Beyond the Ariel

~~~~~~~~~~~~~~~snip~~~~~~~~~~~

To be honest, I would rather have a system that makes excellent performances sound musical and beautiful than one that makes audiophile recordings sound "spectacular." My definition of musical and beautiful is that I should be able to hear everything that is happening in the music (balances not obscured by glare or other artifacts), and that the tone colors of instruments and voices should be realistic, which to me sounds beautiful in real life. The Rethms were operating wide open at the top, with just a HP filter on the RAALs, and the results were fabulous. Lynn ended up with a pair of these tweeters for himself, in addition to the new 9" Lazy Ribbon RAALs. I only have the Aurum Cantus tweeters, but experimented with both frontward and rearward facing. The results have left me really tempted to try the RAAL dipoles, but my wife is thinking I have already spent more than enough on this project. Maybe I can work something out here...we'll see.

I can't say that I have arrived at the final destination yet, but I'm definitely well on the way.

Gary Dahl

UFF DA!

I seem to detect the emergence of conflicting priorities. :D

Hmmmm...Should I listen to beautiful music, or my wife?

Gary, if it were me, I'd tell her to just......

Oh Cripes! I've got to go, my wife just pulled in the driveway and I haven't finished the dishes!
:scared:
 
Hello Gary,
I was wondering about this "musical" description. In my experience, lots of systems can sound musical, Goldmunds to name one, but may not sound realistic. With the effort you are putting in, I assume that you are striving for both.
Not to start any cable wars, but my personal experience is that cables can make and break a system performance. I am currently working on some interconnects, and will have some samples in about a month or so. Really would like to get your listening impressions based on your understanding of the music you are familiar with.
 
Thanks Gary for the update,

Knowing that you have been involved with live music for many years gives you an advantage in speaker design and fine tuning. I am looking forward to seeing the outcome from your time invested in this project.

Your enthusiasm for the new RAAL dipole tweeter has gotten my attention.
I really like dipole speakers, especially the way high frequencies are reproduced. I think it gives a little more live sound quality to the playback.
Please keep us updated with your speaker project and experience with the RAAL dipole ribbon?
I am wondering if the new RAAL dipole is efficient enough to mate properly with a horn mid-tweeter as you are using?

On a side note:

A friend of mine, Steve Brown has just moved back to his home town in Portland Oregon from Tulsa . Steve is friends with Gary P, and was involved in the Oregon Triode Society many years ago. I am hoping that you and the gang in the northwest will eventually connect with Steve, giving him a big welcome home.

Keep up the good work!

NW
 
Last edited:
One of the things that Gary and I were curious to hear at the RMAF was the JBL Everest DD66000 system, which has some design similarities to the new speaker - two 15" woofers with one woofer acting as a baffle-step compensator, large-format compression driver covering 700 Hz to about 9 kHz, and a supertweeter. The differences are beryllium vs aluminum diaphragms, different bass drivers, and a completely different horn, crossover network, and cabinetry.

The DD66000 was one of the few speakers at the show in the same general class as the Benaroya - this show seemed to have more pint-sized monkey coffins with giant transistor amps than previous years. The JBL had the expected "big" sound, but I suspect some of the things I heard in the midrange were due to the Mark Levinson 500-watt Class D amplifier (surely an odd choice for a 96 dB/metre speaker in a small hotel room). Gary and I were both convinced it really should have auditioned with a high-quality PP tube amp - the way the Japanese like to listen to them.

The new dipole RAAL we heard at the show made a big impression on both of us - that's some tweeter! The only tweeter that compares to these are ionic tweeters, in terms of extension, effortlessness, and spaciousness. Sort of like Stax headphones in terms of see-through resolution and HF shimmer. I'm pleased my ship came in and I have both the new dipole and the new ultrahigh efficiency monopoles.

The preliminary measurements of the Benaroya indicate it is very flat, and the AH425 and Radian 745P do not require notch filters or in-band equalization. The TD15M only requires a very modest (out-of-band) notch around 1.5 kHz, and that's to improve the smoothness of the phase transfer between the woofer and the midrange. Gary reports coloration levels are similar to the Ariel, but with a much more effortless sound. He's previously owned Altec A7-828's with the 311 sectoral horn, eXemplars, Ariels, Von Schweikerts, and Audio Notes, so he knows what large-format speakers sound like.

As for old "accurate" vs "musical" debate, I've yet to hear an "accurate" audiophile system that sounded anything like acoustic instruments played in a real room. The tone colors are either flat or weirdly altered, or both. The $tereophile and Absolute $ound Class A-reviewed systems have a mechanical, or artificial quality to the playback, which I don't find realistic at all - kind of like the colors you see on TV versus real life. This seems to be an area where transistor amps really struggle - I've heard a handful that have reasonably good tone colors, but most fall down pretty badly. Still trying to pin down why transistor amps sound the way they do, which was part of the reason I was so thrilled to see Bob Cordell's book, which reveals a lot of challenges of high-quality transistor amplifier design.

The conversation I had at the show with Rene Jaeger was also illuminating - nonlinear charge-storage effects are a big deal in the transistor world, thus the merit of cascoding and ultrafast circuits. Vacuum-tube circuits also have to account for stray capacitances, but the big difference is that vacuum-tube stray C's are linear, while transistor stray C's are very nonlinear.
 
Last edited:
Gary has now reached the point where parts are starting to be audible. I recommended he try the Jupiter wax-paper caps for the HF crossover - once the system is getting dialled into the 1/2 dB to 1 dB correction range, caps start to sound like EQ changes, as well as shifts in transparency. In terms of audibility, the series caps for the HF circuit are the most touchy, followed by shunt elements, parts in the woofer circuit, and notch filter elements.

As for wire, rather than insanely priced audiophile wire, with its absurd 80~90% retail markup, Gary Pimm, Gary Dahl, Bud Purvine, and I like industrial Litz wire and building our own. Braided Litz wire with cotton covering compares well to audiophile wire at silly prices, and corrosion becomes a non-issue with all of the individually enameled strands. The only real hassle is the requirement for a solder pot to terminate the Litz wire. I've looked around for silver Litz wire, but that's probably a little over the top.
 
Last edited:
~~~~~~~~~~~~~~~~~~~~~~Snip~~~~~~~~~~~~~~~

The conversation I had at the show with Rene Jaeger was also illuminating - nonlinear charge-storage effects are a big deal in the transistor world, thus the merit of cascoding and ultrafast circuits. Vacuum-tube circuits also have to account for stray capacitances, but the big difference is that vacuum-tube stray C's are linear, while transistor stray C's are very nonlinear.

Hi Lynn,

I'll be seeing both Rene and Bud Purvine tomorrow and I'll have to ask them about the stray capacitance issue. They've both touched on it in the past, actually more as an afterthought, during conversations.
Just the same, they're both incredibly knowledgeable as well as being extremely nice people, I only wish that I knew as much as they do.


Best Regards,
TerryO
 
Gary has now reached the point where parts are starting to be audible. I recommended he try the Jupiter wax-paper caps for the HF crossover - once the system is getting dialled into the 1/2 dB to 1 dB correction range, caps start to sound like EQ changes, as well as shifts in transparency. In terms of audibility, the series caps for the HF circuit are the most touchy, followed by shunt elements, parts in the woofer circuit, and notch filter elements.

As for wire, rather than insanely priced audiophile wire, with its absurd 80~90% retail markup, Gary Pimm, Gary Dahl, Bud Purvine, and I like industrial Litz wire and building our own. Braided Litz wire with cotton covering compares well to audiophile wire at silly prices, and corrosion becomes a non-issue with all of the individually enameled strands. The only real hassle is the requirement for a solder pot to terminate the Litz wire. I've looked around for silver Litz wire, but that's probably a little over the top.
Ahh, so there is still a journey to be travelled. I remember the old Monster Cable and MIT interconnects used very fine Litz wire, and multiple sizes. The Old MIT 330 Shotguns that I have used as reference for a very long time is that kind of structure. However, they do present some slight smearing of image and less clean sound. To date, I have not seen any measured data of cables and their variation due to different interfaces.

I was recently at a friends place to compare some interconnects. I was trying to decide on which to pick on the final configuration. I listened quietly as he switched the interconnects. After putting on one interconnect and listening for about one minute, he commented that if he made a choice, it would be that one, and explained why and why not. In the meantime, I did not like any performance because it seemed all the material he played were inverted polarity, So I last suggested that the polarity seemd inverted, and recommend we swap the polarity at the speakers. After we did that, the huge difference cannot be described by words. After that, he had a different interconnect choise.;)
 
...
As for old "accurate" vs "musical" debate, I've yet to hear an "accurate" audiophile system that sounded anything like acoustic instruments played in a real room. The tone colors are either flat or weirdly altered, or both. The $tereophile and Absolute $ound Class A-reviewed systems have a mechanical, or artificial quality to the playback, which I don't find realistic at all - kind of like the colors you see on TV versus real life. This seems to be an area where transistor amps really struggle - I've heard a handful that have reasonably good tone colors, but most fall down pretty badly. Still trying to pin down why transistor amps sound the way they do, which was part of the reason I was so thrilled to see Bob Cordell's book, which reveals a lot of challenges of high-quality transistor amplifier design.

...
It seems that most of the shows I've been to really don't take care to playback in the right polarity. Sad situation. Lot's of people just cover their eyes on this issue and wonder why they never hear accurate and musical sound. Even on a little EPOS speaker, the polarity is VERY obvious.
 
The conversation I had at the show with Rene Jaeger was also illuminating

Boy, do you want to come out for a visit to the upper left coast. Rene' has a frankenstein of an amp that is the first amplifier I have heard that rivals the Karna or Gary Pimm's Tabor SS model 2. And guess what...... it uses a 300 B SE tube..... in the most unusual configuration you could imagine. It does also have the level 3 OT's from the Karna.....

No more will pass my lips till Rene' get's all of his commercial possibilities ironed out. He is right there in your class Lynn and I am delighted to hear that the two of you got to talk.

Bud
 
Last edited:
I bet you I can change that in a positive matter.

Sure - in case it acts to compensate for something lacking.
But what I wanted to point at is that silver IMO "adds" to the sound in terms of having a distinct coloration / sonic pattern.

I find your attempt to get a handle on cables interesting - but unless you find a way to clearly document differences by measurements and put up a concept you will always be locked in the "case by case tweaking" department.

As for "absolute polarity" - people are not equally sensitive on all aspects of audio - Lynn has checked for himself and found it to be widely a non-issue (for him) - good for him, I'd say.
We possibly should not conclude that he (or others that have checked and do not care) "never hear accurate and musical sound" - though absolute polarity seems to be an "issue" for you and me - and others too.


Michael
 
Last edited:
Yes, there are some aspects of cables where measurements can be documented against preference, but some not quite yet. I will show some aspects when it's ready. Some preference/characteristic relate with the stucture of the interconnects. When I mention preference, it's not just a decision, but rather characteristics that can be independently identified as preferable.

The importance of absolute polarity mostly depends on speaker performance and type of recorded material. However, I also did find that originally preferred to play from iPod with reversed polarity, but after tuning the interconnects, normal polarity reveals better overall performance. What I can only say is that sometimes certain things are discovered by chance out of curiousity. Once you have discovered a difference that is significant enough, then you can notice it almost any time.

As far as silver material in interconnects, it's not the material alone, but how the different material for a same cable structure alters the cable characteristics. This is why you see that when you look at RG-U cables, they not only have dimensions specified, but material is specified as well.
 
Last edited:
As for "absolute polarity" - people are not equally sensitive on all aspects of audio - Lynn has checked for himself and found it to be widely a non-issue (for him) - good for him, I'd say.
We possibly should not conclude that he (or others that have checked and do not care) "never hear accurate and musical sound" - though absolute polarity seems to be an "issue" for you and me - and others too.

Is there even a widely accepted definition of "correct" polarity, when it comes to mixing recordings? I've built all my own amps and speakers. I check the polarity so that the outgoing waveform matches the input. I check the individual drivers (without the crossover) so that a positive input moves the cone forward. After that, I'm at the mercy of the recording. Not too interested in flipping a polarity switch for every different recording, or even every cut.

Sheldon
 
Is there even a widely accepted definition of "correct" polarity, when it comes to mixing recordings?
Sheldon

Don't think so - especially as there are many XO with reverse polarity as well - so its a mixed bag anyway.

Despite that, there is some reality behind in a physical difference regarding compression versus expansion but its pretty sublime.

Michael
 
Gary Pimm and I did some cable auditioning on his system a while ago. I should mention that GPimm's system was/is fully balanced and transformer coupled, so cable swaps are done with XLR's and do not make pops or clicks even when the volume is advanced. (Transformer coupling does not store DC charges the way cap-coupling does, and there's no make/break problem with XLR connectors.) This allows very quick and rapid A/B/A comparisons without the distraction of clicks and bangs - which have a way of resetting audio memory.

We didn't audition commercial cables - no real interest in them, and without disassembly, no way to be sure what's inside anyway. Instead, we compared various strands of Litz wires with cotton insulation, going from 1 strand, to 2, 4, 8, 32, and 64. Since this was a balanced, line-level interconnect we were comparing, I wasn't expecting to hear much difference at all - the linestage had oodles of current to drive the modest cable capacitance, and all we were comparing was various types of Litz wire.

To my surprise, as the strand count went up, the sound become quite noticeably more coherent and focussed (no tonal difference). Single-strand, in several different gauges, sounded diffuse and incoherent. Two strands, a bit more focussed, four strands, more so, and things leveled out beyond 32 and 64 strands. Rather counterintuitive.

First off, I was expecting no difference at all. Second, if any difference was expected, I expected more strands to sound less coherent, with all those different propagation paths, and no crosstalk between strands thanks to the individual enameling on each Litz strand. But no; 32 and 64-strand were quite obviously more transparent than any single-core wire we tried, and most unexpectedly, more coherent, despite the way things looked. I didn't expect that plain old industrial Litz wire (which has been around since Tesla) with cotton sleeving would sound as good as fancy audiophile cables.

This test would have been a lot more awkward with RCA unbalanced circuits, requiring a muting of the power amps, and remembering to never forget to hit the mute switch before changeover. I know from experience that a single loud bang or system-breaking buzz from a thoughtless RCA swap pretty much wipes out a listening session for me.

At the RMAF show, Keith Johnson mentioned that any acoustical annoyance wipes out high-resolution listening for about 15 to 30 minutes, due to remapping of auditory pathways in the brain. High-resolution listening is only available when the listener is relaxed and in a pleasant, upbeat mood; any degree of stress, even momentary, pretty much wipes it out, and recovery takes a surprisingly long time. If KJ says so, I take him seriously - it's not everyday I meet a Ph.D. physicist that used to teach at MIT, was the prime engineer for the Reference Recordings series, and is the co-inventor, along with Rene Jaeger, of HDCD. Keith Johnson and Rene Jaeger are two of the smartest people I've ever met in audio, or any other field.

Anyway, I'm a bit skeptical of high-end cables - a lot of them seem like audiophile tone controls to me, with a designed-in "house sound" that is specific to that brand. If I want a tone control, I can always rebalance the crossover slightly - it's not that hard. There also seem to be commercial pressures to make them really complex and big, perhaps as a way to justify the astronomical pricing, and I suspect the complexity works against them.

I don't see any reason why a simple twist, or maybe a braid, with a mostly-air dielectric like cotton sleeving, can't be a simple and effective way to make a cable. The holy war against inductance in speaker cables strikes me as really weird, since tweeters have plenty of their own inductance, and this cannot be removed from the tweeter. I'm more concerned about capacitance, the deadly enemy of feedback amplifiers in the 100 kHz to 1 MHz region. The twist or braid reduces RFI pickup, which is a good thing for any amplifier.

Please don't take my experience as gospel. Cables tend to be very system-dependent. What a 200-watt Class AB transistor amp likes isn't going to be the same as 20-watt DHT amplifier with zero feedback. I do get a bit suspicious when an amplifier with a lot of global or local feedback sounds radically different with different cables; I then suspect HF instability or RF pickup getting into the input node of the amplifier.
 
Last edited:
As for audibility of absolute phase (changing both channels at once), well, for me, it depends. With fully-balanced transformer-coupled amplifiers like the Amity or Karna, there's a three-position rotary input switch that selects +IN, Mute, and -IN. Switching is silent, with no DC present on the switch contacts, and Mute centered between + and - inputs. So it's really easy to make a switchover with the amps up and running.

With the Ariels, with their 2nd-order allpass crossover at 3.8 kHz, I hear a noticeable, but still fairly subtle change in timbre on certain tracks when the absolute polarity is changed. I've met audiophiles that bounce out of their chair and exclaim, "Did you hear THAT?" - I don't have that reaction. My only guess is that some people are much more sensitive to absolute phase than others, and I'm not one of them. Weird electronic and speaker-driver colorations both me a lot more.

The new speaker might be better at square-waves, since it has driver offset as part of the crossover design (that was not an option for the MTM Ariel). Perhaps absolute phase will be more audible on the new system, although I grant running up and changing a switch can be a hassle, since many recordings have absolute phase that varies from track to track, depending on which recording studio was used.

Maybe for the ultimate in audiophile convenience and laziness, the phase-switch can be put on a remote control, along with volume and a fine-grained balance control that has 0.5 dB steps. That way, the system can be optimized on a track-by-track basis with the greatest of ease (yes, there are many recordings with slightly off-center vocalists).

While we're at it, why not bring back tone controls? I'd be happy with variable-inflection bass, a moderate-range mid control, and a variable-shelf treble control. On the remote, of course. Wouldn't it be entertaining to have a virtual front panel of a McIntosh C22, Fisher 400C or Quad II preamp on your iPad or Android OS device?
 
Last edited:
Is there even a widely accepted definition of "correct" polarity, when it comes to mixing recordings? I've built all my own amps and speakers. I check the polarity so that the outgoing waveform matches the input. I check the individual drivers (without the crossover) so that a positive input moves the cone forward. After that, I'm at the mercy of the recording. Not too interested in flipping a polarity switch for every different recording, or even every cut.

Sheldon
The same company generally has the same polarity probably 90% of the time. Systems today seem mostly non-inverting. I do recall that DIN requires positive pressure to cause positive voltage at pin 2.
 
Lynn, did the total conductor cross section change as you added strands, or was the cross section for each strand constant, or was both changing? It certainly sounds like old MIT or Monster interconnect structure.

In the development of cable I am doing, which is unbalanced, we did some comparison against a Van Den Hull Jubilie balanced cable on a system that could switch inputs by remote control. I could not here any difference, the other person that was listening though he heard that my cable had a sligthly better bass control, but we both did not feel that the difference between the unbalanced cable and balanced cable were of any concern. There is one different step that I do, is that I have to tune the cable and measure, then tune, this goes in a cyclic manner until I have what seems the best reproduction of sound. Then a reasonable explanation is resolved with some general criteria and specs for the final characteristics are generated.

Got to go now. Will be back later.
 
This effort sort of branched from an active speaker project that is being done.
The general thinking was influenced by a Sumiko OCOS report published probably in the 80's. The difference is that I consider cable as part of the interface, and thus measurements are conducted with various load impedance to ensure that a cable can perform well under a range of load conditions.