Thiel Owners

silvanik

Good to see you and it appears that we have another hands-on, DIY, contributor. Looking forward in reading more about your capacitor replacement project. Perhaps you can join forces with Beetle, Rob and Tom.   You guys are awesome!    Happy Listening!

tomthiel - great suggestions about other caps to replace waiting for your full new 3.6 XO project, I'll give a try to this step of upgrade and will report here as soon as I'm able to build my clear and reliable idea about this modification.
Concerning the list of amps  suitable for the well known Thiel's "bastard" low load I can state that my McCormack DNA-2 is really a nice match, it's capable of 900 W RMS for channel into 2 ohms. Things I like most is the effortless velvet punch and the warm and clear mids/highs it can deliver, I'm very satisfied with it.

In my previous post, when I compared different filter order,
I didn't mean to say that higher order is bad and first order
is good. I think each filter order has its own strength and
weakness. High order filter objective gives you more "clarity"
but I also think high order subtracts the "musical" part of the music.
Higher order tends to give you a more "pin point" image production
vs. first order. Listening to the CS2.4, although it has a lot
of see through clarity of the soundstage, I think I've have heard
better image see through with other speakers, but the CS2.4 is
just more musically satisfy. I guess my English is not good
enough so I want to quote a Stereophile review of the CS3.7
review when he compared the sound of the CS3.7 vs. the
Wilson Audio WATT/Puppy 8:

"While the Wilsons did give Francois Couturier's piano in
Brahem's "Vague/E la nave va" a slightly more vivid presence,
I felt the Thiels did a better job of seeing into the heart
of the music. What does that mean? It wasn't a matter of
soundstaging or holographic imaging—both speakers were
champs at that—but the Thiels had a quality I can describe
only as grace. Grace is like a soap bubble: Try to dissect
it and it's gone. Perhaps a better way of putting it would
be that the Thiels got out of their own way, which is what
a high-end speaker is supposed to do."

That is how I felt about first order filter. It communicates
the heart and emotion of the music better. To me, first
order is like tube amplifier which may not be as clear
or have the impact of solid state, but it is just more
musically satisfying.

When I design my speaker, using the same cabinet and drivers,
if I design the cross over network using higher order filter,
it was more or less an academic exercise. But when I
tried to design using first order, it took me a long time
to get right but it's absolutely worth it. As a matter
of fact, it is easy to make your speakers sound "right"
using high order filter, and it's a lot easier to mess up
the sound using first order filter.

The one major disadvantage of first order, time coherent
speaker is that it really restricts your options as far
as driver configuration. That is why almost all time coherent
speakers are essentially three way. All Thiel speakers
and Vandersteen are essentially three way. Vandersteen
has a couple of 4-way but the fourth driver essentially
acts as a subwoofer crossing over at very low frequency
so it's not a problem. If you want to use multiple bass
drivers then it's probably not possible since it would be
very difficult to integrate the sound coming from different
drivers with first order filter. Also if you want to
build large speaker with multiple midrange drivers,
multiple bass drivers, I would think you have to use
higher order filter.

andy - thank you for your contributions. Indeed you are on it with "first order filter. It communicates the heart and emotion of the music better."

This communication is not some magic of technical accomplishment, and in fact it is one side of the coin of accuracy. This subtlety of communication is primarily a psycho-acoustic effect. That is not saying it is somehow fakery. On the contrary, we hear by synthesizing auditory experiences via very minimal sonic inputs which elicit associations, memories, conjectures and so forth. Then the auditory brain forwards those synthesized packets for storage and assembly into longer, more complex composites such as a musical or verbal phrase, etc. Our research at Thiel led us to commit to first-order slopes because they are the only solution to preserve the phase-time information that the ear-brain uses to believe the input is real. We really believe what we are hearing when the phase information is intact, rather than cognitively conjecturing what we hear when the phase information is scrambled, as it is with higher order filters. Much more can be said about this process, but put it on the shelf for now.

Side 2 of the coin is the technical execution. No doubt, hands down, higher order slopes are FAR more executable for exact frequency domain accuracy. In fact, first order slopes are generally considered non-executable because the drivers must have such a wide range of linear response. Higher order slopes attenuate the out-of-bandpass signal at double, triple or quadruple rates compared to first order. The ubiquitous 4th order slopes attenuate at 24dB / octave rather than our 6dB. So all the grief that the driver goes through at its frequency extremes just goes away with higher order filters, making much cleaner, more controllable frequency domain smoothness.

So, the double-whammy is that the frequency extreme grief of the first order slope is also more objectionable because the ear-brain is trying to process it as real music and not a music-like artifact as it does with higher order slopes. Both sides of the difficult coin gang up against first order slopes. Thiel decided the result of reality was worth the huge grief of execution.

As you eluded earlier, phase coherence without time coherence is not meaningful. The audio brain can only buy in if all the elements of the signal are correct. I believe, and Thiel's position is the distinct minority, that the phase-time aspect of the signal is more critical than the frequency domain aspect. In other words, it is not upsetting if a reproduced trumpet sounds slightly like a different trumpet (frequency-spectral differences), but it is upsetting if a trumpet's harmonics reach the ear at different times and in different phase relationships than real, non-reproduced music (time-phase differences).

An interesting phenomenon is that once a listener (or recording pro) has identified the importance of time-phase, there is no real going back. The artificiality of non-coherent wave-forms is unsettling, even if the frontal lobes convince us that that sound must be a trumpet. As I said before, this discussion encompasses a serious body of study; I hope this response covers the high spots. 

Regarding more than 3 drivers in a phase-coherent system, I'll comment on that later. Think CS5, 6 and 7.

Thank you Tom. I appreciate your opinions very much. I can’t wait for your response with respect to time-coherent in a speaker using multiple bass drivers.