I wonder what kind of horn you were using, and if that had any effect on the 700-1200 Hz response? Here is the raw SPL and response of the Axi2050 on a K-402 horn (on axis). I also have included a group delay/excess GD plot and a normalized spectrogram for your viewing pleasure:
I don't see much happening in the 700-1200 Hz region. However, I did have to boost the output at 16 kHz (which I got away with due to the absence of chattering at this frequency):
Chris
I don't see much happening in the 700-1200 Hz region. However, I did have to boost the output at 16 kHz (which I got away with due to the absence of chattering at this frequency):
Chris
Note that the little throat lens that I identified above was developed by the same guy that did the new Klipsch Jubilee (using the Axi2050 driver on a K-402 horn) specifically to spread out the polars of that combination driver/horn. As I stated above, I didn't get a chance to do a test with the Axi2050 and the lens, since I received the lens much later (December last year). I do know that there was some sort of big deal made of the effects of the lens with that driver by Roy D., which I found (via listening to the TAD TD-4002 drivers I own) to also help, rather dramatically.
You can hear the effect most strongly when you stand up from a seated position at the main listening position (MLP). Without the lens, you could clearly hear a extreme HF timbre shift, and with the lens, no HF timbre shift when standing up. I sit about 10-11 feet from the K-402 horns/KPT-KHJ-LF bass bins, and the K-402s are shortened down such that their centerline is about 47 inches/1.2 m above the floor-which is recommended in multichannel array setup by ITU-R-BS-775-1:
Chris
You can hear the effect most strongly when you stand up from a seated position at the main listening position (MLP). Without the lens, you could clearly hear a extreme HF timbre shift, and with the lens, no HF timbre shift when standing up. I sit about 10-11 feet from the K-402 horns/KPT-KHJ-LF bass bins, and the K-402s are shortened down such that their centerline is about 47 inches/1.2 m above the floor-which is recommended in multichannel array setup by ITU-R-BS-775-1:
Chris
I can confirm that the AXI shoots like a laser beam over 10k
But it does have one of the cleanest impulse responses I've seen.
A with/without lens plot of the same driver and same horn would be a really nice thing to end the discussion 🙂
And if a kind soul here will provide me with the .stl file, I'll print a plug and measure 🙂
But it does have one of the cleanest impulse responses I've seen.
A with/without lens plot of the same driver and same horn would be a really nice thing to end the discussion 🙂
And if a kind soul here will provide me with the .stl file, I'll print a plug and measure 🙂
Last edited:
The driver is inept where it counts -- frequency extension. And I do not believe beaming is the sole reason for this. It may boast low distortion though I perceive that it is also over dampened. In addition non-distortion waves could be lost or cancelled due to the complex pleated titanium diaphragm. I am not convinced by that approach. Also the resonance issue due to poor internal enclosure damping or something is significant. The driver feels prosumer and whimsical in its unique complexity... a stale design prioritizing an homage to heritage drivers like the Altec 288 in look and feel, over logical and simple design choices that could prove to be evolutionary step forward. I am not an Anglophile though I do adore Celestion guitar speakers. I could not care less about peoples nationalistic allegiances which factor way to heavily into peoples brand biases. I find the price especially considering the Klipsch Jubilee, Toys R Us design, and performance somewhat pathetic to be honest. The 288s, TADs and a hand full contemporary drivers thrash that driver in my experience -- achieved by both simple elegance from a design engineering perspective; and achievement of performance fundamentals.
Last edited:
https://data-bass.ipbhost.com/topic/1381-measurements-from-big-cds-and-horns/ I fail to see this 700 -1200 Hz resonance issue. Here are some videos on the design of the AXI2050:
avoiding resonances is a motivation for the design.
what are the things creating THD in a compression driver ? Mismatch of material between the surround and emissive area of the diaphragme ? Material of the diaphragme itself ? (why titanium could be better than aluminium and so on Difusor ? Back chamber ?
I observed that when a certain frequency played through the driver when playing a particular song, I cannot remember which, while I was AB comparing with another driver, it caused what I perceived to be the rear cover of the driver resonating at that very specific frequency. It was not a general resonance but rather a very specific one which I perceive to be isolated somewhere in the 700hz to 1200hz range though maybe a bit lower. It sounded like a feedback loop of a mechanical variety. I am sure there is a mechanical resolution to the issue though that would not resolve the unimpressive nature of the driver.
It is by no means a Grand Jubilee of a driver in my opinion. It is more impressive from an engineering standpoint than a cave painting, though a cinder block driven by a low power piezo electric actuator will produce minimal audible distortion.
It is by no means a Grand Jubilee of a driver in my opinion. It is more impressive from an engineering standpoint than a cave painting, though a cinder block driven by a low power piezo electric actuator will produce minimal audible distortion.
Last edited:
2nd order HD is mostly due to the high SPL in the phase plug channels and the properties of air as an acoustic medium so in intrinsic to the operation of compression drivers. As the throat size is increased keeping a constant compression ratio this distortion is reduced for a given SPL. Its mostly thought this distortion is not objectionable. This is in addition to the normal driver distortion mechanisms.
I think we need more H5 and H7 measurements... I hope H9 to be low enough with good drivers to be harmless enough.
Psychoacoustics and numerous studies suggest that the higher the order of the nonlinearity the more audible it is. Second and third order nonlinearity is dominant in most drivers, but yet it is the least audible. This is what makes THD so pointless. It has been proposed that the orders have level increase of 6 dB/oct. to make their appearance more psychoacoustically relevant. I would not disagree with that as an improvement, but it mises a few things.
The drivers which I am referring to are the Beyma CP755ND and its siblings, the Faital HF108, the BMS 4552, possibly the coaxial offerings from BMS and B&C though I have only tested the midrange sections of both drivers and I would say for dedicated midrange the Celestion is obviously superior either the BMS or B&C midrange. And I would concede the Celestion midrange is definitely superior to even an Altec 288 and maybe Beyma CP755ND -- it is able to be easily crossed over 250hz lower than the Altec and Beyma, and is clearer though maybe less musical than the Altec. In addition, in the video linked above, the Celestion engineer details the ways in which coaxial compression drivers have their pitfalls. With that said I have observed a huge advantage in a horn section of a two way or three way speaker functioning as a point source through using a single compression driver to cover a wide range, or a coaxial compression driver, versus stacked horns and drivers. The difference as it relates to coherence and resolution is night and day in my experience. Given this the only logical configuration I would consider if using a Celestion would be to have that driver cover 300hz to 2000hz with a large horn, and suspend a smaller horn concentric within that horn with the a driver that covers high frequencies properly, my choice would be the CP755ND mounted to a waveguide, and apply a time delay to the tweeter section. More interesting to me than this somewhat complex implementation would be an advancement at the driver engineering to recognize and properly implement the advantages of a point source coaxials allowing for the use of a single driver mounted to a single horn to cover a wide frequency band.
As a side point I feel similarly when looking at something like a meh synergy horn in that the drivers which Danley and the likes are using were never intended to be implemented in such a way. It is great to experiment and I am sure they are pretty dialed in at this point, and they sound good and they bump, though the drivers in said designs are delivering a hodgepodge of off axis information which is further degraded by the sonic energy that is reflected off the ported baffle horn and causing constant deflection of the woofer cone. Some speaker designers implore that any drivers producing frequencies above ~80hz should fire straight ahead. It would be interesting to see a more advanced take on the synergy horn that implements purpose built drivers of a design and implementation that is perhaps radically different in shape and function from conventional drivers.
In addition, a driver like the WE555 which can cover a range of 100hz and up to 15khz with some equalization is a very interesting design even today. There are some contemporary clones of this design. Despite being an old driver, this performance is more useful and versatile for ultra high fidelity applications than a midrange driver like the Celestion. It is worth reminding that many premium compression drivers today are intended to be installed into line arrays of various types for high power handling high output stadium level application. Naturally these drivers will perform better at loud volume and maintain greater composure under these conditions, than drivers like the TAD and Beyma, though as a general rule they will not be as resolving or refined sonically as their midfield studio monitor intended counterparts. The diaphragms of the class of more resolving drivers tend to be extremely thin and fragile which reinforces the intended usage and factors largely into a perceived superiority with regards to refinement and resolution.
Last edited:
So, why not use the AXi2050 in a big horn* 300-10000Hz and add a super tweeter.
*JBl 2360, Electro Voice HP9040 or big JMLC 200 pr 270 horns.
*JBl 2360, Electro Voice HP9040 or big JMLC 200 pr 270 horns.
One could. I cannot stand the sound of a JBL2360 and I imagine I would feel similarly about the EV. I am not after a horn loaded sonic experience. While just terminology, the idea of a waveguide as something that guide a wave, as the name implies, has yielded far better sonic results than horn loading. I think horn loading for anything but the most long distance applications is idiotic. I think a good pair of 10 inch sensitive low mass mid woofers per channel would yield equal and maybe better performance compared with the Axi2050 for midfield listening if seeking to reduce the crossover frequency of the woofers to ~100hz.
I find the CP755ND midrange outstanding and in no way inferior to the Axi2050.
There are disadvantages to your suggested approach, primarily phase cancellation and time alignment issues introduced by adding another driver. Historically it was the standard formula for high output systems -- old large Western Electric systems, Altec VOTT, JBL monitors, various fancy horn systems of today with supplementary super tweeters emulating these vintage systems -- before contemporary drivers like the aforementioned TADs and Beymas which can operate from a low crossover point relatively linearly to 20khz. It is retrograde to do what you are describing when the advancements achieved by the TAD TD4001 with regards to frequency extension and similarly performative drivers like the CP755ND are available. I believe the Beyma is more extended than the TAD. Said drivers which feature very powerful motors coupled with lightweight diaphragms that are large relative to any tweeter will yield better hf sound at midfield distances than any conventional tweeter or super tweeter. This is what makes a small handful of compression drivers vintage and contemporary so superior to anything else. While limited in some respects relative to the TAD or Beyma, I would opt for an Altec 288 for its hf extension and performance over an Axi2050 even if the Celestion offers superior midrange performance.
In other words, with the right choice of wide band single diaphragm compression driver, or a coaxial compression driver, mated to a proper waveguide, one is able to observe the sound emanating from the horn with a microscope if one desires. This quality is reduced with the addition of non concentrically oriented drivers.
It is uncommon for a 1.4 inch throat compression driver to extend to 20khz as the CP755ND and a few others do. The CP755ND exhibits excellent dispersion. The correlation of the diaphragm size and thus the amount of air being moved, the compression ratio, and the phase plug geometry all contribute to excellent treble response and dispersion. And that which is better than any super tweeter or direct radiator which would sound feeble by comparison.
Celestion does not publish the specs of the magnet used in the Axi2050 though I would bet it is not as powerful as the 2.2t tesla magnet in the CP755ND. Neodymium magnets of that strength are quite expensive. The diaphragm of the Beyma is equal in quality to the motor in my observation. This combination, operating from 500hz-20khz linearly with low distortion is a better formula than any other configuration I could think of. There is nothing else to explore in the realm of conventional drivers for midfield listening for me. I intend to experiment with modifying the enclosures of the Beymas, and explore some different waveguide configurations, though I am satisfied with my discovery of this driver and the current wave guide which I am using as is. I eagerly await a coaxial compression driver that can play from 100hz-20khz or better yet a single diaphragm driver that can do the same. Until then, the only other path of interest for me is electrostatic drivers which are superior to conventional transducers for near field and near-midfield listening.
I find the CP755ND midrange outstanding and in no way inferior to the Axi2050.
There are disadvantages to your suggested approach, primarily phase cancellation and time alignment issues introduced by adding another driver. Historically it was the standard formula for high output systems -- old large Western Electric systems, Altec VOTT, JBL monitors, various fancy horn systems of today with supplementary super tweeters emulating these vintage systems -- before contemporary drivers like the aforementioned TADs and Beymas which can operate from a low crossover point relatively linearly to 20khz. It is retrograde to do what you are describing when the advancements achieved by the TAD TD4001 with regards to frequency extension and similarly performative drivers like the CP755ND are available. I believe the Beyma is more extended than the TAD. Said drivers which feature very powerful motors coupled with lightweight diaphragms that are large relative to any tweeter will yield better hf sound at midfield distances than any conventional tweeter or super tweeter. This is what makes a small handful of compression drivers vintage and contemporary so superior to anything else. While limited in some respects relative to the TAD or Beyma, I would opt for an Altec 288 for its hf extension and performance over an Axi2050 even if the Celestion offers superior midrange performance.
In other words, with the right choice of wide band single diaphragm compression driver, or a coaxial compression driver, mated to a proper waveguide, one is able to observe the sound emanating from the horn with a microscope if one desires. This quality is reduced with the addition of non concentrically oriented drivers.
It is uncommon for a 1.4 inch throat compression driver to extend to 20khz as the CP755ND and a few others do. The CP755ND exhibits excellent dispersion. The correlation of the diaphragm size and thus the amount of air being moved, the compression ratio, and the phase plug geometry all contribute to excellent treble response and dispersion. And that which is better than any super tweeter or direct radiator which would sound feeble by comparison.
Celestion does not publish the specs of the magnet used in the Axi2050 though I would bet it is not as powerful as the 2.2t tesla magnet in the CP755ND. Neodymium magnets of that strength are quite expensive. The diaphragm of the Beyma is equal in quality to the motor in my observation. This combination, operating from 500hz-20khz linearly with low distortion is a better formula than any other configuration I could think of. There is nothing else to explore in the realm of conventional drivers for midfield listening for me. I intend to experiment with modifying the enclosures of the Beymas, and explore some different waveguide configurations, though I am satisfied with my discovery of this driver and the current wave guide which I am using as is. I eagerly await a coaxial compression driver that can play from 100hz-20khz or better yet a single diaphragm driver that can do the same. Until then, the only other path of interest for me is electrostatic drivers which are superior to conventional transducers for near field and near-midfield listening.
Last edited:
Edit: I think horn loading for anything but the most long distance applications is idiotic.
I think -- extreme -- horn loading for anything but the most long distance applications is idiotic. A large horn can be advantageous for a variety of reasons in any size listening space though I would opt for incorporating a higher crossover frequency, or adding another driver, than relying on a horn like the 2360 to achieve a desired low crossover frequency. I heard a wooden clone that horn on one occasion and it sounded very resonant and tonally off.
I think -- extreme -- horn loading for anything but the most long distance applications is idiotic. A large horn can be advantageous for a variety of reasons in any size listening space though I would opt for incorporating a higher crossover frequency, or adding another driver, than relying on a horn like the 2360 to achieve a desired low crossover frequency. I heard a wooden clone that horn on one occasion and it sounded very resonant and tonally off.
Last edited:
@gedlee,
I agree with you only to a point. I have read your papers and textbook.
Yes 2HD and 3HD are masked and they have the largest contribution to THD.
The rule, not the exception says that if 2nd and 3rd Distortions are present so are the higher order Distortions. 2nd and 3rd HD’s do not show up by themselves, the higher order related Distortions also show up to the party.
So a logical assumption is that if there is a high THD higher order Distortions are also elevated.
THD need not be a faulty measurement after all.
Thanks DT
This is point where i'd sharply diverge since i value midrange performance much more than highs.
It is logical, but does it always work out that way in reality?
I find this comparison of different compression drivers to be interesting and the distortion graphs show some differences.
http://www.justdiyit.com/grand-comparatif-de-compressions-1-pouce/5/
Most of the compression drivers have very dominant second order distortion with all of the other orders being quite close to each other but very low.
The DE250 is a bit of an outlier where each increasing order drops lower in general.
If these were compared percentage wise with dome tweeters the THD might well be higher but the higher orders lower.