Frequency Response

[scott fitlin]

I dunno, I read, and read, and always read more, I play with different things, and differing brands, I look at graphs and specs, and claims and so on. But, in the end, its ALWAYS my ears as my final judgement and decision maker. Too many times, I have read the specs, and according to them, the drivers are this and that, get em into your system, and for whatever reason, they do something you werent aware they would, or dont do what the specs led you to think they would.

I know what I like, and I know when I hear it, and If I like what I hear, I almost dont even care what the spec sheet says!

I just got done re-installing my JBL 2441,s. I JUST had to play with my TAD 4002,s again, so I did.

Whatever JBL did when they made this series ( 2440 and 2441 ) they did right, really, really right, and even if the graph says otherwise, my ears tell me I love my 2441,s!

[Steve Schell]

This is a great discussion Widget; I just hope that it doesn't get derailed by red herrings such as whether people who like horns can hear well. I agree completely that a frequency response curve tells only a small part of the story. It seems that in audio the measurements that have become the standards are those that are easy to make, not those that might be the most telling.

When Wente first designed his experimental driver that was later reincarnated as the Western Electric 594A and JBL 375, I believe that he was trying to obtain perfect pistonic behavior of the diaphragm to the extent possible. As designs have progressed (regressed?) into the modern era, designers have encouraged and embraced suspension resonances as a means of extending high frequency output. This works on paper, but tends to result in tizzy, fatiguing sound quality. The response measurements look great, but the listening can be awful.

I believe that it is no coincidence that many critical listeners prefer drivers that use Wente's simple half roll compliance. The W.E. 594A and 713 series, the Lansing Mfg. Co. large and small format drivers and the JBL 375/2440 all used this elegant design. The half roll is not perfect, as the paper you referenced pointed out; there is a broad resonance that extends response on the top end. Seems to me though that it is generally more benign and relatively well behaved than the tangential and especially diamond pattern surround resonances, which I suspect are more peaked and narrow in frequency.

For years now I have preferred a response that is rolled off over one that is extended with such parlor tricks. Diaphragm material also makes a big difference, as we all know from listening to aluminum, titanium and beryllium diapgragms in otherwise fairly similar drivers. I have become most fond of composite diaphragms, which seem to self-damp to a greater extent than metal diaphragms and sound cleaner as a result. Most audiophiles have shunned these due to frequency response limitations, but there is still much potential in this approach. Some of the best sounding vintage tweeters, such as the EVs and the Jensen RP-302, used linen and phenolic diaphragms.

BTW, is the JBL technical paper easily accessible? Is it up on our site somewhere?

[Ian Mackenzie]

Steve,

I like what you post...your insights are impressive!

On the subject of diaghrapm damping as you say it plays a big part in tonal quality and rendering of fine details based on my recent experience.

I recently opened up my 2420 for a new diaghragm and it was quite educational to watch the pistonic behaviour under drive condition not normally associated with in use application.

Ian

Ian

[scott fitlin]

Hi Steve, always good reading, you provide!

I have been reading your post about your Cogent drivers. When will these be available?

Years ago, a freind of mine had some JBL 1in drivers that used a linen/paper type diaphragm, they were white color, and they did have a distinct sound in the midrange!

I got eyes for those beauties your making!

[Mr. Widget]

BTW, is the JBL technical paper easily accessible? Is it up on our site somewhere?

Here it is:

http://www.audioheritage.org/vbullet...ead.php?t=4410

I have wondered about the surrounds too. I have used the 2440, TAD 4001, and TAD 4003 all with the half roll surrounds. I have used the 2441s with their diamond surrounds and have listened to many of the JBL Ti diaphragms with their diamond surrounds, and I have also used the Altec Symbiotic and Meyer Sound drivers with their mylar surrounds. I feel I can make generalizations about Ti vs. AL vs. Be, but I am not comfortable making generalizations about the surrounds.

Widget

[scott fitlin]

FWIW, I have some 2440,s. Ive looked at the half roll surround, and Ive listened to them. They do have a 9K peak, gives this driver a high pitched sound up there, I always thought a bit more so than the `41. I like them, and theres yet another driver with a half roll surround I like too, the Gauss HF4000.

Many guys used to say the Gauss had the sweetest sounding upper mid, all I know is, you listen to three different drivers, and they all sound different, pick the one you like!

I also liked the Altec 288,s! and as I said, they all sound different from each other.

[Zilch]

I'm sick of hearing about the graph as the "end all". Sure, an RTA gives us insight to a driver's Fs and overall performance- but it won't tell you squat bout how it sounds!!

Well, yes, but the corollary is also true: If you don't have the requisite FR, it doesn't matter -- you don't have squat to begin with....

[Mr. Widget]

Well, yes, but the corollary is also true: If you don't have the requisite FR, it doesn't matter -- you don't have squat to begin with....

I absolutely disagree. I quite possibly may have made that statement myself a few years ago, but having listened to some very fine speakers with seemingly terrible response curves and some absolutely crap speakers that measure virtually textbook perfectly... no the corollary is not true.



Widget

[Zilch]

I quite possibly may have made that statement myself a few years ago, but....

You may be a couple of years ahead of me, then, but, for now, you'll not convince me it's all a wank.

BTW, we probably should distinguish between drivers vs. systems here.

[I'll admit I don't understand the THX "X-Curve," tho....]

[Steve Schell]

Ian, I do think that suppression of resonances plays a big part in lifelike reproduction. This applies to well behaved suspensions as well as the behavior of the major piston area of the diaphragm. An exciting aspect of all this is that compression driver design is still a young science, and I feel that many improvements in these aspects are still possible. Having said that, a major reason that we like compression drivers on horns to begin with is that these distortions tend to be lower than with competing driver technologies, due to the heavy loading and tiny excursions.

Widget, thanks for the URL. I have found also that there is sometimes little correlation between smoothness of frequency response, measured bandwidth and perceived quality of sound. One time my friend Rich and I listened to three cone speakers in mono on the same Voigt pipe: an RCA 8" full range with Olson's dual voice coil, a 6" Fostex full ranger and a 4" Jordan aluminum cone full ranger. The RCA sounded best overall, very lively with the greatest subjective high frequency extension. Next was the Fostex, not as lively but still pretty good. The Jordan suffered by comparison, sounding constrained and lifeless. When we ran LMS frequency response curves we were in for a surprise! The Jordan had low sensitivity but awesome f.r., nearly ruler flat to 18kHz. The Fostex had higher sensitivity, fairly flat response that died above 14kHz. or so. The RCA had the highest sensitivity by far, but dropped off rapidly above 8kHz.

Scott, thanks for asking about the Cogent drivers. There has been great progress recently. Rich revised our voice coil winding tool, and we have made changes in materials as well. Sensitivity of the midrange driver is up by 2dB and the sound has improved as well. Our main logjam has been in completing the patent application on our phasing plug, and it was finally mailed to the USPTO today. We are about to ship drivers to our first customer, and are in the midst of our first small production run.

Note: my comments on the benefits of composite diaphragm materials need to be taken with a grain of salt, as I am involved with a company selling same! I do feel, though, that the original RCA design of the center-suspended cone compression driver is possibly the most noteworthy and underutilized invention in the world of compression drivers. Imagine a driver with a 4.5" cone for a diaphragm, with a 2" voice coil located midway between center and edge and a cloth skirt for an outer suspension. The diaphragm is damped like crazy, between the basic construction, cloth surround and choice of materials for the diaphragm. RCA's original was silk cloth and phenolic resin, ours is carbon fiber and epoxy. The bandwidth is unreal; the diaphragm resonance is 125Hz., and use to 200Hz. is entirely possible with the right crossover. With our improvements to the RCA radial phasing plug response is fairly flat to 10kHz., and usable above that with EQ.

I will be able to say more about all of this once our patent application is acknowledged by the USPTO. We plan to launch our Cogent web site soon; I will announce it here.

Know how to become a millionaire? Start with two million and go into the speaker business!

[Mr. Widget]

You may be a couple of years ahead of me...

This isn't a competition. I see it as a shared journey.

...you'll not convince me it's all a wank.

That wasn't my point. All I am saying is that we tend to put too much value in these simple curves. They are not irrelevant, not at all. But you need to open your mind to the possibility that when objective listening is also used, you may find that the speaker with the flatter response may not be the one that we would prefer to listen to. I am not talking about room curves or tonal preferences. I am saying that there are other important aspects to sound quality too. Think about those beautiful specs coming from the 70's amps with gobs of negative feedback. Most of us would agree that an amp with less negative feedback and poorer specs SOUNDS better.

BTW, we probably should distinguish between drivers vs. systems here.

Agreed. It is much easier to glean useful info on a simple driver curve than one from a multi-way speaker system... and more difficult still to glean useful information with a full range system taken in a reverberant room.


Widget

[Ian Mackenzie]

Steve,

Another great post bringing fresh new insights into the thread.


Ian

[Robh3606]

That wasn't my point. All I am saying is that we tend to put too much value in these simple curves. They are not irrelevant, not at all. But you need to open your mind to the possibility that when objective listening is also used, you may find that the speaker with the flatter response may not be the one that we would prefer to listen to.

Hey Zilch

I have 3 systems I can set up to measure the same and they all sound different. Even set up for flatest on axis response. It's useful information but doesn't tell the whole story. Do you really think an L20T/4406 and a 4344 sound the same?? Above 60 Hz you can get them surprisingly close and the L20 has the smoothest response. The 4344 can eat them alive. So Widget I agree it's useful but not the whole story.

Rob

[Ian Mackenzie]

Giskard,

Time step function on the FFT might be an indicator of damping.

My tests of the horn out of the box the other day were remarkably clear of aberrations if you know what I mean.

Ian

[scott fitlin]

So what shall we do? Brush up on adjectives and adverbs instead of relying so much on visual aids I suppose.