Well, since every other speaker in the world sounds just awful, lets see if we can find why EnABL'd speakers sound so good!!!!!
Ron,
Can you use your machine to look at a square wave passing by, out at the outer 1/3 or so of the main cone? Can you repeatedly introduce just one pulse at a time, and over a period of time, look at the initial wave, and the reflections of that wave on an untreated and EnABL'd driver? I would ask for the inner 1/3 too, but I don't see how you could get in there. I want to find out how long this initial transverse wave rings.
dlr,
In the fft impulse that uses a known sound similar to white noise for it's stimulus, where is the leading edge information? What I see on a CSD plot is the maximum signal, shown at time zero. The loudspeaker is not providing that signal at that time in a true pulse or transient bit of music. The devices have a latency in both the motor and the traverse time of the transverse wave, that creates the rest of the signal, that is added to the initial piston portion of the total compression wave. Can this information be derived from the fft used in current impulse tests?
John K,
What I am looking for is how the reflected transverse wave corrupts the forming compression wave leading edge, because of it's ringing within the cone and emission of same frequency information. Have I asked Ron for a test that will show me a comparison of ringing that will show the cones activities unambiguously?
All,
I have ordered six Tang Band 3 inch FR drivers, for test purposes at RAW acoustics, PN is W3 3195 F. They are a paper cone with a shallow dome dust cap and a santoprene surround. All will be characterized with on and off axis fft from Praxis. Al also has Sound Easy and Clio available, along with a speaker rotation table and access to a gymnasium, just to get the walls out some distance. After initial characterization, two will come to me for EnABL treatment and two to Dave Dlugos, one for his coatings and one for both coatings and EnABL. Then, back to RAW for another round of tests. Then off to Texas to another interested party who will do tests on just the finished units to provide comparative data set for the RAW data.
This data will be made available in unsmoothed format for all of the various derivations. I will ask for a distortion test, with 9 orders broken out, on these units. Al has agreed to do the tests only so long as they do not directly benefit anyone's current sales operations, including his own.
Bud
Ron,
Can you use your machine to look at a square wave passing by, out at the outer 1/3 or so of the main cone? Can you repeatedly introduce just one pulse at a time, and over a period of time, look at the initial wave, and the reflections of that wave on an untreated and EnABL'd driver? I would ask for the inner 1/3 too, but I don't see how you could get in there. I want to find out how long this initial transverse wave rings.
dlr,
In the fft impulse that uses a known sound similar to white noise for it's stimulus, where is the leading edge information? What I see on a CSD plot is the maximum signal, shown at time zero. The loudspeaker is not providing that signal at that time in a true pulse or transient bit of music. The devices have a latency in both the motor and the traverse time of the transverse wave, that creates the rest of the signal, that is added to the initial piston portion of the total compression wave. Can this information be derived from the fft used in current impulse tests?
John K,
What I am looking for is how the reflected transverse wave corrupts the forming compression wave leading edge, because of it's ringing within the cone and emission of same frequency information. Have I asked Ron for a test that will show me a comparison of ringing that will show the cones activities unambiguously?
All,
I have ordered six Tang Band 3 inch FR drivers, for test purposes at RAW acoustics, PN is W3 3195 F. They are a paper cone with a shallow dome dust cap and a santoprene surround. All will be characterized with on and off axis fft from Praxis. Al also has Sound Easy and Clio available, along with a speaker rotation table and access to a gymnasium, just to get the walls out some distance. After initial characterization, two will come to me for EnABL treatment and two to Dave Dlugos, one for his coatings and one for both coatings and EnABL. Then, back to RAW for another round of tests. Then off to Texas to another interested party who will do tests on just the finished units to provide comparative data set for the RAW data.
This data will be made available in unsmoothed format for all of the various derivations. I will ask for a distortion test, with 9 orders broken out, on these units. Al has agreed to do the tests only so long as they do not directly benefit anyone's current sales operations, including his own.
Bud
Bud,
Why don't you do some testing with baffles and ports at the same time?
For baffles - two identical baffles - one EnABL'd and one not EnABL'd and swap the same driver between the two baffles.
For ports - have one test box where the port can be swapped.
I am convinced that this will be more revealing as the driver would be the same one for each test.
While we're at it, how about a compression horn test as well?
Regards,
Alex
Why don't you do some testing with baffles and ports at the same time?
For baffles - two identical baffles - one EnABL'd and one not EnABL'd and swap the same driver between the two baffles.
For ports - have one test box where the port can be swapped.
I am convinced that this will be more revealing as the driver would be the same one for each test.
While we're at it, how about a compression horn test as well?
Regards,
Alex
Re: Subjective testing
Notice how all these 'conclusive tests' are done with multi-way speakers?
I think this says more about how unrevealing multi-way speakers are than anything else.
With a decent FULL RANGE speaker and no crossover the difference may well be audible.
Of course, it may well be that the coat hanger still wins...
mightym said:
Notice how all these 'conclusive tests' are done with multi-way speakers?
I think this says more about how unrevealing multi-way speakers are than anything else.
With a decent FULL RANGE speaker and no crossover the difference may well be audible.
Of course, it may well be that the coat hanger still wins...
All I know is that I sat front row in the EnABL demo at RMAF. I heard a difference - most of the time.
In fact I was sorely disappointed, because I did not like the EnABL'd drivers much. Sometimes they seemed as good as the untreated versions, but mostly they didn't - to me.
Then I learned that what I thought was EnABL'd was untreated and vice-versa. Ooopps! Guess I like the treated drivers better after all.
I could not hear the difference on every passage, but on some it was pronounced.
That's far from a double blind test, but at least I can say that I really didn't know which was which.
Most speaker drivers are round. Perhaps that's why the arguments tend to go round and round...... I dunno.
In fact I was sorely disappointed, because I did not like the EnABL'd drivers much. Sometimes they seemed as good as the untreated versions, but mostly they didn't - to me.
Then I learned that what I thought was EnABL'd was untreated and vice-versa. Ooopps! Guess I like the treated drivers better after all.
I could not hear the difference on every passage, but on some it was pronounced.
That's far from a double blind test, but at least I can say that I really didn't know which was which.
Most speaker drivers are round. Perhaps that's why the arguments tend to go round and round...... I dunno.
Re: Re: Subjective testing
No, I didn't notice that. What I did notice is a "fullrange" electrostatic panel with a "helper" woofer @ 250hz MartinLogan SL3 . Far more "revealing" that something clouded with amplitude, IM & Harmonic distortion, like the 4-8" paper cone dynamic drivers favored here. Perhaps Bud should try his magic on Stats to improve upon the wire testbed even further?
cheers,
AJ
Alex from Oz said:
No, I didn't notice that. What I did notice is a "fullrange" electrostatic panel with a "helper" woofer @ 250hz MartinLogan SL3 . Far more "revealing" that something clouded with amplitude, IM & Harmonic distortion, like the 4-8" paper cone dynamic drivers favored here. Perhaps Bud should try his magic on Stats to improve upon the wire testbed even further?
cheers,
AJ
dbt's
This is only true because no one will ever agree on the criteria that must be used to evaluate the response results of the necessary trials amongst the participants. In no way does this invalidate the methodology, just condemns the subjects due to their lack of understanding the protocol
John L.
JacquesToo said:
This is only true because no one will ever agree on the criteria that must be used to evaluate the response results of the necessary trials amongst the participants. In no way does this invalidate the methodology, just condemns the subjects due to their lack of understanding the protocol
John L.
panomaniac said:
Sounds like a really well prepared demonstration, everybody was organized and well prepared ... designed to definitely expose both preferences and objective reality regardless of predilections of all the participamnts.
In fact, you've revealed your predilections with the comment "sorely dissapointed" in your post. No expectation bias here....
Sign me up.
John L.
Re: Re: Re: Subjective testing
Well I guess I would answer that be saying it doesn't corrupt the leading edge of the wave because the reflections occur later in time. However, the point is moot because it has already been shown unambiguously that enable doesn't prevent reflections in the cone.
You know Bud, I don't know what your back ground is but considering that you have been tinkering around with this for 30 years I think it may be about time you bit the bullet and at least made an attempt to educate yourself in this area, if you haven't already. I don't mean this as any kind of put down, but there are a lot of fundamental aspects of wave propagation, reflection, dissipation, etc. that have been studied and analyzed to death and are presented in many reference books. For example, I mentioned inhomogeneities in the cone structure. Well, to get some idea of how this would affect the cone, maybe it is appropriate to look at a loosely related problem, a vibrating string. The string has some similarity to a radial cut through the cone so we might expect some similarity in the behavior. And the inhomogeneous string has been well analyzed. Here is what is said about the effect of inhomogeneity in a string, again in Theoretical Acoustics by Morse and Ingard:
BudP said:
Well I guess I would answer that be saying it doesn't corrupt the leading edge of the wave because the reflections occur later in time. However, the point is moot because it has already been shown unambiguously that enable doesn't prevent reflections in the cone.
You know Bud, I don't know what your back ground is but considering that you have been tinkering around with this for 30 years I think it may be about time you bit the bullet and at least made an attempt to educate yourself in this area, if you haven't already. I don't mean this as any kind of put down, but there are a lot of fundamental aspects of wave propagation, reflection, dissipation, etc. that have been studied and analyzed to death and are presented in many reference books. For example, I mentioned inhomogeneities in the cone structure. Well, to get some idea of how this would affect the cone, maybe it is appropriate to look at a loosely related problem, a vibrating string. The string has some similarity to a radial cut through the cone so we might expect some similarity in the behavior. And the inhomogeneous string has been well analyzed. Here is what is said about the effect of inhomogeneity in a string, again in Theoretical Acoustics by Morse and Ingard:
This is pretty much consistent with what has been observed in the simple measurement which have been presented.
JacquesToo said:
Only for audio preferences that include visual aesthetics, quite the contrary for the acoustics. The primary purpose of double-blind testing for a speaker company is to use them as a tool to determine what people truly can detect and what they prefer sonically when they don't know system details so that the designers can then focus on the acoustics alone. The visual aesthetics come into it later, but on a marketing level. They may even have to sacrifice some sonics for desired visual aesthetics. It's been long ago proved that sonic preferences differ between sighted and unsighted testing. Double-blind is designed specifically to make that distinction. It separates visual preferences from sonic preferences.
It's just like a car. You concentrate on the performance separately from the paint color. Red paint doesn't make a car go faster, but sleek, red sports cars have that reputation. They don't go faster, but they sure look like they do, don't they john?
Dave
BudP said:
Actually, Bud, the MLS signal and any applied signal is showing the reaction that the driver has when the signal is applied while the driver is at rest. All signals, even a step, are nothing more than some combination of sine waves. Fourier analysis tells us that.
Now if you use music as the source, unless you connect the signal to the driver while the music is already playing, all natural music signals start from magnitude zero, therefore they are nothing more than a combination of sine waves, all starting at zero, but with different time deltas. The driver's response will have filter actions evident in the frequency response to include phase (its minimum-phase response curve). Of course, voice coil heating has effects and that is more complicated to analyze. Maybe you should invest in a Klippel system.
The CSD simply shows what happens when a steady-state signal (music or any other) is suddenly shut off. The full resonant nature, evident in the FR of the driver, is then displayed sequentially in time increments. Everything occurring in the driver is being shown with regard to its actual acoustic decay response within the limits of the measurement system, whatever measurement system is being used, and within the constraints imposed by the measurement conditions, such as proximity to reflective surfaces over which the tester has some control.
It's all there for you to see now.
Dave
Bud,
You may wish to see the step response that shows the reaction of the driver to a step signal. Evaluating that is not straight-forward, but before/after measurements will show how quickly (or not) the modified driver moves to the step position with whatever overshoot and ringing may have been altered. However, it's still just another presentation of the same impulse data, the same as for the FR and the CSD. It just shows it from another perspective, so-to-speak. IF you have one, you can determine the others.
Again, it's all available now.
Dave
You may wish to see the step response that shows the reaction of the driver to a step signal. Evaluating that is not straight-forward, but before/after measurements will show how quickly (or not) the modified driver moves to the step position with whatever overshoot and ringing may have been altered. However, it's still just another presentation of the same impulse data, the same as for the FR and the CSD. It just shows it from another perspective, so-to-speak. IF you have one, you can determine the others.
Again, it's all available now.
Dave
Interesting, John B. Reading down the discussion I found this:
The last point might well explain my DIY obsess...uh...fascination. I am pleased that I can build a CHEAP amp or cables and they therefore sound better to me (in some cases, anyway). But in the end it's my pleasure that matters. And I do think there are real differences in certain cases depending on equipment, recordings, room, etc etc.
And because I'm not currently able to build a cheap speaker driver, I still hope to find the time to try EnABL (as well as fullrange drivers, box designs, and maybe even dealing with my hardwood floors, plaster walls and tin ceilings!!!).
Carl
The last point might well explain my DIY obsess...uh...fascination. I am pleased that I can build a CHEAP amp or cables and they therefore sound better to me (in some cases, anyway). But in the end it's my pleasure that matters. And I do think there are real differences in certain cases depending on equipment, recordings, room, etc etc.
And because I'm not currently able to build a cheap speaker driver, I still hope to find the time to try EnABL (as well as fullrange drivers, box designs, and maybe even dealing with my hardwood floors, plaster walls and tin ceilings!!!).
Carl
hallowed dbt's
typical...
Rather than get your undies in a bind over this, why not try to understand the terminology and method, as opposed to the knee jerk response?
The whole point (as this thread demonstrates) is reaching a consensus on how to evaluate the {subjective} results in a manner everyone can come to a consensus on. That's what the word "criteria" means. Takes some time to define it in an experimentally correct way. Dave (dlr) alludes to just such testing at speaker companies to differentiate audio vs. visual responses.
John L.
JacquesToo said:
typical...
Rather than get your undies in a bind over this, why not try to understand the terminology and method, as opposed to the knee jerk response?
The whole point (as this thread demonstrates) is reaching a consensus on how to evaluate the {subjective} results in a manner everyone can come to a consensus on. That's what the word "criteria" means. Takes some time to define it in an experimentally correct way. Dave (dlr) alludes to just such testing at speaker companies to differentiate audio vs. visual responses.
John L.
The whole problem with using human senses to try and empirically quantify things, is that human beings are very inconsistent. They can be very easily rendered unreliable. Criminal pyschology is full of examples of how erratic humans behave under stress, anxiety or duress.
Example: Ask anyone confidant in a blind test to test something. He confidently tests it, and scores well and can accurately differentiate between A and B.
Now simply tell him that unless he scores as well in the next series of random A/B tests, he will lose one year's salary.
Any predictions?
Now run this same test with 100 volunteers, and chart the outcome.
I think that would be an interesting test. Test the various testing methods themselves.
I see potential for inconsistent results in any test involving human perception under a structured test due to the stress of the test itself.
Cheers
Example: Ask anyone confidant in a blind test to test something. He confidently tests it, and scores well and can accurately differentiate between A and B.
Now simply tell him that unless he scores as well in the next series of random A/B tests, he will lose one year's salary.
Any predictions?
Now run this same test with 100 volunteers, and chart the outcome.
I think that would be an interesting test. Test the various testing methods themselves.
I see potential for inconsistent results in any test involving human perception under a structured test due to the stress of the test itself.
Cheers
so you agree ;-)
I'm actually a great believer in scientific methods. Have a slight problem with idiocy.
I have always assumed the double blind epithet referred to the participants in trials, but it seems to fit the believers just as well. Blind to their own method even more than blind to their subject.
Let me help a bit: do you think the way most DBTs are performed favours the brain hemispheres equally? Is this irrelevant?
Do you think pointing out stupidities in other systems in some way proves your point of view?
Why don't you intelligently respond to Daygloworange's posts? He seems to have an enquiring mind.
BTW weasels do get aspirated by jet engines, literally and figuratively.
I know we're all busy, but I'll have to excuse myself now.
J
I'm actually a great believer in scientific methods. Have a slight problem with idiocy.
I have always assumed the double blind epithet referred to the participants in trials, but it seems to fit the believers just as well. Blind to their own method even more than blind to their subject.
Let me help a bit: do you think the way most DBTs are performed favours the brain hemispheres equally? Is this irrelevant?
Do you think pointing out stupidities in other systems in some way proves your point of view?
Why don't you intelligently respond to Daygloworange's posts? He seems to have an enquiring mind.
BTW weasels do get aspirated by jet engines, literally and figuratively.
I know we're all busy, but I'll have to excuse myself now.
J
Daygloworange said:
prediction #1) Given an entirely different set of criteria <-- (note the use of the evaluation schema) this would represent a different test, with little relevance to the results of the first go-round.
prediction #2) This would not be a test of the testing method at all... simply a different dbt with different criteria and subsequent conclusions.
Confusion here seems to be the unmitigated effort to discredit multivariable analytical methods as unreliable when, in fact, the proposed experimental design is flawed due apparently to a lack of undestanding all that is involved in setting up and doing said tests correctly.
(my last post on testing methods ... )
John L.
- Status
- Not open for further replies.