The Reproduction of the Upper Octaves of Music

[MylesBAstor]

I was listening to the system tonight with the Kaguras--whom every time I listen to them--continue to impress me and redefine the term "naturalness." A naturalness that unlike many other SE ended tubes amplifiers that have passed through my system, doesn't result from a rolling off of the upper octaves and a midrange emphasis. The proverbial one thing led to another and that got me to thinking about upper octave reproduction. Yes, the highs are there with the Kaguras but they don't call attention to themselves. But the upper octaves are certainly there when they need to be. That got me to thinking about two recent quotes from Alon Wolf that appeared in my Magico S7 preview.

Alon strongly believes in tweeter performance and the driver's effect upon the speaker's overall sound. He observed that, "some audiophiles want speakers to excite them." In Alon's estimation, "that's the last thing they [speakers] should be doing. The excitement should come from the disc itself."

"We are used to," Alon mused, "what tweeter artifacts have to do to produce high frequencies."

Yes it was of those a-ha moments that had been staring me in the face for the longest time but hadn't put 1+1 together. But it ties in with the loss of the midrange and trying to compensate with accentuating the highs. No, it's not losing the mids because we are trying to get a wider frequency; it's just just being aware that the mids need to be improved. But more than that, as Alon pointed out, what we call detail is just anything but that. It's all artificial and leads to that mechanical quality in our reproduced music. That quality that always keeps us at arm's length from the music.

It also interesting to hypothesize what our systems might sound like without this artificial detail. Would we need cables to tune [not in the way of say MIT or Transparent here] the sound of our systems? Those with analog based front-ends might also find that they are still setting their VTA/SRA to compensate, no matter what the scope shows, to compensate for these artifacts. Would even digital front-ends benefit from that lack of artificial edge?

[Rust]

Myles - Good post. Good point. The growing trend in exaggerated highs, an extra dB or two under the power curve is exactly what I find annoying in long term listening. Aside from that, boosted highs cause the midrange to sound thinner and more recessed, even though in isolation or with a more balanced system the mids likely might be decent. Or not.

A good speaker should not add to nor subtract from.

Dynamic tweeters have to be a bear to design, the dome material issues (stiffness and weight requirements/issues are well known) but dealing with the way inductive coupling and thus impedance in the motor system changes as frequency rises has to be tough. The only two options are a better tweeter design or forcing the tweeter into a performance envelope by way of compensating networks.

Forcing drivers into a performance envelope seems to kill the sound. Many years back a designer was showing off a massive crossover with compensating circuits which resulted in a very flat frequency response. Don't get me wrong, I always thought he was a really good guy, but the resulting speaker to me was just dull in the midrange and the highs were a little weird.

[Guest]

I don't want to hear one compensating for the other. I don't want to hear drivers at all.
Knowing how to match drivers and driver frequency ranges breeds coherence.

[MylesBAstor]

I don't want to hear one compensating for the other. I don't want to hear drivers at all.
Knowing how to match drivers and driver frequency ranges breeds coherence.

I'm not sure how coherence and compensation are mutually exclusive. I think you can have a coherent speaker eg. seamless transition between drivers yet the tweeter's artifacts color the sound. That said, a tweeter should be seen and not heard. Almost as if it's not there except when it should work.

[Guest]

I'm not sure how coherence and compensation are mutually exclusive. I think you can have a coherent speaker eg. seamless transition between drivers yet the tweeter's artifacts color the sound. That said, a tweeter should be seen and not heard. Almost as if it's not there except when it should work.

Sure - I don't disagree, but doesn't it turn on the particular implementation? If the drivers are matched well that doesn't mean there can't be issues beyond the crossover area. But I was thinking of scenarios where the midrange is stretched further into higher frequencies than maybe it should be or vice versa with the tweeter. I ran into the former with an early model Thiel. Despite the time-alignment, the lower highs / upper mids always seemed pushed forward, while the higher highs (?) were not. With the tweeter disconnected I could hear the mid straining at its upper end. If the tweeter could have gone a bit lower, the mid range would not have been pushed to compensate for its inability not to. Or so I thought! That was with a Hafler 500 - a while back. :-)