I would be very surprised to learn that it's all the horn, despite your study. Certainly the horn contributes, but how can it be the sole cause? Are there any studies which demonstrate (audibly) a compression driver on a certain horn having no audible distortion when it clearly does on another?
The compression driver performance and it's termination are inseparable. You test always with some kind of termination, once you change the termination the whole system is different. I believe that design should consider the system as a whole to get optimum performance. I am much surprised why so many think they can just design independently and hope for the best. It must be some kind of audio conspiracy, to let you buy more in the process of trail and error.
Last edited:
Plane wave tube was damped. What load did this place on compression driver at 800Hz v load on driver in typical horn? Is 6dB/octave boost above 2kHz used with these drivers on typical horn?
Hello Barleywater
It depends on the type of horn. Exponentials as an example equalize the response flat by using the DI. CD type horns need the 6db boost to make them flat. But you can also attenuate the 1-2K region and use only the attenuation to do the EQ. Some horn and driver combo's don't conform the 6db rule and you need almost a parametric type EQ. There are lots of possibilities
Rob
It depends on the type of horn. Exponentials as an example equalize the response flat by using the DI. CD type horns need the 6db boost to make them flat. But you can also attenuate the 1-2K region and use only the attenuation to do the EQ. Some horn and driver combo's don't conform the 6db rule and you need almost a parametric type EQ. There are lots of possibilities
Rob
A very precise observation. There is such a thing as scale distortion (playing music at a different level than it is recorded), but I think what matters is the level the music was mastered at; that's the final 2-channel mix that meets approval with the musicians and the producer.
If the playback loudspeaker has good dynamic tracking, it should be easy to discover the level that the recording was originally mastered at. That's the level where the balance sounds "right" ... not too overbearing, but with full resolution and depth of bass. On a Compact Disc, the correct playback level will be different track-to-track, since there's commercial pressure to raise the level of every track so that at least brief portions are right at clipping, or perhaps right below it. On LP and the occasional ultra-fi PCM or DSD download, there's less of that kind of level-pushing monkeyshines, and you get to hear the track-to-track dynamic levels that the producer intended.
I've invented the term "dynamic tracking" to describe an even-tempered loudspeaker that sounds the same over a wide range of dynamic levels. I've heard the Summas, and they are exceptionally good at this ... low diffraction, low HOM's, efficient, cool-running drivers, and of course Dr. Geddes' overall design get the credit.
Many of the highly-reviewed (and very expensive) low-efficiency audiophile speakers have poor dynamic tracking. Gritty and harsh above a certain level, and conversely, if the level drops too low, the tonality and resolution drops into a black hole. This forces the listener to gain-ride every recording, and duck the level on sustained loud passages.
Electrostatics are an interesting special case. There's no drop-off at low levels, but high levels can create the dreaded sparking and crackling that signifies diaphragm punch-through and imminent destruction. So you're always wary with the volume control when you play a high-dynamic recording for the first time. Anyone that's ever burned out a direct-radiator driver is also a little cautious the first time you play a new recording with an unknown dynamic range.
Owners of horns with low diffraction and low HOM's don't need to worry about damage (except to their ears). The system can handle any transient, and resolution does not drop at low levels ... it stays the same, regardless of volume control setting. I grew up in the Fifties; hi-fi systems back then were about dynamics, when Klipsch, Altec, Bozak, and the Electro-Voice Patrician ruled the roost.
In the quest for neutrality that began in the late Sixties and extended through the Seventies, dynamics took a back stage, and overall efficiency dropped by 5 to 15 dB. As efficiency dropped, transistor amplifier manufacturers spread the rumor that the most powerful amplifier (which was the most expensive model) was also the best-sounding, something I have not found to be true. There are now demos at hifi shows where the amplifier is bigger than the loudspeaker, a definite case of misplaced priorities.
The most advanced horn designs (which certainly includes the OS) have finally conquered the ancient problems of diffraction and HOM's, which lets us hear horns in a new light. I'm aiming for optimized impulse response with a rapid settling time that is free of resonance, a classical goal of direct-radiator systems. From what I've heard of the first prototypes, dynamic tracking, freedom from horn coloration, and spatial presentation are very good.
Last edited:
Certainly certain items have to be considered as a set, and so compression drivers and horns form a set and so would a direct driver in a horn. We can draw pracrical analogies from elsewhere, but the process is then individual changes in discrete components to logically chosen multiple combinations until the whole final result achieves a genuine best result for a reasonable % of a random or specifically focussed population i.e professional sound reinforcement to the guy who likes decent stereo to relax to.The compression driver performance and it's termination are inseparable. You test always with some kind of termination, once you change the termination the whole system is different. I believe that design should consider the system as a whole to get optimum performance. I am much surprised why so many think they can just design independently and hope for the best. It must be some kind of audio conspiracy, to let you buy more in the process of trail and error.
From this thread like many others we can pull out others consensus and it is usually confusing.
With the latest drivers we need to see some comparative tests that really show the current state of play is.
If only I could afford a DHT SET or PPT with mumetal o/p transformers with direct coupling at least for a trial ( perhaps unsafe due to inherent stability risk of direct coupling). And this could then be put into a combination of Drivers/Horns/wave guides could we make any progress.
We cannot disregard 2nd order HD or any HD. We can have virtually no HD or IMD from amps and speakers as a goal should be the same. If you like 2nd order HD then go for it but that a separate track from the mainstream.We do not want HD or IMD from any of our equipment.
We seem to be scared of true fidelity. But it is there beckoning.
The quality achievable with linear panel speakers really does exceed anything else IMHO so my goal is to get as close as possible with easier techologies at a reasonable price. I do not want a cost no object 10kV electrostatic system. But wouild like to see if the best possible horn system is the way. So what is going to do this without spending space rocket money.
Hello,
Well, there is some truth in what you said in first approximation (but never completely verified in true measurements and at various levels, specifically at large SPL levels . Remember that a direct radiating loudspeaker will not be able to radiate more than 112dB SPL 1W/1m...) if we compare impulse response of the direct radiator driver itself (eg. mounted on a measurement baffle) to the compression driver mounted on a horn providing a resistive acoustical load.
But when the direct radiator is mounted in an enclosure this is no more true specially when the goal is to reproduce the bass frequency range, then the difference between the IR of a direct radiating loudspeaker mounted (e.g.) in a bass reflex and a compression driver mounted on a horn is clear.
And waterfall and CSD are here to demonstrate everyday the superiority of Compression driver + horns over direct radiating loudspeaker + bass reflex
Best regards from Paris, France
Jean-Michel Le Cléac'h
Hello,
I deeply disagree with the "less sweet" character of an added H2 distorsion, specialy when 3rd order and higer orders distorsion components are also there.
H2 will help by some masking effect to reduce the audibilty of the 3rd order component. Generally audiophiles agree on the harschness due to H3, then if H3 is less audible due to the masking effect brought by H2 the sound is surely less harsh.
Lovers of triode SE amplifiers will surely disagree themselves that the use of their SE is to obtain a" less sweet sound". IMHO its is the inverse, the addition of some H2 tend to give a less harsch sound (even at the price of some simplification of the sound).
Best regards from Paris, France
Jean-Michel Le Cléac'h
Last edited:
Hello Soongsc,
For sure reflections from mouth to throat are visible on IR measurements and also on the frequency response curve (by the comb effect they induce).
Give a look toe the lesurements on 16 horns I performed in the exact same conditions durnf the European Triode Festival sevral years ago.
www.forums.melaudia.net/attachment.php?aid=1760
(as an example compare N°5 and N°6).
Truncated horns (= with uncomplete mouth) behave more like tuned pipe and the reflection is often 10 dB more than horns a having a rollback at the mouth.
Best regards from Paris, France
Jean-Michel Le Cléac'h
Last edited:
May be my statement seems unclear. Please read again
I am saying removing th 2nd order HD this will result in a less sweet sound compared to the high 2nd order sample as it can no longer mask th higher order. We are the same on that. It is well established. But get rid of the higher order and then you do not need the 2nd order HD.
SE is to get a more sweet sound I thought that was clear from what I stated but better no HD or IMD. If we start with very low HD IMD in the whole system up to the speakers then some nice 2nd order can add a little more spice Eh. It would be easy to keep summing the 2nd order harmonic and higher even order harmonics away from fidelity with the source.
Thanks for this excellent study. It does not surprise me.
The horns are significant and so are the drivers TAD2001/4001 on the best examples. It would be good to see the Le Cleach with a reasonably priced compression driver to achieve similar results.
I return to the direct drivers. If a number of $40 tweeters can rival the best tweeters inc Be from an audition point of view, then it should also be possible for compression tweeters and Le Cleach horn or similar. And if they are out there now what are they.
Can you tell me what you believe is a mid priced driver that will achieve virtually the same result. Be is not the ubiquitous material but just one of a number of materials. I ride an aluminium framed bike. A beryllium frame would be better but by a smidgeon. The old steel frame is also pretty good. It needs a Behringer approach to distribute effective new perfomance compression drivers to match the mass market simple MP3 to DSD stuff at a realistic price.
Hello Jean-Michel,
I have listened to the Avant Garde horns many times. In fact, they present first good impression of the potential of horns to me. I have also listened to them driven by their own amplifier which was even better. However, I think they cross the tweeter at quit high frequency, and what we mostly hear is the mid range. Although my current small active speaker is more neutral and penetrating than their speakers, I will have to get a system of compatible physical size to have a more correct comparison, low level listening is slightly cleaner and clearer which correlate with the penetrating characteristics.
To avoid reflection and reduce defraction, my instinct is that the curvature radius should increase as it expands; however, this may not match well with compression drivers. This is why I think compression drivers may not be the best selection for in home listening. Compression drivers were developed to sent sound far out, that design works fine.
Remember I had a test article that used a combination of OS throat and LeCleach expansion? That actually showed the fastest decay in the higher frequency, but the lower frequency still could not decay fast, the exact reason still needs further investigation, but I am sure that amplifier design as well as driver design issues cannot be ruled out. Smaller curvature radius near the throat expansion seems to impose loading variation not so ideal.
I only know what we studied and I can only draw conclusions based on the results that I have. The results say that the horn is the major cause. Are there any scientific studies that say otherwise?
Last edited:
Here is response of sealed woofer before equalization:
An externally hosted image should be here but it was not working when we last tested it.
Here is response of dipole woofer, woofer mounted in 'U' frame before equalization:
An externally hosted image should be here but it was not working when we last tested it.
Here is response of sealed woofer stacked on top of 'U' frame woofer before equalization:
An externally hosted image should be here but it was not working when we last tested it.
The stacked system forms Cardioid pattern.
Here is sealed woofer response after equalization:
An externally hosted image should be here but it was not working when we last tested it.
Here is 'U' frame woofer response after equalization:
An externally hosted image should be here but it was not working when we last tested it.
Here is response of stacked Cardioid response after equalization:
An externally hosted image should be here but it was not working when we last tested it.
This is impulse response of monopole (top track) and 'U' frame before equalization:
An externally hosted image should be here but it was not working when we last tested it.
And these are impulse responses after equalization:
Sweeps employed in measurements were 26V rms.
Better impulse response behavior of drivers doing what they are suppose to be doing is poor criteria for comparison. Likewise, efficiency is poor criteria of comparison.
Earl,
Could be HOMs , could be air non-linearity distortion at high throat SPL levels makes compression drivers on horns sounded worse as they are turned up.
I haven't concluded which is the major culprit, I suspect both play roles.
Your saying that your waveguide loudspeakers do not sound worse as you turn them up "to the point where they are actually painful" really does not tell me whether they sound any better at that SPL level than the horn/drivers I tested.
If you were to do the same test procedure as I did, record the output of the horn driver playing a musical recording, raise the input level incrementally up to the thermal limits of the driver, note the peak SPL level of each iteration and allow us to compare the results to the original recording, it would be easy for all of us to hear whether the conclusions you have drawn are valid for our hearing.
If, listening to a volume normalized recording, your waveguide sounds the same through all the iterations, that would be conclusive proof of your statement, without that, you are, as you like to put it, "hand waving".
But I do like the way you wave
.Art
Execpt of course that I have heard the Gedlee speakers get rough at high levels. Sounded just like distorting compression drivers to me.
The waveguides are very nice, but drive that little 1" Cd hard enough and you'll hear it.
A test like Art's would be great. An actual test of the audibility that any one can access. That would go a long way to convincing me and probaly others. Remember the old saw about extrodinary claims and proofs? I'd be very happy if its true - but remaine skeptical.
The waveguides are very nice, but drive that little 1" Cd hard enough and you'll hear it.
A test like Art's would be great. An actual test of the audibility that any one can access. That would go a long way to convincing me and probaly others. Remember the old saw about extrodinary claims and proofs? I'd be very happy if its true - but remaine skeptical.
That's a convenient opinion for you and not one shared by any of the owners that I know.
If you are unsatisfied you do the tests and post the data.
Arts test would be great, but I am not interested in proving anything that is not fully scientific and subjective tests where you know what you are listening to are not scientific.
You guys always just "dig in" when someone posts something contrary to what you want to believe. It becomes pointless to discuss it any further. I'll stand by the data that exists, not some beliefs that have no supporting evidence.
Why do you think that we did those tests at B&C? Because I used to believe just like you, only I accepted the data and changed my mind.
If you are unsatisfied you do the tests and post the data.
Arts test would be great, but I am not interested in proving anything that is not fully scientific and subjective tests where you know what you are listening to are not scientific.
You guys always just "dig in" when someone posts something contrary to what you want to believe. It becomes pointless to discuss it any further. I'll stand by the data that exists, not some beliefs that have no supporting evidence.
Why do you think that we did those tests at B&C? Because I used to believe just like you, only I accepted the data and changed my mind.