Jump to content
The Klipsch Audio Community

Chris A

Heritage Members
  • Content Count

  • Joined

  • Days Won


Chris A last won the day on November 23 2020

Chris A had the most liked content!

Community Reputation

3885 Legendary

About Chris A

  • Rank
    Music Enthusiast

Profile Information

Contact Methods

  • Website URL

Recent Profile Visitors

41247 profile views
  1. What looks great on a glossy sales brochure (or an EASE data file, as the case may be) and what actually makes a difference to movie-goers scattered throughout a commercial cinema auditorium is always confounded (in the lean/six sigma sense) with human decision making and organizational behavior. I'm pretty well versed on that subject as that's a major portion of the decision making of what goes into the product and what doesn't (and how much it costs) in the field of system architecting. You sell to one group of people (customer's decision makers and/or buyers), but another group takes delivery of the product/system (the system installation sites), and yet another group actually uses the products/systems (the users: you and I in this case), and all three groups do not have congruent needs/wants. So the price of the product rises, but the utility of the product for the end users doesn't usually reflect their optimal needs. In the case of cinema, unless the buyers are also sitting in the (fully populated) audience, the price rises or, considering lowest-cost acquisition approaches, the sound quality suffers accordingly. An observation: every cinema that I go into over the past 10 years or so that has the name "JBL" on the loudspeakers has been a big letdown in terms of the sound quality (for instance). And what usually happens is that during the physical setup and dialing-in of large cinema loudspeaker arrays: somewhere in the center of the audience apparently gets the best sound, but the other seats...not so much. Even though the users don't really care about some capabilities of the components (e.g., very high frequency coverage), the loudspeakers chosen will have them, nevertheless, if the decision maker is reading or looking (with their eyes only)...instead of listening critically and thoroughly. Each customer functional group gets what they want to some degree, and it's the acquirers/buyers that usually have the loudest voice in that process. And they largely sit in the prime audience seats (I would imagine) when making those decisions--or they're sitting at their desks trying to figure out which offering has the best "technical merit" based on the EASE data, and make their decisions accordingly, perhaps without actually listening to the resulting systems in a full audience. Chris
  2. Roy has apparently designed an in-throat acoustic lens to spread out the polars above 6-8 kHz and he's stated that it works in that regard only up to ~13-14 kHz. The Axi2050's 3-part phase plug (which is extensive) is apparently untouched. I think it's easier in a home environment just to point the loudspeakers at the listening position a little more because at those frequencies, they're not really bouncing off the room's boundaries and arriving at the listener's ears to increase the perception of "image broadening" that Toole refers to. At those frequencies, you're only getting the direct arrivals, and you're not hearing very much from in-room reflections, anyway. Pointing the loudspeakers at your listening position is much more effective in that regard. Now, if you're designing a loudspeaker for commercial cinema duty, that post-phase plug in-throat lens might make a lot more sense, since the audience needs uniform illumination across that frequency band at every seat in the auditorium. They can't all be at a "sweet spot". But in a home environment, the listeners (more than one) can easily accommodate being within 15-20 degrees on-axis by positioning the loudspeakers accordingly at the geometric center of the listener positions. So in other words, that money may be spent better elsewhere--like with a bass bin that has a bit more driver area to reduce modulation distortion due to the higher excursions of the fully horn-loaded single 12" woofer (even with a rear wave rectifier--the internal bass reflex ports), or adding a nose to the bass bin to spread out those lower midrange polars more effectively. YMMV. I think that I said that, above. The Axi2050 cost is lower (in general) than other well-designed 2" throat beryllium diaphragm drivers, and the f3 of the horn/driver is an octave lower than all other 2" compression drivers on a K-402 horn. If the affinity for "W" section dual mouth folded horn bass bins is still overriding, then using the Axi2050 addresses the one true Achilles heel of the Jubilee (present and future version): its lower midrange crossover interference band jump from the centerline of the K-402 to the centerline of the bass bin, which comes at a point that's in the middle of the vocal range (400-600 Hz) in the present model. I simply move the K-402 horn closer to the centerline of the bass bin to reduce this issue, and while this is a significant improvement in that area, doesn't actually solve that issue. Crossing over an octave lower is another way to lessen this issue (and the related issue of the truncated horn mouths that do not rejoin). Elimination of that issue is available by using a full-range MEH approach. Chris
  3. My comment on MEH was that I think the MEH portion of the loudspeaker is almost an afterthought, and one that can easily be replaced by a single 2" compression driver (with two diaphragms, perhaps), and reduce the cost and complexity quite substantially, and probably increase the performance in terms of avoiding higher order modes. FIR filters can be employed to get the flat phase response that they are showing above. It's the off-axis polar response that I was referring to. Longer than 1/2 wavelength corresponding to the horn mouth size (about 15-16 inches horizontally and vertically), the horn loses directivity control, so placing the woofer ports in the MEH portion of the horn means that the 8" woofers run out the end of the horn below 400 Hz and illuminate everything in the nearfield. They also require a lot of power (with resulting higher modulation distortion). This is my observation. I'm not a fan of this sort of thing. The reason why Danley has done this is not for its acoustic performance, but rather because he's selling (like so many others) to those that listen with their eyes rather than their ears... He's designed a loudspeaker that doesn't "walk the talk", so to speak. It only pays lip service to the MEHs that he's been trumpeting for over 20 years. But he has broached the subject of FIR-filtering-achieved flat phase response and its effect on the listener. Apparently, no one is paying attention on the web, but in-person listeners can hear the effect of the flat phase. I think that the only other loudspeakers that have flat phase are certain DSP-corrected (i.e., use of FIR filtering to achieve flat phase and SPL response...truly neutral transfer function) studio monitors--the little ones that are often placed on top of sticks, and are in the extreme near field of the mixing and mastering listening positions. This is what Danley has found, and it doesn't required MEH to achieve it, but it is significantly aided by full-range MEH designs, as I've found empirically with my MEHs. Chris
  4. It's a 4-way that's 17.4 inches wide and 45 inches tall, weighing 195 pounds(!), and is an active loudspeaker that loses directivity at 400-500 Hz (the 8 inch "woofers"). Dual 15" subwoofers are built into the base to extend the frequency response down to (a claimed) 14 Hz. It has a lot of amplifier power capability (400w to the tweeter and midrange, 750w to the woofers, and 1500w to the dual 15" subwoofers). It clearly uses DSP and FIR filtering to achieve this kind of SPL and phase response: It's not my cup of tea, but it is interesting. I bet the price will reflect all of the above. I wouldn't call it a multiple entry horn (MEH) design, per se, but one that makes use of a MEH top end. It has all the disadvantages of a direct radiating loudspeaker and few if any advantages that a full-range MEH has: efficiency and full-range controlled directivity. The only thing it really has is flat phase and SPL response due to the extensive use of FIR filtering. Chris
  5. A thread on the new Celestion Axi2050 Axiperiodic driver that directly applied to the K-402-MEH: Chris
  6. One set of plots that I didn't plot above is harmonic distortion. I see no real differences between the TAD TD-4002 and Axi2050 drivers across their respective passbands. Since THD is basically only showing you second harmonic distortion, in order to see the higher order (and therefore, more audible) harmonics, I have to plot those separately. First the TAD driver on a full-up Jubilee with bass bin at 93 dB (at one metre microphone distance, crossed at ~525 Hz) : Then the same harmonic distortion levels at 93 dB for the Axi2050 with Jubilee bass bin (crossed at 225 Hz): Chris
  7. Now for a discussion of the relative advantage or differences using the Axi2050 on a K-402-based MEH: Already, I think you can see that there probably won't be a way to measure and listen to the Axi2050 on a K-402-like horn until the off-axis ports are moved farther away from the horn throat toward the horn's mouth (about double the distance in order to cross over to the woofers an octave lower). So what does moving the woofer off-ports closer to the mouth result in? Well, the farther the woofer ports are from the horn mouth, the lower the first-bounce cutoff frequency of the woofers to the throat return bounce (about an octave lower). What's usually forgotten in this discussion is that the woofers themselves will gain lower end response due to the fact that they will be located at a point on the horn with a lower area expansion rate, which is a function of straight-sided horns, commonly called "conical horns" but which are more properly called "straight-sided" because their area expansion formulae along the central horn axis is not generally a strict "x2" area expansion rate. So what? Well, this means that the K-402-MEH based on an Axi2050 driver can actually play a little lower than the equivalent horn using compression drivers that cross about an octave higher. A plot of the distance along the central axis of the "loading frequency" of the woofers vs. their distance from the horn throat (as plotted by Danley): It's not a lot of difference, especially considering that as the frequency decreases below the point at which the first 1/4 wavelength is fully contained within the K-402 horn. Below that frequency, the first 1/4 wavelength begins to use the room boundaries to support the fundamental part of the wavelength formation. But it turns out that every little bit helps, as is true with almost all considerations of woofers and horns below 200 Hz. So to see and hear the effects of moving the woofer ports farther away from the horn throat, another K-402 horn will need to be modified with off-axis ports in the right places. One more consideration: the horizontal off-axis polar coverage of the K-402-MEH prototype at and just below the internal crossover frequency suffers just a little. In the prototype K-402-MEH, the off-axis polars are somewhat decreased from 450-500 Hz, as can be seen below in its horizontal polar sonogram a little below 500 Hz as a little local "waistbanding": The interesting thing is that by moving that little disturbance down one octave to just above 200 Hz, the ability of the human hearing system to pick out that polar disturbance is less discerning. In the case of the Axi2050, it may will be that it becomes inaudible, since it might fall below your listening room' Schroeder (a.k.a. transition) frequency. This is a very good thing. So the first question might be: "do I need to move the ports closer to the mouth of the K-402-MEH?". The answer is, "no", you can leave them where they are at and simply cross the Axi2050 compression drivers at a lower frequency to the woofers. What is the effect? Very, very small, in my estimation. Chris
  8. Listening Impressions: As far as I can tell, the Axi2050 is pretty much a match for the TAD TD-4002, with the understanding that there is a very, very small apparent difference in "sparkle" or "air" on the very top end. It is so small that I doubt I could pick it out on a blind-blind A-B comparison test, unless the test music used was very specific to show the differences. My guess is that no one else would hear it either since it would take a very careful and meticulous dialing-in process in both cases (in-room EQ and delays). This implies to me that I could be making my judgment between the two based only on sighted bias. I think the two drivers are fully interchangeable in terms of their resulting sound quality after being very carefully dialed in. I have been playing the TAD TD-4002 on the left channel and the Axi2050 on the right channel for a couple of days now, and can make no observations about left-right sound quality/balance other than the sparkle comment I made above--subject to sighted bias. The difference in crossing the K-402/Axi2050 an octave lower than the K-402/TAD TD-4002 is not generally audible, although I have lowered the K-402 on top of the Jubilee bass bins in both cases to minimize the vertical separation issues between the two horns. It seems to me that the Axi2050 on the right, crossed an octave lower than the TAD TD-4002 on the left, makes a very, very subtle difference, and only if moving well outside of the listening position(s) laterally. The right channel (Axi2050) may sound a little more "full" in the lower midrange (225-500 Hz) which is difficult to pick out, but I would expect this might become more audible over time if listened to extensively. I can't make a judgment presently, but the idea (in my mind's eye) of crossing an octave lower is one that I would welcome--it's just that I'm not sure that I can really hear it presently. I would say this driver can't really be beat in terms of sound quality, equal to the sound quality of the TAD TD-4002s, and having an extra octave of low end response that the K-402 horn can easily take advantage of. One more point: I have seen no other horns that can fully take advantage of this driver's very broad-band performance than the K-402 horn. Even the largest SEOS 30 horn isn't big enough to take advantage of the extended low end of this new Celestion Axi2050 driver. Chris
  9. The Celestion Axi2050 is a new 2" throat compression driver with a very large diameter ring radiator diaphragm that uses advanced design techniques to avoid the problems of other dome-type titanium diaphragm 2" throat drivers. More on this subject can be found on the web in the form a fairly lengthy YouTube video. This is the driver that Roy has announced will be the compression driver used in the new upcoming "The Jubilee" that's slated to be released this summer, again using the K-402 horn and a redesigned bass bin with an internally ported single 12" driver. Some plots of the Celestion Axi2050 on a K-402 horn on top of my right Jubilee bass bin follow. First the raw, uncorrected response of both drivers. I use this plot to help determine the best crossover frequency: and the phase plot of the same data: Note the the suggested crossover point between the bass bin and the K-402 is about 225 Hz by these plots. I've found that 225 Hz frequency is a fairly good place to cross, or a little higher--about 300 Hz. There is a tradeoff in harmonic distortion levels at crossover, but it's not a strong factor, in my opinion. You might not be aware of the issue with the "W" section bass bins that have a truncated mouth (i.e., a flat baffle between the two horn mouths side-by-side). This introduces polar coverage issues above the frequency where the separation distance of the two horn mouths is more than 1/4 wavelength. For the Jubilee bass bin, that frequency is about 225 Hz. So crossing over at that low frequency (225 Hz) has the advantage of avoiding the narrowing of the bass bin polars that occurs in the current two-way home Jubilee, and the polar coverage of the K-402 and the bass bin are roughly about 90-100 degrees included angle. I can see why Roy liked the Axi2050, since it easily crosses about an octave below the other 2" compression drivers and can avoid the polar narrowing problems of the "W" section bass bins. The phase plot also shows the large overlap between the two horns/drivers, which indicates the design flexibility of where the crossover point can be located. From that plot, the crossover frequency can be about 190 Hz (where the Axi2050 on a K-402 horn really begins to lose on-axis SPL response) all the way up to about 800-900 Hz, where the Jubilee bass bin begins to lose on-axis SPL response. That's as wide a response overlap band that I've seen, and indicates great flexibility. The reason, as I said above, to set the crossover frequency at 225 Hz is to avoid crossover interference band polar narrowing with the "W" section bass bin. In an MEH, the choice of where to put the crossover is dependent on the placement of the woofer ports (I call these the "off-axis ports"). In my K-402-MEH horn, I cut these ports so that the crossover frequency would be about 480-500 Hz (which is 1/4 wavelength axially from the beginning of the off-axis ports to the mouth of the K-402 horn). There is a natural acoustic cutoff frequency due to the placement of these off-axis ports that you determine by the port placements. In an K-402-MEH using the Axi2040 drivers, this placement of the woofer off-axis ports could be about double the distance from the throat as my prototype MEH--about 5" axially presently to about 10" axially. If you measure along to the horn wall (which is flat as a board in this area), it can move beginning radius the off-axis ports from about 6" from the throat entrance to about 12". This is the reason why I tested the Axi2040 on a full K-402 horn (without off-axis MEH ports), because I had anticipated the low frequency performance of this compression driver to be quite a bit lower than all other 2" compression drivers--and it was: about full octave lower. Another view of the difference between the two drivers in terms of their raw response on a K-402 horn. Note that the relative drive level and resulting SPL of these two measurements are not the same: Here is the corresponding phase response of the two drivers: and the step response (a one-to-one comparison here): This shows that the moving mass of the Axi2050 is still higher than the TAD TD-4002, but the offsetting low end response and the lower cost of the driver (about 2.5x lower in new condition) is an offsetting factor. Here is the "predicted SPL response", in other words, the response based on initial flattening of the two drivers' responses, without extra effort to extend their responses on either end of the spectrum (high or low) based on the REW EQ facility's initial responses that I ran: This plot really shows you the relative response of each driver in as close to an "apples vs. apples" fashion. Here is the present case of flattened SPL response of the Axi2050 vs. TAD 4002 compression drivers on the same K-402 horn in the same room/microphone position, with the orange trace having an "in-phase" bass bin with the Axi2050 and the green trace having flopped phase in order to pull the phase of the K-402 and bass bin closer together, using the same PEQs to flatten the response: That isn't as remarkable as the following phase response, which tells a different story: So the option of having flatter overall phase response also implies using more PEQs to push the SPL response back to flat SPL again (not compensated above). In the traces above, no named crossover filters are used to cross the bass bin to the K-402 high frequency horn. Perhaps more discussion on that trade to come. Here are the Xilica settings that I used for the Axi2050 channel: Here are the relative PEQ settings use for the TAD TD-4002 driver and plotted (yellow trace) over the top of Celestion Axi2050 driver PEQs on a K-402 horn (white trace): Chris
  10. Not so much if the woofers are horn loaded, i.e., horn-loaded woofers use ~15 dB lower power than the same woofer being used in direct radiator mode (i.e., the horn-loaded woofer only has to move ~1/5th the distance of the direct radiator woofer to achieve the same SPL on-axis). So if you've chosen to have a small loudspeaker, you're going to pay a price. This is but one aspect of that choice. Another one is notably higher modulation distortion of the direct radiating woofers. PWK's quotes on that subject with Bruce Edgar here: https://community.klipsch.com/applications/core/interface/file/attachment.php?id=96455 That's a little bit away from the OP's question, but not very far... ___________________________________________________________ You also haven't identified the same issue within passive crossover networks themselves: ohmic heating due to their having to take the full power of the amplifier to achieve their function--as opposed to using DSP crossovers which eliminate parasitic losses on the amplifier--and instead direct couple the woofers to the amplifier output channels. A third effect is the microphonics of the passive crossover components--notably the capacitors become microphonic at some SPL, probably below the SPL mentioned just below when used inside the loudspeaker box. The case for a closed box woofer from S. Linkwitz (RIP): Chris
  11. Do you have them on a concrete slab floor, or is it suspended? Chris
  12. Here's a screenshot plot directly from DATS for the AMT-1 (after break-in). The flatness of both the magnitude and phase impedances vs. frequency is pretty spectacular in my experience: I suppose I could do it against an unused AMT-1 that I have on hand. It will probably be a couple of weeks before I could get to it, however. The real changes that I've seen with drivers is the acoustic phase and group delay response from REW. Below you will see the group delay (the plot of the slope of the phase line vs. frequency) for a brand new AMT-1: (courtesy of Rudy81) and the same driver model after being broken in a week of normal in-room play: (my measurement) Since the AMT-1 is only good down to about 600 Hz, that's a lot of change over a week of playing time at ~1800 Hz...a lot. EDIT: The threshold of audibility of group delay is something like 1-2 milliseconds at 1800 Hz. The AMT-1 diaphragm has to loosen up, and that takes a while. Chris
  13. You didn't read the article that Cory linked to...right? Some of the drivers' T/S parameters are changing a lot more than you might imagine--even out to 80 hours of break-in signal, and that's just for little 8" woofers. Big woofers could change even more in terms of T/S parameters. Chris
  14. Where did PWK document that? I haven't been able to find it anywhere in Dope from Hope or the other articles he authored, doing a brief search. What is the gauge of typical voice coil wire? How long is the voice coil wire? This is a good point made by Greg B. (Edgar). I think it's wise to point out the sensitivity of, say, Khorns to amplifier output impedance. Here is a notable input impedance plot of the Khorn from PWK's time (1986 from Richard Heyser's famous KHorn review for Stereo Review magazine) that shows the rather wild swings in input impedance: If the amplifier you're using has an output impedance that's anything like 0.5 ohms or even higher--which is typical for SETs, you're going to see SPL response changes in the Khorn relative to an amplifier having 1/100th of an Ohm output impedance--which is typical for transistor amplifiers. No one can really say which is better, however. And it's probably easy to state that smaller diameter/cross-section wire is just as good as larger, since the added resistance of the wire under load will raise the effective resistance of the loudspeaker (without materially affecting the reactance), which is exactly what one does when using a current-feedback amplifier (transconductance type). The voice coils will certainly gain much more heating than the loudspeaker wires--especially with loudspeakers having 85-94 dB/m sensitivity. The advantage of doing it with long loudspeaker wires is that it probably won't change its resistance very much while the loudspeaker is under great load (i.e., playing quite loudly) since the cables will be effective at dissipating the ohmic heating into the surround room more effectively than a voice coil. Just pointing out that our assumptions on what's better and what's worse doesn't always have to track with dollars spent. Chris
  • Create New...