Uncommon taste in speaker voicing? Need advice

[poldus]

Since I lack any technical preparation whatsoever I would like some comments from the experts on what I think is an uncommon preference of mine when it comes to voicing speakers. I listen to classical music for many hours every day. I find that the only way for me to achieve an acceptable sound that won´t cause listening fatigue is to create a dip in the mid-high frequencies.
After a couple of years tweeking my diy speakers I ended up with a first order filter: a 1.5 mh coil to feed the vifa pl17 mid-woofer and a 2 uf capacitor on the tweeter (ribbon foster).
I´ve just got myself a pair of quad esl-57 and I am basically doing the same thing to them: 1 mh coil before the quads and then fill in the high frequencies with the foster and the 2 uf coil.
My source is a denon 2900 for cd´s and sacd´s.
Is the frequency dip as pronounced as I think? Is this normal?

[Cal Weldon]

Everyones ears are different and that's what yours tell you to do. Nothing wrong with that. In fact the louder I expect a speaker to be played, the bigger upper mid dip I introduce to it. Small indoor stuff doesn't need a dip as I rarely play it loud enough to experience fatigue but the outdoor gear is different. I have a heavy mid dip so they may be played loud and long without faitgue. They sound very dry and lifeless at low volumes though. In the end it's all in what sounds good to you, not what the graph looks like.

[poldus]

Thanks for your reassuring comment Carl. I sure was beginning to think I was ´hearing things¨

[Drew Eckhardt]

Originally posted by poldus
Since I lack any technical preparation whatsoever I would like some comments from the experts on what I think is an uncommon preference of mine when it comes to voicing speakers. I listen to classical music for many hours every day. I find that the only way for me to achieve an acceptable sound that won�t cause listening fatigue is to create a dip in the mid-high frequencies.

Is the frequency dip as pronounced as I think? Is this normal?

Human head-related-transfer functions mean that we're less sensitive to reverberant fields arround 3KHz than sounds from in front of us. Recording microphones don't make the distinction so live recordings are often overly hot in this area.

[Foxx510]

I've often wondered how each persons hearing response curve differs, I find that I just can't handle any peaks in the 1-3k area, and generally need a small dip in this area. A "flat" speaker to me generally sounds too pronounced in the midrange. It would be great if we all came with spec sheets for our ears!

[Jay_WJ]

According to equal-loudness contour --- http://en.wikipedia.org/wiki/Equal-loudness_contour ---, human ears are sensitive to sound in 2 kHz - 5 kHz. So, a dip in this range seems to be a key to building non-fatiguing loudspeakers.

For instance, this kind of dip is present in all of recent XO designs by Curt Campbell.

[Dave Bullet]

Two other points that might be relevant

1. On-axis gated measurements (with a dip) can become off-axis listening without a dip

I've found that a designed on-axis 3KHz - 5KHz dip can actually become flat off-axis.

2. Driver distortion increasing perceived brightness.

Zaph mentions this in his mantras. Some drivers with higher distortion in their passband might even "brighten" the presentation. Zaph mentions low distortion drivers "voiced flat" sound good to him... could be when lesser drivers are used, a relaxed presentaiton is required due to increased distortion.

So far, I prefer a more relaxed voicing in the upper mid / lower treble too.... until I try some better quiality drivers to see if it is distortion related meaning flat might be fine?

David.