I think I've made up my mind

After having auditioned the following :

1. Avantgarde Uno G2 + Audio Note SET amp/Pathos Twin Towers integrated amp
2. Magico V2 + Soulution 720 pre + 710 power amps
3. Sonus Faber Elipsa/Wilson Sasha + ARC Ref 3 + Ref 110
4. Thiel CS3.7 + Parasound JC1 pre + JC2 power amps

...and after weighing all factors like room interaction, degree of difficulty of setting-up, value for money, magazine reviews, and comments posted here and AVguide.com....

...the winner is: Thiel CS3.7 + Parasound JC1/JC2!

Now for the cables. MF and JA of Stereophile noted that the specs of Parasound show ultra-wide bandwidth and high-current capability which reminds me of Spectral amps. Therefore, would MIT Magnum MA speaker cables and interconnects be a good match? Anyone has any other suggestion? Thanks.
That was a pretty strong group of contenders. Why the 3.7s?


PS - I'm considering the 3.7s, Sophia 2s and the Studio2s.
As I said before, the Thiel's are great speakers. If you value soundstage, timbral accuracy, overall clarity, and value, then this is definitely a top choice.

If you have the opportunity to listen to the Avalon Indra before your purchase, that would be one last speaker to consider. The driver integration is more cohesive and refined on the Avalon. I went from Thiel to Avalon, and it is a very logical progression.

Also, don't rule-out a tubed preamp to go with your amp.

Finally, I would recommend you get some inexpensive cables to start. When the speakers are broken-in, demo several cables to fine-tune the sound. Cables are not a band-aid; they are the icing. Cables are very system and user dependent. You've learned that reviews are a guide and not the answer, so don't get caught-up in marketing concepts. Always use your ears when investing significant money.
BTW... were these amp + speaker combos you mention connected up for demo purposes when you auditioned pieces?

or were all listed items ala carte... here and there pieces that you heard separately?

.... just wondering.

If so, then what cables were in play when you chose the combo you seek out now... 3.7 + JC 1's?

that might be your answer too.

I liked the MIT Magnum ICs, myself. Though only for spot duty... on the source. I found more than one link of MIT M's was too too much and the sound lost it's moisture and began heading towards less vibrantly colored and more towards a pastel of the original timber & tome. you may feel otherwise however.

Again, recalling those wires that were in place when you heard those speakers might aid you a lot here.

From what a friend of mine says about the 3.7's... the previous post is onto something here pointing out the possible addition of a tube preamp for those Halo's & Thiels. With as much control as the mono blocks ought to provide them, perhaps trying Nirvana SX or HT Magic IIs ICs should be a very good thought IMHO.

Good luck.
Wow,you got to audition all that different equipment together.You are indeed lucky.Jim brings up some very good points in his post.
Ritmo, you can find my reasons in my other threads: 1. Let down by Wilson Sasha/ARC Ref 3 combo 2. My findings on Magico....what now?

Rtn1, I have considered the Avalon Indra but have no chance of auditioning them as no dealer in my country carries them. I've inquired from a dealer in a neighboring country and the cost of getting the Indra's is prohibitive (about $25k which is $5k more than retail) because of the shipping charges due to the very heavy wooden crates they come in.

Blindjim and Tpreaves, they were set up for demo purposes although at different dealers over the past few months.

Thanks to all for the helpful suggestions on the cables.
All very good stuff. Looks like you considered a nice variety of system types before deciding, which is very smart.

Two things worth considering I might add:

1) A tube pre-amp might work very well with that amp!

2) Did you consider OHM Walsh speakers by chance? Might be worth taking advantage of OHMs extensive in-house trial period first if you might be interested in perhaps halving or better the cost for the speakers, depending on room size. Sound should compete well with the Thiels I would think however styling and appearance is way different.

How many of you believed Spectral when they said their amps would go into oscillation if you didn't use MIT speaker cable?
They said that? Great marketing if your looking for an unstable amp.
In my humble opinion, the Ref 3 is an extremely dull sounding preamp. I wouldn't trust any speaker evaluation with it.
"Extremely dull" sounds extremely extreme.
if you do invest in a cheap cable and then compare it to a more exotic. Don't be surprised that you just might stay wit the inexpensive cable because it is that good.
Pubul57, you have a point. Perhaps I should have left it to one adjective 'dull' and not try to overemphasize it.
Did you mean it is not dynamic or that it does not have high frequency presence?
I agree with Madfloyd. When I owned the REF3, and used it with Pass Labs XA-60.5 amplifiers, the resulting sound was bland, bland, bland.

The REF3 did not come close to meeting my expectations based on all the fawning reviews.
Tomorrow might be a good day to buy a used Ref3:)
Madfloyd just curious what tubed pre-amp do you prefer in the price range of the REF3? I have LS26 and would like to upgrade to the REF3, but am open to suggestion.
Can't speak for the Parasound components, but I've had good results with Alpha-Core Goertz, StraightWire and WireWorld speaker cables with various Thiels with various amps.
could you, please comment on what aspects of the sound of Thiel system, you preferred to Avantgarde Uno G2?
I understand, you liked Avantgarde in the first audition, and Thiel is so very different, just the opposite of that presentation.
Congratulations on the 3.7's! I have owned 3.6's for 10 years and would lean towards a 3.7 if I ever got the serious itch to move to something slightly different. However, I have auditioned many speaker systems and have learned once the Thiel sound gets in your blood its hard to move on!

I have auditioned and owned tons of cables, more than I would like to admit. Although they are system dependent (you can view my system here) I have found Analysis Plus Pro 9 SC to be a match made in heaven for me. They allow my system to deliver a coherency from top to bottom with a full and satisfying midrange - something many speakers cables cannot achieve with Thiels. They are also virtually impossible to beat for the money. However, all tastes are different, systems are different and cables while a vital link/component are system dependent, you will have to piece that last link together with auditions. Hopefully some of the advice on this thread will help get your further along than scratch - Jim and Unsound both have great suggestions. Jim nailed the Magnum description from my experience and I have used both Goertz and Straightwire and lived with both for a while. Goertz and SW are also great bargains in the performance/price ratio. But as you know there are a lot of choices! Good luck.
Rrog said: > How many of you believed Spectral when they said their amps would go into oscillation if you didn't use MIT speaker cable?

I did and do, and it's easy to demo - they run considerably hotter w/o MIT cables. That's because - if you cared to look into it - the amps offer no input choke or output coil, leaving those functions to the cables, and therefore the MIT/Spectral cables are specifically and specially spec'd.

What's your point?
Maril555, I still like the Avantgardes. They are very revealing but in an effortless way. Vocals are very well served by the Avantgardes. I find that the Thiel CS3.7 matched them in that respect. The downside of the Avantgardes is the bigger than life but rather narrow soundstage. Various instruments seem to accumulate together. Also, when listening to the Avantgardes, I feel like I'm at the front row of a concert where I have to cock my head up to watch the performance. It's thrilling but can be fatiguing for long listening sessions. I find that I get very realistic soundstage and imaging with the CS3.7. Finally, I still prefer passive speakers to active ones.

The truth is they don't want you to screw up the sound with a cable that doesn't mate well with their amp. So, they tell you to use MIT or it will void your warranty.

I have heard Spectral with many other speaker cables and the amplifier worked just fine.

Maybe more companies should do this to ensure satisfied customers.

My point is; what's at the end of the speaker cable?
Mad floyd FYI any tube pre amp with a highish out put Imp will be dull into a 10k or 15 K input imp
Cheers Johnnr
John, that would cause a high frequency roll-off, no? I'm not sure if that is what is meant here by dull, but the Ref 3 certainly sounded dynamic as heck, but that was into a 100kohm load, I think Pass amps need SS preamps or tube pre with 500ohm or less Zout.
Rrog: > So, they tell you to use MIT or it will void your warranty.

Nonsense. I suggest you start by understanding SS amplifier designs, especially those with extremely wide bandwidth and no chokes, then you will perhaps understand that the MIT/Spectral cables limit that bandwidth to something manageable in the range of 800kHz, that keeps the transistors within acceptable operational thermal limits.

The amps will not die immediately without the MIT/Spectral cables, but long term they will cease to operate - has happened way too many times to those that think they know better.
Rrog: > My point is; what's at the end of the speaker cable?

The missing output inductor, with a value that compensates precisely the length of the cable purchased that carries its own RLC characteristic values. In the end, they limit the bandwidth of the amplifier to 800kHz (-3dB) and they optimize all electrical characteristics from the input cables to the tip of the speaker cables.

Spectral could have easily incorporated bandwidth limiting in their amplifiers and allowed the use of other cables. So, what do you think the reason is for the MIT speaker cable requirement? Could it be money from MIT for telling their customers they have to use MIT speaker cables? They are also telling you to use MIT interconnects.

This is an example of where highend audio has gone wrong.

Don't get me wrong. Spectral makes very fine equipment. I just don't agree their philosophy on some things.
Rrog: > Spectral could have easily incorporated bandwidth limiting in their amplifiers and allowed the use of other cables.

NO. Per my previous response, the inductor in the cable box optimizes the electrical RLC characteristics of the LENGTH PURCHASED. No fixed inductor in the amplifier would ever achieve perfect matching with any but one cable of a one particular length.

They like to optimize the entire chain, it's as simple as that. Same goes for the interconnects. If you don't like, don't buy their products.

Rrog: > This is an example of where highend audio has gone wrong.

I think this is where your analysis falls flat.

Ok, tell me how this benefits you in your system and what the advantage is over other amplifiers.
Great choice Jtein. I love those speakers.

BTW - What country are u located in. I have been reading your posts and see that you have travelled great distances to audition different systems.
Rrog: I and others care about an optimized set up. It's that simple. Lots of other people want to play with cables and tweak the sound to their liking and/or add or subtract "things" to the sound for better match in their systems - that's a valid approach AS WELL (but not necessarily optimal). They may even get "better" results. None of this implies that what Spectral is doing is wrong or not based on solid engineering principles - it's just not to some people's tastes and goals. I personally have no qualms about the performance of the top MIT/Spectral cables, and am glad I don't have to overspend on cables. In fact, I have compared them to higher-priced MIT cables and still find them superior (that could also mean that MIT is generally producing colored cables, and I am a strong believer of that).

Advantages over other amplifiers? I haven't analyzed other amplifier designs, other than to observe that other edge of the art SS gear, like the Soulution, are also wide bandwidth designs - wide bandwidth achieves (or aims to achieve) extremely fast circuits, which theoretically follow the signal more closely - the sound Spectrals and Soulution achieve is probably testament to that. I don't know if these other amplifiers filter the output inside the amplifier or not (and an output coil is very easy to spot but I have not seen close-up pictures of the units), but I have also observed they include ventilation fans, so there is a possibility they don't filter but achieve optimal thermal operation by blowing air in and out of the unit, which would let one play with cables - that's fine too.

Another edge of the art vendor, FM Acoustics, also believes in fine-tuned amplifier/speaker interfaces. You should read http://www.fmacoustics.com/pdf/audiophilpoweramp.pdf and take notice of a couple of claims:

1) "One of the most important characteristics that determines the quality of the entire system is the interface between amplifier and speaker."
2) "In amplifier-speaker interfacing a variety of criteria have to be observed, and the ideal cable has an optimal mix of these criteria."
3) "The criteria that has the most significant influence on the transmission quality of cable - the transfer characteristics - is often neglected for some other - often visual - aspect."
4) "As no wide-band width cable of this calibre is available as standard, engineers at FM ACOUSTICS in co-operation with the world's leading experts on cable technology have designed a unique cable that features a variety of proprietary characteristics. FOORCELINES guarantee optimal signal transfer, truly highest damping
and perfect control of the speakers."

So there you have it - another manufacturer that uses specific, optimized cables for their needs.

The bottom line is that what Spectral, FMA and others are doing is nothing but *optimal implementation of transmission line theory*. To everyone else, it's just a cable that can act as a tweak. Personally, I have been sold on wide bandwidth designs and transmission line theory decades ago.

You may still think Spectrals are not worth the trouble - and that's fine with everyone. I am surprised, though, that you brought up the cables as being the limiting factor and not the requirement for their preamps (and a lot of that has to do with stability into the MHz, proper output impedance and ample amperage to drive the amps), except for the inferior "S" versions of the amps...
What about Transparent's higher end cables;I have music wave supers with soundlab m2's and audiovalve challanger 180 monoblocks;seem to do quite a nice job for me.
The supers are the middle of the lineup I can only think the ones above must even be better;just my opinion though;enjoy your system.
Every manufacturer of electronics worth their salt should have optimally tuned their products with a single kind of cable (or a range of cables that are extremely similar to each other), and they should be able to tell you which one. I have no experience with tubes and extremely little with Transparent - I used to run a McCormack DNA-1 with some $800 Transparent speaker cable over 15 years ago, but honestly I did not like the construction inside that box (looked very sloppy). Basically, I can speak with confidence with products I have used since then.

The people manufacturing tube amplifiers don't make a big deal out of speaker cables. They just say to choose an efficient cable and keep the runs short.

I keep thinking I want to get back into solid state, but good solid state sound is both complicated and expensive. Then there is this issue with Spectral and MIT which I am convinced is a marketing ploy. How convenient, you don't have to experiment with cables because it is already done for you. I still think Spectral did this so their customers would be locked into a sound they approve of.

It's easier to get good sound with tubes. Besides, solid state manufacturers continue to say their latest amplifier is the most tube-like product yet.
Yes indeed you don't have to experiment with cables - it's done for you. What do you think we as consumers are doing by experimenting with cables? We try to find synergy. Well, synergy is also an electrical engineering science that can be calculated. It's that simple.
Tvad and Madfloyd,
I used to have the ARC Ref 3 before I sold it for the dCS Puccini which doubles as a digital pre-amp. It was driving the Bryston 28B SST monoblocks and was certainly dynamic with huge dynamic contrasts. So if it sounds bland or dull, it is quite likely the impedance mismatch with your amps, as suggested by Pubul.
04-09-10: Jon2020
...if it sounds bland or dull, it is quite likely the impedance mismatch with your amps, as suggested by Pubul.
I used the REF3 with Pass Labs XA-60.5 running balanced. The input impedance of the XA-60.5 in balanced mode is 30K ohms. The REF3's output impedance is 600 ohms in balanced mode. Even allowing for high output impedance spikes from the REF3, the impedance match was well within range.

The match may be within range but 30K ohm is still towards the low end of the SS amp input impedance range from what I've seen, which seems to be around 10K ohm. I've seen balanced input impedances on SS amps up to 200K ohms or so.
Well, I guess we can say the ARC Ref3 may be dull and bland with Pass XA.4 amps, and that might very well be true with that combination. I've just heard the preamp sound anything but dull and bland which makes me think it is not an inherent quality, but like much else a question of matching components.
The match may be within range but 30K ohm is still towards the low end of the SS amp input impedance range from what I've seen, which seems to be around 10K ohm.
Mapman (System | Reviews | Threads | Answers)
I don't understand this comment. In the case of the Pass XA-60.5, which is the only observation I am making in this thread, 30K ohms is well above the minimum required to match a preamp with an output impedance of 600 ohms.

John Atkinson's test measurements of the XA-30.5, which correlates to the 60.5, stated:
The input impedance of the balanced XLR jacks was 29k ohms, close to the specified 30k ohms; the input impedance of the single-ended RCA jacks was higher than specified, at 20k ohms. Both figures were constant across the audioband.
The important statement being "both figures were constant across the audioband".

Therefore, the 10K ohm input impedance figure you mentioned in your post is not applicable to my observation of the REF3/XA-60.5 impedance match.

I think I recall you indicating in prior discussions that you believe higher amp input impedance generally leads to directly observable improved bass performance, or something to that effect, which technically sounds viable to me.

In that case, 30K ohms is a lot closer to the low end for SS amps I have seen (about 10K ohms) than it is the max I have seen (about 200K ohm balanced).

It would be more of a factor for higher output impedance tube pre-amps than lower impedance SS pre-amps.
Mapman, what I have posted here are specifically the impedance numbers for the REF3 and XA-60.5, including test measurements from John Atkinson for the XA-30.5, which correlates to the XA-60.5 (comparing the specs on the Pass Labs website).

Atkinson measured an output impedance spike from the REF3 at of 1437 ohms at 20Hz in balanced mode. That spec still provides at 20:1 ratio into 30K ohms.

These numbers are well within the generally accepted range for impedance matching.

Nevertheless, if you have specific numbers that would suggest a non-optimal match, then I'd be interested in seeing them and learning more.

Certainly, my experience with the REF3/XA-60.5 match was that the sonic results were not the best I had heard.
"Nevertheless, if you have specific numbers that would suggest a non-optimal match, then I'd be interested in seeing them and learning more."


I'm only suggesting (as unless I'm mistaken I recall you have suggested to others as well in past threads) that if what you heard was not good, and a tube pre-amp was used, an amp with 60K, 100K, or 200K input impedance would probably sound different and might sound better than one with 30K input impedance, all other factors aside.
an amp with 60K, 100K, or 200K input impedance would probably sound different and might sound better than one with 30K input impedance, all other factors aside.
Mapman (System | Reviews | Threads | Answers)
It might. Not having heard a match with those specs, I can't say for certain.

The numbers of the REF3/XA-60.5 combination suggest there should not have been any mismatch whatsoever if one agrees with the thorough contributions of member Almarg (and others) on this topic.

Anyway, the horse is dead.


My recent move from an amp with 62K input impedance (Musical Fidelity A3CR) to one with a 100K input impedance (BC ref1000mkii) produced a difference in sound of the type and magnitude that would seem to support this theory and the physics behind the theory makes sense to me.

That's not to say that such a change in a single spec assures better sound, but the effects may be somewhat predictable in the case of impedance matching when heard I believe, which is of benefit.

The change in sound in my case was a marginal but noticeable improvement in overall clarity, dynamics and detail. That is pretty much what I was looking for in going to a SS amp (class D in this case) with higher than average input impedance specs that appeared to be a safer if not necessarily better match with a tube pre-amp.

Whether the resulting sound is better or not is always a matter of taste and preference and other technical factors as well. Specs are a guideline to better sound perhaps in some cases (like impedance matching) but only tell part of the story.

I think one of Bel canto's rationals for the design of the mkii version of the Ref 1000s was to help assure an optimal match with more pre-amps, tube pre-amps in particular. The input impedance was raised from standard IcePower 10K input impedance in the original ref1000s to 100K (unbalanced)in the newer mkii versions, which is largely why I sprung for the much more expensive mkiis over the originals, to help assure optimal results with the ARC sp16 pre-amp in my system.
Since higher input impedance always appears to be better, why don't SS desingers provide 100kohm input impedances. I would think the problem is that to do so requires a tradeoff they are not willing to make, but I don't know what that might be -- anyone?
Ack, The discussion regarding the ARC Ref 3 and Pass amp is exactly what Spectral wants to avoid. I am in fact a believer in not only the sonic matching of components, but also the electric matching which is where the system is either make or broken.

Some (like ARC in their SS amps) and others do provide that.

Some do not.

Hopefully someone knowledgeable in SS amp design can chirp in and clarify what the tradeoffs are.

I guessing it has something to do with what is required to deal with distortion cost effectively then in the amp itself (negative feedback , etc.) in conjunction with driving lower impedance speakers effectively as well.

And/Or it might be that certain amps go for a certain type of sound that can be achieved more consistently with lower input impedance and a SS pre-amp.

I suspect part of the bottom line is just that no two amps work exactly the same way and synergy among components (pre-amp on one side and speakers on the other) then is always key as a result.
Pubul57, Perhaps the better question is why; don't high end manufacturers comply with the highly compatible pro standards?