|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
69.59.200.72
In Reply to: RE: I look at it slightly differently... posted by John Marks on August 17, 2015 at 12:56:28
It was interesting to compare the measurements of the Wilson Benesch speakers and the Falcon LS3/5A in the previous issue. Below 1.2kHz or so, the on-axis response is nearly identical: significant humps around 100Hz (like probably the majority of small 2-ways) and 1kHz, with a deep saddle between them. While they are voiced quite differently above that range, both speakers' response peak at 1kHz (an artifact of the woofers used, since it is both well above box loading and below crossover frequency?) HAS to contribute significantly to the sonic signature or "personality" of the speakers.
As for the treble responses, I see different things going on. The Falcons place both drivers down -6dB at crossover and in reverse polarity, a quasi-LR2 alignment which has become fairly standard practice in recent years. They sum reasonably flat through the crossover region, with a slightly lifted HF response (an obvious design choice) and a lot of hash in the upper treble, probably attributable to diffraction nasties off the protruding cabinet edges. The WB appears to be using Butterworth transfer functions, with both drivers in positive polarity, down -3dB and 90 degrees out of phase at crossover. Measured on the tweeter axis, this will quite predictably result in a peak at crossover frequency and a phase cancellation dip just below. On the "correct" axis, somewhat below the tweeter, this should level out a bit as the phase falls into place.
As a speaker DIYer who works long and hard to achieve the flattest, most neutral response with the best possible phase tracking an octave or more either side of crossover, I find it intriguing that these rather gross response anomalies should sound pleasing to two highly experienced listeners. I've got an old graphic EQ in a closet somewhere. I may have to dig it out, put it in the system, and dial up a big spike at 1kHz just to hear what that sounds like. Who knows, I might like it!
Follow Ups:
Just a thought:
You need to re-create all conditions of that peak. It is pretty sharp. I believe a GEQ will have a much lower q cenetered around each frequency. Remember: A feedback destroyer works by finding reaonant frequencies, and filtering them out with ultra-sharp filters. Concidered to be inaudible, but a microphone set to 1/48th or higher woukd likely pick it up. Cool, but most times too much information is no good either because we then take it and fool ourselvs things are there, or not there, that are or arent....at a single point in space, one speaker driven, and so on. The complex sum of what we hear with both channels driven, room interaction, and HRTF's does not always mean a single responce measurement is quantifiable as "good information". Its a snapshot, like the Alamo, from one direction you see majestic history, from another you see 'meriga. More important that anything, and those very few I trust can weigh everything involved, and the summation of ALL data points concidered with good, well educated coorelations formed yeilds a very good picture. I am unsure any reviewer or publisher gets it right all the time however I am always inclined, when the data seems marginal, to aire tword the human expirence.
Edits: 08/18/15
Plus it's op-amp based and probably degrades the signal, adds a tiny amount of hiss, the bands aren't narrow enough, and everything you said.
1,000 Hz is a funny neighborhood...
If the recording you are listening to is of a baritone singing Schumann's "Dichterliebe," there's one song with an unaccompanied prologue, "Ich hab im' Traum geweinet."
The first five notes are B-flats at circa 474 Hz, and being low male voice, the overtones are weak. If you were to EQ that phrase with a narrow 6dB peak circa 1,000 Hz, I rather doubt it would sound shockingly different at the listening position.
Whereas 1,000 Hz is near the fundamental of the first notes of the violin solo part of the Mendelssohn Violin Concerto (B = 987 Hz), and a peak up there may at the listening position just seem that the soloist's mic (or mics) was goosed up in the mix somewhat.
(I say "up there" because even though 1kHz is below almost all applicable crossover frequencies, I do not consider it "midrange" but rather, looking at a piano keyboard, it is treble. Are the any music listeners who think that he opening of the solo part of the Mendelssohn Violin Concerto does not take place in the treble region?)
So, I will continue to state that in its price range, I think that the Wilson Benesch Square One is a very respectable contender with much to recommend it--at least from the listening chair.
But most people already know that, as previously mentioned, I have loved the sound of Shahianian's larger speakers; but in the interest of alerting the unwary, I did go so far as to call their sound "wide-screen" and "Technicolor."
At the end of the day I think this is another case of Horses for Courses (please follow the link below, as it is one of my best all-time efforts), and that the most important thing is that no French people were harmed in the making of my column.
ATB,
JM
I like your term "intriguing," rather than the term JA used, "surprising." Why would JA be "surprised" that a reviewer enjoyed a speaker that didn't measure the way JA thinks it should? As mentioned multiple times in this thread, it happens all the time, and there's nothing wrong with that.
I'd also like to comment on some of the IN-ROOM speaker frequency response measurements that JA has published in the past. I think these are fascinating in that, in just about every case, when two even vastly different speakers are compared in-room, the graphs seem to track each other almost perfectly. I would say we are seeing a frequency response graph of the room, just as much as the speaker.
I've also noticed that in JA's room, there is a frequency bump at 50-60 hz and a drop-off at 40 hz with just about every speaker he tests there, then a rise again at 30 hz. Based on the Stereophile Test CD test tone track, this is the same type of response that occurs in my room. Perhaps this is more common than not?
Edits: 08/18/15
And the second most important is whether the amp is well-suited to driving the speaker in terms of its electrical load and back EMF.
I don't think that there is any question that having flat measured frequency response is one desirable characteristic for loudspeaker performance. But it is not the only one. The character of the dispersion is important, and the presence or absence of distortion is important.
I think that it is a respectable position (I did not say it is the majority position) that phase coherence in the crossover region is so important that if in order to maximize that in a particular design at a given price point, flatness of frequency response might have to give way.
I have heard at least one loudspeaker that claimed to measure flat anechoically that could be problematic in real-world rooms, with the result that the speaker was tri-amped with active crossovers... which I am sure were meant to be used.
I am not a loudspeaker designer. However, I think that I have a fair amount of experience as a listener, and some of my recordings have found favor. I wish that I could say that the sum total of all the wisdom I have gained since 1966 or so could be expressed as: "Buy a loudspeaker with the flattest quasi-anechoic measurements you can find, and live with it and be happy regardless of what your room is like," but that is not the case.
Which is why I am a big fan of room treatments, which I last dealt with at the attached link.
ATB,
JM
Those diffusors make me dizzy ............ yeow
I really like those concentric circle diffusors. Maybe for the next project, I'll be spending a LOT of quality time with my router!
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: