How I Transformed My Home Listening Experience (Part II)

As we are writing I am doing some early morning listening to some Sonny Boy Williamson with Matt (Guitar) Murphy doing The Sky Is Crying and Don't let your Right Hand Know on The Fi/ Analogue Productions Sampler, I have never heard a blues harp sound so darn real, every breath, every note, they are worth whatever the cost!

 

That would have to be a huge difference in speaker efficiency (tweeters are inherently more sensitive which works against this notion) and 10db accounts for double the perceived loudness, so by that point, any perceived benefit of the super-tweeter wouldn’t be worthwhile.

If you had main speakers with a sharp roll-off of the highs around 8kHz or so, with something like a 4th-order slope, then you could probably get away with bringing the super-tweeters online at 16kHz without much detriment. Maybe this could work with some vintage speakers, but wouldn’t work with any of the warmest of modern speakers I’ve encountered. I own some Stirling SB-88s (LS3/6 minus the super-tweeter), speakers that are considered very warm by conventional standards, and relatively inefficient. Even when those were paired with the Aperions the detriment was clearly audible, but there was indeed some trade-off — more sparkle on the high end vs a loss of imaging precision and upper midrange presence.

 

But in the case of the Aperions this does not apply. Their roll-in slope is far too shallow.

 

I did some listening tests today and I did not detect any loss of soundstage, and I love my soundstage! There is just this increased sparkle and dynamics when switched on at the right level (for me). I tried as of 10k, 12k, 14k Hz ( at -2dB setting) no difference in imaging.

Possible explanations:
The Aperions are located about 10 inches from the tweeter, way within the 1.9 ft quoted in the paper I posted earlier “......comb filtering could be the result of a single reflection picked up... or could be the result of adding the signals of two microphones of which one is 0.6 m (1.9 ft) further apart compared to the other.”

I also measured the loudness at my Canton speaker and it decay is about 18dB from right next to the cone to my listening position. I am sitting in a large room, hence there should be no combing effects due to reflections.

You mentioned that you had experienced audible effects due to comb filtering. What was your set-up and which (super) tweeter did you test?

 

I tested the original Aperion super-tweeter, but the physics of comb-filtering will apply to both models. They were set atop both my Stirling SB-88s and Spendor SP2/3R2s. Both pairs have dome tweeters with excellent dispersion and a roll-off consistent with soft dome tweeters. I moved the speakers well away from any room boundaries and took gated, near-field measurements for a quasi-anechoic response. No appliances running in the house at the time—about as quiet as non-sound-proof rooms can get. The on/off switch of the Aperions made back-to-back comparisons a cinch. With them on, regardless of level, crossover setting, and position on the top of cabinet (always facing forward) there were massive dips and peaks in the response, even with 1/6 octave smoothing applied.

I think that paper you keep quoting is misleading you somehow, or you’re not quite grasping what it’s discussing. While microphones are similar to speakers in that they’re both transducers, the physics is not quite 1:1 because of the way sound waves propagate. It’s actually the close proximity of your Aperions to your main speakers that guarantees the comb-filtering. You want greater distance to mitigate the effect, not less. The solution is to turn the Aperions around and have them firing to the rear. That way their sound waves would be arriving late enough that they wouldn’t impede on your main tweeter output. The downside to that configuration is of course the aesthetics.

You may not hear the comb-filtering but it is indeed there. We can agree on that or not but the reality is it’s the truth.

 

As I pointed out, very high frequencies are impossible for our ears to localize. Because of this, like super low bass frequencies, they do not contribute to "imaging".

Which is why you can turn super-tweeters completely to the rear.

You can bounce them off the wall and because their wavelength is so short, it matters little that they are in phase or not. We simply can not discern any difference, so it does not effect the soundstage.

 

Right, when you're bouncing them off the wall it doesn't affect soundstage, BUT, when speakers are setup away from walls for ideal imaging (as most speakers require for ideal performance), and both main and super-tweeters are facing forward, the comb-filtering does impede on the upper midrange frequencies. The proximity of the listener to the speakers has negligible effect on the overall outcome in this scenario. The comb-filtering is an interaction between the tweeters, not a three-way triangulation where you take the receiving device out of the equation and it no longer occurs. The phenomenon is not isolated to the top end, even when the Aperions are brought online at 16kHz, they're still producing frequencies only 6db lower in level at 8kHz, which is why response valleys in the graph show up even lower than that.

As much as I wanted to not believe this guy's argument, his assertions are true and I verified them for myself:


There are work-arounds in the crossover domain, like implemented in some speakers I mentioned earlier, but those were designed with a super tweeter from the ground up. The problem with the Aperions is both the crossover slope they employ, and the absence of any low-pass filtering on the tweeters of most speakers they'll be paired with.

In your particular case, with the horn-loaded Altecs and horn super-tweeters, I'm not surprised there is no audible detriment, those are different animals than what this thread is mostly discussing.

 

And look guys, don't take me the wrong way, it's not my intention to rain on a parade. If you enjoy your systems more with the super-tweeters rather than without, then that's really matters, in your particular cases.

I'm only sharing my experience because I believe listeners should be aware of the physics involved and what they might encounter with this product. I believe they should at least conduct A/B comparisons and attempt to listen for any detriments. If they don't hear any, great. If they do hear some, then they'll at least know the likely culprit and can then try a placement somewhere other than the top of their speakers.

 

What should we do when several listening tests with several family members has been conducted with lots of different music, and the sound stage seems to suffer no ill effects from the front firing super tweeters being added and the top end is simple more alive, more brilliant more believable, more spacious, upper register clarity, and overall, just provides a fantastic, improved listening experience, what do you do when that happens?

I am all for the perfect curve on my computer, but when I take the super tweeters out, the sound is just not as "complete." I completely enjoy the clarity these speakers add and yes, I have tried the turn around and fire backwards so called fix, just no comparison, and I gave it a good shot. So, when all is said and done, I am simply not hearing the comb filtering problems that you speak about, and I'm real finicky if I hear something alter the sonic presentation negatively, and I just don't hear anything but added space, brilliance, upper register clarity, and improved and expanded sound stage, so for me and others that have listened to these speakers, if lovin these super tweeters is wrong, I don't want to be right!

 

I guess I am not understanding what fundamental principles are different?

While the speakers have different geometries, the physics doesn't change.

Altec says that their 511B horns and their drivers have frequency extension to 20-kHz., I beg to differ. The 908 series driver I use has the Altec Simbiotic driver diaphragm, which I also had in my 808A drivers some 50-years ago. They definitely roll off above 10-Khz., if not earlier.

They make a 902 series driver. It handles half the power and has a metallic diaphram and is supposed to be the more "accurate" of the two. While this may be true, to me it has the unmistakable sound signature of a metal diaphragm. I have these in one of my other pairs of A7's.

This is one of the custom ALK crossovers for the A7's. The rear crossover is for the super-tweeter. (The front crossover is the main 500-cycle crossover from ALK Engineering.)



The rear crossover is piggybacked to the main crossover and sends program material from 6-kHz. to the super tweeter.

It depends only of the natural roll off of the main 511B horn and 908 driver, so technically speaking, there exists an overlap between the main HF horn and the super-tweeter.

This was something that initially concerned me. But, I found that in actual practice, the two speakers blended perfectly. Surprising, the super tweeters volume is just right, as opposed to being too hot.

 

I suppose it’s sort of like how we perceive the Earth is flat when traversing the surface, when in fact it’s not. Make no mistake, the comb-filtering is there.

 

The difference is the horn loading which I suspect plays a large role in your perception. I’ve yet to encounter a horn speaker that performs the imaging trick particularly well in the first place. They produced more of a wall-of-sound presentation in my experience, which is sort of what happened to my system when the Aperions were in play. The imaging became more diffused. In the case of the conventional speakers members are using, I’m not surprised if they don’t notice an imaging detriment if their speakers are near the walls as in the OP’s case.

 

Well, they can do both. Being that they control and focus sound, they eliminate most first reflections, allowing the listener to hear music that emanates directly from the speaker drivers.

In addition, legacy speakers like the Klipsch K-Horns, La Scala's and Altec theater type speakers closely time align the drivers on the same vertical axis.

Here is a close up of one of the A7's. Notice how the bass speaker is is mounted deeply recessed into the cabinet.



Notice on my previous photo of the crossovers, you can see the driver at the rear of the 511B sectoral horn. The driver lines up with the voice coil of the bass speaker, maintaining phase coherence, which an important consideration in imaging.

Tower speakers of any kind are not well suited for imaging, due to their multiple vertically stacked drivers.

The best imaging comes from a point source, such as the Altec 604 Dual Duplex drivers, which incorporate a HF into the center of the woofer. This design also provides precise time alignment of the drivers.



The Tannoy Dual Concentric Drivers are another good example of a point source design.



"Sound energy is propagated from exactly the same point and delivered through the centre of the low frequency cone- a true point-source. The Dual Concentric delivers a spherical wave front that ensures even dispersion in the horizontal and vertical planes, providing exceptional off-axis performance."



"As you can see in the cutaway drawing (below) of Tannoy's Dual Concentric™ drive unit, the flare profile is continuous from the high frequency Pepper-Pot Wave Guide™ out across the low frequency cone to provide optimum dispersion across the audio frequencies to give a smooth transition at the crossover point. One of the most important aspects of the Dual Concentric™ drive unit design is that it has all frequencies emanating from the same central axis, with the high-frequency exponential Pepper-Pot Wave Guide™, located in the center and behind the low frequency driver, perfectly integrated to produce polar sound dispersion that is symmetrical in both horizontal and vertical planes, a design aspect which has a huge advantage over multi-driver systems in coherency due to being in essence a point source, which I'll talk about more as I describe my listening impressions."



For those who aren't vested in $50k speakers like the Westminster Royal SE's, the KEF LS-50's function on the same principle for $999/pr. (The link is to Positive Feedback and these Westminster Royal SE's belong to Jeff Day.)

The optimum performance from most speakers are obtained by positioning, usually this means pulled out from the wall and other room boundaries.

In our modest listening rooms, this is often not a possibility, unfortunately...

 

I’m pretty well-versed in the fundamental physics of such designs. I just haven’t heard a horn-based system that does pinpoint imaging particularly well. Though I agree that nearly all speakers benefit from space, horn or otherwise. It’s my supposition that first reflections due to speaker/wall proximities are why many owners of the Aperions cannot detect the comb-filtering. Which is all well and good, but to claim there is no comb-filtering occurring simply because they cannot hear it, well, that’s just unequivocally false.

 

The new EISA award winning Dali Callisto series features a new hybrid tweeter system consisting of standard dome tweeter and a ribbon supertweeter. The 2 work in parallel up to ca. 15kHz and the supertweeter extends up to 30kHz.

The frequency response also seems quite balanced, here measured by German audio mag.

This is a quite similar set up compared to using the Aperion on a speaker with a conventional tweeter, except that the Dali’s supertweeter cannot be adjusted , whilst the Aperions can be adjusted within a wide range.

 

If you had a good understanding of speaker design, you’d realize the Aperions have zero advantage over these Dalis that were designed holistically to include a super-tweeter. The Dalis still exhibit a pathetic response on-axis (Figure 2 in the linked measurements below), with a difference of nearly 10db between 2.4kHz and 10kHz, and about an 8db difference between 2.8 and 3.4kHz. That’s the sort of result one gets from comb-filtering! However, they managed to mitigate this in the off-axis behavior through careful consideration of the two tweeters’ dispersion patterns, basically forcing owners to listen off-axis for a balanced response. It’s also likely that the soft dome tweeter has a relatively sharp mechanical roll-off, unlike a conventional dome.

DALI Callisto 6 C wireless loudspeaker Measurements

The takeaway here is that Dali had to take BOTH tweeters into account when designing this speaker. They didn’t take a random soft-dome and decide later to throw the ribbon tweeter into the mix. It was a holistic approach that didn’t turn out as bad as it could have, yet many speaker designers would still consider this a dismal performance given the on-axis behavior. What you actually get in the Callistos are active speakers that don’t even take smart advantage of the active part. They could have achieved a flatter response through use of an active crossover between the soft-domes and ribbons.

The Aperions offer no advantage, unless you consider the ability to shift the point at which the combing comes into play an advantage.

Anyhow, I’ve beat this proverbial horse to death. Enjoy your Aperions. Ignorance is bliss after all.

 

Yes that might be a good advantage. I do like the idea of opening up the soundstage and widening the on-axis areas without eq of the signal or any altering a digital signal with dsp. And I do hear a lot of dull sounding recordings and mastering, which I figure is not all (only some) my aging hearing, because it is not on every recording I play.

 

My middling jbl l890s go 28hz-40khz and already has a super-tweeter.
Would this actually make a difference?

 

I honestly don't know, however, I'm going to take a shot here and tell you NO, the L 890 seems to be loaded with tweeters, horn tweeters, I'm going to say the Aperion would probably not be a good match for your particular loudspeaker. If for nothing else, you would really be tweeter heavy.

 

In your case the objective could not be extension of high frequency as your very nice speakers already run to 40kHz. If the objective is to boost the very high frequency range, maybe?

Here some FAQ’s by Aperion:
Will adding a Super Tweeter cause phase issues?
Answer: While adding any type of additional speaker to a signal chain has the potential to cause phasing within a room, our 1-step crossover design mitigates any phase issues. This 1-step design is not to save cost and space because if we were to add multi-step crossovers into the chain, it would cause a major phase effect and the sound test results would not sound good.

What kind of speakers will the Super Tweeter work with?
Answer: Most passive bookshelf or tower speakers are great candidates for the Super Tweeter. But, you can also use them with center channel speakers and surround speakers.

This would be your situation:
"Will the Super Tweeter work with my electrostatic, full range, or professional audio speakers?"
Answer: We advise you to check the sensitivity level and frequency response of your current speakers. A speaker that has a higher sensitivity level rating than the Super Tweeter might not be a good fit. However, the we've received feedback that the Super Tweeter works great with Bose 901's, Quad Electrostatic 2805, and other full range and electrostatic speakers.

Your JBL’s have a quoted efficiency of 91bB. The small, medium and new large Aperion models have 94, 96 and 98dB, respectively. On paper it could work, but the best chances are with the bigger models. You would definitely have to test the Aperions on a trial and free return basis to find out.

 

I was speaking with another member (Fish) of the forum that was very interested in the Aperion super tweeters and I suggested he try the less expensive Planar Magnetic design tweeters. He has two sets of speakers, some floor standing Alons as well as some very nice KEF's, a very nice setup with a great vintage receiver, he gave me permission to post this pic of his equipment!

He is still tweaking his attenuation and best roloff frequency when used with his KEF loudspeakers. Like me, he is very pleased with the added brilliance and clarity he is hearing in the upper octaves. The last time we spoke he was at 12KHz at -2db. The efficiency of the KEF's is 92db.

Here is his system:

 

If you have not tried the Full Ribbon super tweeters, they are just extremely nice. I started with the great Planar Magnet design, which I still use on my bedroom system, but I went ahead and moved into the full ribbon design on the larger speakers in my main listening room, they are a bit larger and just add more beauty to that top end. My timing was not so good, as I got the full ribbon design just a short time before Aperion came out with their newest offering, the dual AMT ribbon super tweeter, would certainly love to try that one!

 

I'm wondering exactly what thoughts Aperion would have with regard to their ribbon tweeters causing comb filtering?

 

I have asked several people about this Larry, and as of yet, no responses. I did a very practical re-test again. I got in my sweet spot, made sure the Polk 707's were towed in just right, closed my eyes and took the tweeters in and out during many different passages, I literally tried to hear problems, muddy altered sound stage, loss of midrange, gaps in the high frequencies, and no matter how hard I tried, I just could not hear anything but improvements in everything, eyes closed, Telarc, Pope Records,
Reference Recordings, Chesky, vinyl, CD, SACD, I really tried my best to find a problem, and the problem just did not come up.

I did send some notes to different companies asking all the questions, and when I hear back I will shoot you the results. Most of these companies do not get in any hurry. The Ribbon tweeters just add another dimension, clarity, real clarity.

 

I have the same experience and I wonder if the comb filtering issue is not overrated?

For starters, if the height of my ears is exactly between that of the tweeter and supertweeter the signal should arrive at the same time at both ears. Even if one is slightly closer, say the tweeter at ear height and the supertweeter 20cm above ear height they should still arrive with 1ms. There are many speakers that use 2 woofers on the same speaker and they would usually be sitting below ear height, hence the signal of both woofers arrive at slightly different times at the ears.

The same goes for the the stereo set up. For example, listening to a (mono) record I don’t have to be exactly on the spot between the speakers to get a good sound. I mentioned in another thread that a single coaxial mono speaker in front of you would be ideal(, but it is just not practical).

This site has a sound example for spoken voice with 0 to 50ms delays. https://audiouniversityonline.com/comb-filtering-explained/
Even at 10-20 ms it is not really awful, and the voice is still quite pronounced. And this is in the important frequency range of a voice and for the worst delay time of 10-20 ms, whilst for super tweeters we are talking very high frequencies only, just on the edge of 1 ms of it becoming an effect.

Furthermore a music signal is not just a pure single sine wave, where a second identical delayed one simply adds old deducts from the signal, but consists of many harmonics and a mix of other signals.

I know the science says phase interaction exists and one should be aware of it and try to limit it, for example keep an equidistance between 2 speakers for serious listing. Also, Aperion mentioned that they tried different roll off filters to minimize the phase effects. As far as the supertweeter seems to be concerned the huge benefit of elevating very high frequencies seems to outweigh the potential (minor) phase effects.