This system is a culmination of many years of realizing what I value most in music. With audio, the only way to move forward is to know what you want and then work your way to achieve your goal. With that in mind, every component in this system was chosen to deliver what I value most in music. I strongly believe that the first step in any system is to choose the speaker that delivers what you want, and then choose all upstream electronics to make the speaker work.
The Acapella Violon delivers what I want. The plasma tweeter is simply without equal. It works by ionizing air from a high voltage electrode so that the air is sufficiently energized to erupt into a ball of glowing plasma. The size of the plasma flame is modulated to produce sound. No moving mass, no inertia, no coloration from cone material. This tweeter sounds like nothing. There is no metallic ring like you would get from a metal dome tweeter, and no softness associated with textile tweeters. Sound simply emerges fully formed, the speed and dynamics is unmatched. It just sounds real. This is without doubt the best tweeter technology in the world.
The midrange horn very nearly keeps up with the tweeter. It covers a large range from 500Hz-5kHz with shallow (-6dB/oct) slopes either end. The dynamics from the horn match the tweeter, and it excels in delivering instrumental timbre and loads of inner detail.
The weakness in the system is obviously the woofer. The overall quoted sensitivity is 91dB/W/m, and I am certain that the horn and tweeter were padded back to match the woofer. The drivers are 10", and arranged in a pseudo-isobaric configuration. Strangely, the external woofer is run from the crossover (500Hz, -6dB/oct), but the internal woofer is run directly from the speaker binding posts - i.e. full range. The bass is muddy, sluggish, and there isn't much of it. It obviously sounds disconnected from the horn and tweeter.
Furthermore, the speaker exhibits a wicked impedance curve with 600 Ohms for the tweeter, 8 Ohms for the midrange, and down to 2 Ohms for the woofers. This further exacerbates bass problems when run with valve amps.
I was obviously aware of these problems when I bought the speaker, so why did I choose such an obviously compromised speaker? The answer is - I strongly believe that midrange and treble should be gotten right first. If your midrange and treble are not right, you will spend endless sleepless nights trying to fix a problem which may not be fixable. Bass, in contrast, is relatively easy to fix.
I am about to embark on an adventure to fix the bass on my speaker. The first step is to purchase a solid state amplifier to bi-amp. The next step is to purchase a DEQX to control the phase response and group delay of the individual drivers. After this, I will remove the crossover in the bass unit (converting the speaker to active), and change the woofer to a more suitable driver. Finally, I will build a custom crossover to go before the midrange/tweeter, which will prevent the ADC/DAC stage of the DEQX from contaminating the all-important midrange and treble.
Last but not least, I will return to an analogue source, but such a purchase is still very much on the horizon.
Watch this space.
Previous entry ---
Thanks to AudioGon, I became aware of this speaker. After salivating at some of the nicer Acapella setups here I had to go have a listen. Boy, was I captivated. The most revealing speaker I have ever heard. The plasma tweeter is something else!
I am currently also upgrading from the Cary CAD-805AE to the Cary CAD-211AE, which should allow me to control the Violons a little bit better.
I named this system "Southern Lights" after Eliotswede's Acapella Violon system ("Northern Lights"). After all, i'm in Australia!
-- Old description below --
Years ago I became aware of the sweet combination of Cary amplification and ProAc speakers. It has been my intention for a long time to replicate this system and I finally got the opportunity to build my dream system from scratch.
3 way speaker with plasma tweeter, horn midrange, and conventional woofer. The most astonishingly natural sound I have ever heard. Sounds seem to emanate from thin air, free of distortion and coloration. This picture was taken next to my old ProAc D38's.
Cary Audio Design SLP-05
This is an outstanding preamp with faultless sound. It takes forever for the tubes to run in. The first time I powered it up it was shockingly harsh and unmusical, but this settled down after 24 hours and is now gone.
Cary Audio Design CAD-211AE
110W of push-pull triodes. Incredibly grunty.
Marantz RC-9500 universal remote
The HT and 2 channel systems are tied together with this programmable remote. One press of the button starts a macro which ensures that all the correct settings are chosen.
Acrolink Mexcel 7N-S9000
Simply unbelievable transparency
Acrolink Mexcel 7N-A2070 XLR
Astoundingly transparent. With my last cable I was wondering where the midrange had gone. Well, now it's all back :)
Microseiki BL-99v
Micro-Seiki BL99V turntable with vacuum hold-down.
Lyra Lyra Dorian
Lyra Dorian cartridge
SGR EL30S
SGR EL30S solid state stereo amplifier, 300W. For bass duties only.
JL Audio F110
Pair of JL Audio F110 subwoofers.
DEQX HDP-3 preamp/processor
The DEQX HDP-3 is used to measure the system and will be removed from the loop as soon as development is done.
I don't think that the midrange driver relies solely on the natural band pass offered by the horn so the resistor must be within the crossover network located inside the enclosure that the horn is attached to. I have no idea how to get to it. I wish I could help but I never opened up my Violons.
How about going a step further and removing the padding resistor on the midrange horn? The midrange horn in the Violon is heavily padded to bring the efficiency down to the level of the original woofers (91dB), however it is capable of at least 94dB efficiency (the same midrange driver/horn is used in the 94dB efficient Triolon). By removing the resistor you would end up with a high efficiency system that can be easily driven by a 300B or 45 SET. A properly executed low powered SET amp should sound much better than your Cary amps. Crossed over at 600Hz it will be driving a limited bandwidth so even a 45 should be fine.
A few months ago I heard the CAD 211AE's on some custom built 95dB efficient horns. The sound was very good. The owner then switched to a 2 wpc Yamamoto A-08 45 based amp and there was simply no going back to the 211's: the Yamamoto was so much more tonally refined, transparent and alive, simply magical. The Yamamoto is significantly cheaper than your Carys. Just some food for thought.
Have you considered not using the Marchand as a low pass for the horns and instead using it only on the woofers to match the high pass on the horns (probably a 12dB slope)? In your current configuration you have two crossovers (the Marchand and the passive inside the horn box) in the signal path to the midrange horn and the tweeter. The Marchand is pretty good but I'm worried it may not be completely transparent. Just some food for thought...
Hi Keith, What slope did you end up on the 600Hz high pass: 6dB, 12dB 18dB or 24dB? You mention that the SS amp is hooked up directly do the driver. Your speakers have two woofers per cabinet. Is the internal woofer still hooked up? Did you experiment with using both woofers vs. a single one? I would think that there would be no benefits of using the internal woofer now that everything below 80Hz is routed to the JL Audio subs.
I would be very interested to see your measurements.
All the additional analog to digital and digital back to analog conversion that the DEQX performs simply cannot be without an impact on the sound. The loss of resolution and dynamics that you're describing is thus something to be expected when the DEQX is fed with an analog signal.
I think that the ideal way to experience the full potential of systems such as DEQX or TacT is to feed them with a digital signal, let them handle all the necessary crossover, time alignment, phase adjustment and room correction functions and then use their digital outputs to feed quality external DACs followed by a nice multichannel preamp (in your case four channels would be required) for analog volume control. Of course this way you would not be able to use SACD or analog sources. Based on my recent experiences however, the future belongs to music servers. With the increasing availability of high resolution downloads I just can't imagine SACDs surviving for too much longer. If, however, you're thinking about getting an analog rig, then I don't think you would ever be satisfied with the DEQX system. There are other options though. In my system I am currently using an analog parametric EQ device made by Rives Audio (PARC). I am bi-amping my Violons (Audio Aero Prestige SETs on the midrange and treble and Belles 350A Reference on the bass), and the PARC is inserted between the preamp and the bass amp, thus having no impact on anything above 500Hz. Rives claims that the PARC introduces no intrinsic time delay but I haven't been able to verify it. All I can say is that in my rather difficult acoustically room, the PARC brought huge improvements in bass definition and articulation. The bass still doesn't quite match the amazing speed of the midrange horn and the plasma tweeter but it is pretty good.
With the right electronics the Acapellas' midrange and treble are the best I ever heard and although bass is not as outstanding as the rest of the spectrum, I could happily live with it. I think that part of your problem lies in the electronics and you're hearing too much of the Cary signature sound in the bass. I used to own the Cary SACD 306 and although it is a very good player, I found it's bass slow and muddy compared to players such as Wadia 581 or EMM CDSA. The Cary pre doesn't have a world class bass either and I think that everybody would agree that the Cary 211 amps cannot quite deliver what the dual 10" woofers need.
Your experiments with the DEQX are very interesting! Please keep us posted. Considering the pros and cons of the DEQX system, it would seem that the best approach would be to have it in the signal path in front of the bass unit only but I am not sure whether it would serve its intended purpose in this configuration. The DEQX by itself introduces a few microseconds of delay, which is the opposite to what you would be trying to accomplish. To time align the midrange/tweeter with the bass driver, you need to introduce some group delay to the midrange/tweeter section, which means you would have to use the DEQX in front of horns.
