different flare rate horns (JMLC) with vastly different levels of colouration. I'm convinced the throat matching is critical. And obviously not crossing over too low... which gets all the more tempting the bigger the horn. If you didn't address these aspects, I'm not surprised at the outcome.
Speaking of the JMLC - I recently had a brief listen to an E-JMLC 600. No horn colouration that I could detect whatsoever. Something of a first to my ears.
How much of that is due to the elliptical profile vs (probably fortuitous) flare match I don't know. Certainly the round JMLCs I've heard can come very close, and I suspect with proper attention to the thoat angle matching would be just as good.
I'd love to see an E-JMLC calculation option is hornresp some day
I have made direct comparisons between the TPL150 with wave guide, and without wave guide. The difference can be heard immediately.
So my conclusion is : even a very shallow wave guide colors the sound.
The question should not be, if horns color the sound or not, in my view, they all do. But how much coloration you are willing to live with.
Most horn enthusiats do live with the illusion, no coloration is there, and are happy with it. And this is perfectly fine......
If we agree that either system can be adjusted to be neutral, horns will still benefit from their directivity. When it comes to the point where the speaker is causing less problems than the room, directivity is important.
Horn builders are also less likely to allow the baffling to be over-run. Not that it can't be done well with a direct radiator, but it can be done very well with a horn system.
Horn builders are also less likely to allow the baffling to be over-run. Not that it can't be done well with a direct radiator, but it can be done very well with a horn system.
I prefere the spaciousness of wide radiating speakers. I like a higher participation of indirect sound. With horns, i was able to move my head just a littlebit, and perceiving the directivity, and subsequently the change of portray of the musical event. That makes direct radiators being somehow softer and more forgiving, less sharp, precise , and less dynamic.
original QUADS
Hi Angelo,
Have you ever listened to the original QUAD electrostatic, either single, or double panels, used as a Mid-Range only ? (such as an HQD)
After listening for a while, then switching back to direct radiators, the colourations of boxed direct radiators becomes quite apparent.
Hi Angelo,
Have you ever listened to the original QUAD electrostatic, either single, or double panels, used as a Mid-Range only ? (such as an HQD)
After listening for a while, then switching back to direct radiators, the colourations of boxed direct radiators becomes quite apparent.
hi Scott
thats why i am quit curious about the Bohlender & Graebener Planars. There is a RD50 for sale at the sales section, price is ok. But shipping to brazil is very expensive. The Quad's are very directional, aren't they ? i don't know either about dynamics.
It sounds as if the horns were making an improvement in spite of the other issue.
I concur completely with Scott. Although there are dynamic limitations, the original ESL57, in singles or pairs, have the most natural midrange ever made (and measure very well too). They do not sound like the later ESL63, or modern electrostats. They basically sound like real life ... up to about 92~95 dB SPL.
If they don't sound like real life, there is something horribly wrong with the replay system, or the ESL57's have not been correctly restored (and they all need restoration after all these years). They should sound like an open window into the performance hall.
The real shocker is a direct A/B comparison with a pair of conventional audiophile speakers; it makes the taken-for-granted colorations very, very obvious, and will permanently re-calibrate your perceptions of loudspeaker coloration. The only other analytic tool with this kind of power is a pair of Stax electrostatic headphones in good condition (the same applies to the working pair of ESL57's). The A/B comparison should be instant and obvious.
The gotcha with the Quads, of course, is dynamics. When they crackle it is too loud and you are risking damage to the speaker and the amplifier. Using them for midrange-only helps, but the dynamic limitation is a fundamental physical feature of the loudspeaker.
In comparison to the Quads, many, if not most, horns have obvious colorations, and a peculiar exaggeration of dynamics (which may be part of the tonal coloration). But a LeCleac'h with accurately matched exit and throat angles, combined with a crossover that is not first or second order and also well away from the cutoff frequency, has some of the lowest coloration of any driver ... ribbon, dome, direct-radiator, as well as many commercial electrostats.
Horns are less forgiving of crossover errors than any other type of driver; inadequate slope, or crossing too low, instantly turns the sound into grainy, harsh, low-quality PA sound, or the screechy mess you hear at modern movie theaters. If you hear any grain at all with a horn, then something is wrong: diffraction in the throat (or mouth), too low a crossover frequency, or a slope that isn't sharp enough, which allows too much excursion for the compression driver.
That's why I look at the decay tail of impulse measurements: that's where the ESL57 is better than nearly any other loudspeaker (it has about the quickest, and most non-resonant, decay). This is the measurement that separates good horns from bad ones: many other horns have long and resonant decay characteristics, worse than many direct-radiators, while the LeCleac'h does a quite creditable imitation of an electrostat.
I'm in the minority here, but I feel that quick, non-resonant decay is more important than the directivity pattern, and the LeCleac'h trades good impulse response (and even diaphragm loading) for the more popular constant-directivity pattern. (Even though a LeCleac'h is not considered a "constant-directivity" horn, the dispersion pattern is very smooth and free of sharp edges or ripples in the main pattern.)
Last edited:
Quick followup: some readers may be puzzled by my advocacy of high-slope crossovers while at the same time emphasizing the importance of rapid decay in the tail of the impulse response.
The time distortion of an all-pass crossover (such as LR4 or higher) mostly affects the shape of the transient itself, producing a quick overshoot and return-to-zero. This is a separate issue from discrete reflections from diffraction, standing waves in the horn or on the front surface of the enclosure, or diaphragm resonances. These problems create much longer decay-to-zero, and typically have several resonant frequencies that appear in the decay interval.
Since the ear/brain uses decay information to discern instrumental timbre as well as decode spatial information, clutter in the decay interval creates noticeable artificial colorations that are not heard in an acoustic performance. We tend to accept these colorations uncritically since they are nearly universal in electronic reproduction; it takes a direct exposure to a very low decay loudspeaker to become aware of just how bad most loudspeakers really are.
Side note: It doesn't really look like the TPL150 "waveguide" is optimized for low diffraction: it looks suspiciously like two-dimensional conical horn (in the horizontal plane), which is not going to have low diffraction. In the vertical plane, there's no horn at all, and the mouth acts like a diffraction slot.
The time distortion of an all-pass crossover (such as LR4 or higher) mostly affects the shape of the transient itself, producing a quick overshoot and return-to-zero. This is a separate issue from discrete reflections from diffraction, standing waves in the horn or on the front surface of the enclosure, or diaphragm resonances. These problems create much longer decay-to-zero, and typically have several resonant frequencies that appear in the decay interval.
Since the ear/brain uses decay information to discern instrumental timbre as well as decode spatial information, clutter in the decay interval creates noticeable artificial colorations that are not heard in an acoustic performance. We tend to accept these colorations uncritically since they are nearly universal in electronic reproduction; it takes a direct exposure to a very low decay loudspeaker to become aware of just how bad most loudspeakers really are.
Side note: It doesn't really look like the TPL150 "waveguide" is optimized for low diffraction: it looks suspiciously like two-dimensional conical horn (in the horizontal plane), which is not going to have low diffraction. In the vertical plane, there's no horn at all, and the mouth acts like a diffraction slot.
Last edited:
Looking at Bjorn's BEM acoustic-impedance simulations, an exit-matched LeCleac'h has about the smoothest transition in and out of the cutoff region of any profile simulated. The other profiles, particularly conicals and near-conicals, were much worse, with very large ripples extending up to two octaves above the nominal cutoff. These acoustic-impedance ripples unload the diaphragm, create impedance peaks, disturb the time response, and create ripples in the polar pattern.
We didn't see any horns that didn't have a very sharp entry into the cutoff region: when they're gone, they're gone, and it all happens in less than an octave.
The implication for low-slope crossovers is there will be a region just below cutoff where the diaphragm is nearly unloaded; in other words, it "sees" free air, as if there was no horn at all. Some diaphragms tolerate unloading better than others: cone drivers don't care much, since they work pretty well as direct-radiators anyway. Plastic diaphragms in compression drivers also seem to tolerate this condition pretty well.
The metal diaphragms are another story. Grit-n-grain rise very fast if the crossover allows much electrical input in the unloaded region. My hypothesis ... really a wild-***ed-guess ... is that as the diaphragm approaches the spherical back surface of the phase plug, air turbulence (around the slit entrance) and compression start to become significant, and this is audible as a "roughening" of the sound.
The plastic diaphragms bend a little as pressures get higher, which alleviates the roughening of the sound. Looking at it another way, plastic diaphragms are akin to soft-dome tweeters, which are always distorting and bending a little, while metallic tweeters stay rigid until the abrupt onset of breakup, and then enter a region of very high distortion. I'm doubtful that aluminum and titanium diaphragms in compression drivers are actually physically breaking up at the microwatt and milliwatt levels that audiophiles are using, but something is indeed happening in the unloaded region of the spectrum. Maybe the suspension is rocking: I don't know, but it's audible as rough, grainy sound.
Last edited:
I'd like to ask a basic question: how do you calculate the exit and entry angles for compression drivers and horns?
For CDs, I don't think they are available on spec sheets, nor are they discernible (AFAICT), from the CAD drawings of the drivers. Regarding horns, maybe I need to do more reading about how the entry angle is calculated. Some pointers would be helpful.
Thanks!
For CDs, I don't think they are available on spec sheets, nor are they discernible (AFAICT), from the CAD drawings of the drivers. Regarding horns, maybe I need to do more reading about how the entry angle is calculated. Some pointers would be helpful.
Thanks!
I've sent email to the manufacturer's tech support email address asking for the information and have got reply's both times. Once with Radian who sent a cad drawing of the throat with the angle marked and another time from Celestion about the small CDX1-14xx series. Celestion sent the requested angle information in the reply email.
Gary
Gary
That's one of the main reasons I like the mutli-cell horns - a wide sweet spot. They tend to give a wide, even sound.
That Beyma tweeter does look superb. Might be just the tweeter I need, except for the price!
That Beyma tweeter does look superb. Might be just the tweeter I need, except for the price!
The TPL excells in the upper midrange, from ~ 1,5khz to 9khz.
Just for treble duty, it seems the RAALs are better. Someone send me a p.m., saying following :