Frequency Response 23khz and above??

[John Warren]

Here's the WF for the EV T35. The surface contour is "bullet nosed" within the first ms of the impulse.

[John Warren]

Same input power, same conditions. Here's the JBL 2404H. The first millisecond contour is more like a "1/4 cylinder" which demonstrates that the tweeter is capable of dissipating sound energy over a broader bandwidth with uniform decay (say above 4kHz). It has a higher upper limit is a consequence of being a better tweeter. Note too the nasty hang-over in the T35 (above) at the resonance (~3.5kHz). This illustrates why the tweeter should be crossed over at about an octave above Fs.

[ZAKO]

Thank you John,,,I have always enjoyed your expertees....You,v been away too long.

[DrWho]

A stand-alone tweeter is a minimum phase system

That isn't necessarily true.

Filters cause EXCESS phase and complicate this.

Most any electrical filter is minimum phase. There are ways to make non-minimum phase filters, but the kinds we see in audio applications tend not to be.

[DrWho]

The first millisecond contour is more like a "1/4 cylinder"...

That is no different from saying that the 2404 has a wider/flatter frequency response. We certainly don't need a waterfall to tell us that.

While the 2404 is arguably a better tweeter, I don't think there is anything in these waterfalls that shows one as being absolutely superior when it comes to ringing.

Just for kicks, try a lowpass on your 2404 to match the frequency contour of the T35. Apart from the difference in Fs, you're going to see very similar spectral decay.

[PrestonTom]

This issue of bandwidth and transient response is actually fairly well-known. Let's not get caught up in measurement systems and instead ask some very simple questions. What is the ring time of a filter (and yes a transducer is a "filter" also) as you change the bandwidth? The narrower the filter (higher Q) the longer the rise and decay time.

In very, very simplistic terms: strike a cheap wine glass with a fork and you will get an abrupt and short-lived "clunk" (a broad band response), Now, strike a crystal wine glass with a fork. You will get a tonal response (narrow band "ring") that decays very slowly. Hopefully, this analogy will provide some intuition for those not familiar with systems analysis.

Now the obvious question is whether the "filters" in our own auditory system will still allow you to "appreciate" the temporal aspects of the two transducers that John Warren demonstrated?

-Tom

[Jeff Matthews]

Its the .5 that i question.

Your killin' me, Maron! [Y][][][]

[artto]

Anyone who says they can hear a 28.5Khz tone is either a liar, or doesn't have a clue what they are talking about, and needs to be put to the test ~ by yours truely.

That being said, while none of us can actually hear that high of a frequency tone, we can often hear the effects of hose frequencies, if they are present on the recordings, of which they alomost always are not. However, lets assume for a moment that these higher frequencies are present or capable of being reproduced from a certain recording. Even though we can't hear these tones (or most likely noise), these upper inaudible frequencies can cause frequency modulation in the audible range.

As an example, a friend of mine with some of the most expensive high-end equipment around had some Tannoy super-tweeters on his system. He claimed that since the tweeters went up to 40Khz, and because he could hear the difference with them switched on or off, that he could hear all the way to 40Khz, and therefore the super-tweeters were required. Interesting, because I had recently had my hearing tested (I'm 56) and it rolls off pretty rapidly after 12.5Khz, yet I too could hear the difference. The difference was caused by the super-tweeter filter (crossover), frequency intermodulation, probably also by some phase interference between the drivers. And quite frankly, I thought it sounded better, more real, without the 40Khz super-tweeters. As a point of reference, a week later he visited my place and couldn't believe "how realistic" it sounded. He went back home, emailed me later and said how "small and distorted" his system sounded.

[IndyKlipschFan]

artto..

Nothing beats a perfectly matched system too. hehehehehehe I have heard yours in incredibe.. For YEARS now!

When I go into a high end jewelry shop, I can "feel" the glass breakers on. Sometimes it is uncomfortable, but I do indeed "feel" something. A dog whistle the same thing. Those negative ion energizers too.

Same thing is true when I have my own hearing checked... I have always been over 20k in both ears but I can't tell you I hear a specific point in a note and then I can't above that next note? But I can tell a test signal that is very high is on or off in my left and right ears. BTW, most hearing tests if your lucky go to 12k as well.

I have been blessed with an incredible range in hearing, Which is not always a great thing. (I sometimes, even if I am not trying to hear things in the next room or two over, I have to deal with not trying to... but overhearing other tables close by. It is not louder.... Just always been able to do so. I have to concentrate to only hear what I want sometimes... It made school in larger lecture halls tough to concentrate. e.g. movie theaters too!)

Hearing different parts in an orchestra or choir has always been easy for me to follow. Complex Jazz or other complex music is always a treat too.

I have no idea if it might be a sad or happy day when it drops to that 10- 12k zone.. ?? I have been told it always does in older men.

I agree with the analogy of your friend.. You feel something up there but is it a good thing? Or is it added harmonics and or hiss too? I woudl take a lot of Klipsch speakers clen to 17k any day than a hiss to 20k30k40k too!

[John Warren]

A stand-alone tweeter is a minimum phase system

That isn't necessarily true.

m

minimum phase. There are ways to make non-minimum phase filters, but the kinds we see in audio applications tend not to be.

m

Edited April 22, 2014 by John Warren

[John Warren]

This issue of bandwidth and transient response is actually fairly well-known.

Correct Tom but I want it to be painfully clear

Wide bandwidth means better transient behavior. A tweeter with say, 2 1/2 octaves of bandwidth (4-24kHz) has better transient behavior that one with 2 (4-16kHz).

[Klipschguy]

I can only hear up to 28.5KHz @ 0dB down and I find there is very little musical information up there on most recordings.

OK I'm impressed

Me Too!

Klipschguy,

How old are you? With all due respect, and forgive me for asking this, ........28.5KHZ? Are you in the Guinness Book?

Woof

tc

I'll be 94 in the fall, but after working in the boiler room for all those years, 28.5KHz is about all...

Of course I am kidding about my age and the high freq hearing. The high frquency claims of some audiophiles (not anyone here of course) has aways given me a chuckle, so just a little lighthearted jest. Along the same lines it is funny on some of the musician sites I visit where everyone has perfect pitch. Those guys are fun to rib a little.

Andy

[BadDad]

I asked my pet parrot about hearing 23khz and she nodded her head and kept squawking.

[DrWho]

Oh it isn't? Are you trying to tell us that I can't take the frequency response from a standalone driver and perform a Hilbert transform on it and get the phase response? Are you going to tell us that that's non-sense? Specifically, what does *isn't necessarily true* mean? What circumstance are you trying to highlight here? I'd like to know. When is it *sort of* the case that a driver is minimum phase and when does it exhibit excess phase, you tell me.

reflections

[DrWho]

Wide bandwidth means better transient behavior.

So 20Hz to 640Hz (5 octaves) has "better" transient behavior?

[John Warren]

reflections

Of what, times gone by?

Isn't that why we've got anechoic chambers? OR time-gating MLS systems like MLSSA, CLIO and Matlab(?),

How about just doing outside?

[John Warren]

Wide bandwidth means better transient behavior.

So 20Hz to 640Hz (5 octaves) has "better" transient behavior?

Compared to a driver that has 4 octaves of linear pistion output (20-320Hz), yes.

[jacksonbart]

I am blowing my dog whistle as I type this, can anyone hear it?

[ZAKO]

The only reflections Mike can hear are bouncing around in his little RED suitcase. He has always been a suitcase full of missinformation.

[Don Richard]

Wide bandwidth means better transient behavior. A tweeter with say, 2 1/2 octaves of bandwidth (4-24kHz) has better transient behavior that one with 2 (4-16kHz).

Technically correct, practically irrelevant. Harmonics of the fundamental are what gives rise to steeper rise times (pun intended). A measurement mic may have the necessary frequency response to detect the frequencies which would steepen the rise time of a complex waveform displayed on a graph. An FFT done on that waveform would separate and display the amplitude of those frequencies comprising that waveform at that instant in time. If these frequencies (harmonics) that cause the rise time to steepen are beyond the range of a human's hearing, no information will be presented to the brain which would cause any detection of improved transient response.

If one is interested in drawing curves on a sheet of paper, the transient response is improved by measuring the tweeter with the wider bandwidth. Any audible difference, if any, between that tweeter and the other would be caused by other factors. Considering the information given above, those factors have not yet been determined accurately or completely, in the above case. Considering that different people have different hearing limits, blanket statements on the audibility of transient behavior are completely incorrect.

What is correct is the analogy of human hearing to an FFT display. The psychoacoustic mechanism of human hearing hearing behaves like an organic FFT because the cilia that move in response to the vibrations of the eardrum are frequency-specific. The cilia that are moving stimulate the auditory nerves, and the brain is then presented the varying amplitudes of the frequencies comprising the waveform at that instant in time. The brain then decodes that information which we percieve as sound. If no cilia are stimulated there will be no information presented to the brain pertaining to those frequencies so we cannot percieve "better transient response" in that case.