planet10 said:
Looking at these, I would say there is much difference, but IMO, not improvement. It might be possible that one will get the impression the driver sounds louder, however, when you try to distinguish the instruments, the timre will sound a bit mushy. Loud listening levels might sound a bit fatiguing. This is just the guess from looking at the data. I wonder what the actual listening impression is.planet10 said:
BudP said:
Hi Bud,
John said what I had in mind and quite a bit more, as well as better than I might have, so I won't try to add to it.
However, I do have a few questions. On seeing the first FR curves above, I noticed the extension down to 20Hz. This surprised me since I thought that most true anechoic chambers don't go this low. As a single measurement (it doesn't appear to be a close-mic spliced to a quasi-anechoic measurement) I don't see how this could be obtained with a window set to stop at the first reflection.
Further study shows that you used a Blackman window, a full Blackman window, indicated by LAUD at the bottom as . Can you describe the measurement conditions (driver, room, etc.) and why you used the full Blackman? Is this a single FFT or a spliced response re-imported? The version of LAUD that I have shows the width in time of the window that was used when the FFT was run, but it doesn't show. I'm having a hard time squaring some of what I see with what I know of LAUD.
Is this a closed box measurement or an infinite baffle type?
Dave
Yep, probably inverted with some loss of accuracy at very low frequency where phase levels out to -45 degrees instead of approaching zero.
FrankWW said:
FrankWW,
This is an excellent post!
Perhaps because I am neither a scientist nor engineer and have no experience with rigorous empirical measurements and/or testing procedures, I have not always understood the confrontational posture taken by some in this discussion.
It seems to me, as FrankWW suggested above, the best investigation of many hypotheses would include both subjective and empirical data. When I first began reading about Bud's EnABL process I thought ok, I haven't heard a pair of treated drivers so I can't really say anything is happening but, on the other hand, I haven't seen any measurements that tell me something can't be happening either. I probably didn't say that right but the point is I continued reading with an open mind.
The only tool some of us have to make sense of the world around us is our subjective experience. To completely discount someone's subjective impression of an event and call it meaningless because - that person - can't quantify or measure what may or may not be happening defies common sense.
Don't many discoveries begin by someone observing, feeling, tasting, hearing or smelling a change or difference in something then deciding to bring out the 'tools' to try and understand why the subjective changes they experienced actually are or are not happening?
But then again I have a liberal arts education. Subjective, warm and fuzzy description and evaluation are my stock in trade. I don't even know which end of an O scope to look at. Still, just because - I - can't measure it doesn't mean it isn't happening.
Listen with an open mind!
YMMV -
Bruce
abpea said:
No one is discounting subjective impression, I and others have said all along that there may easily be a change in the driver output, as indeed there appears to be in the examples shown, and that they may be audible. The discussion has been about quantifying the change and the mechanism effecting the change. This can only be done by measurements.
Absolutely, I agree 100% with this sentiment! I said pretty much the same thing in an earlier post (1892) extracted below, with the caveat that the best method is controlled testing. No one seemed to notice this. It may be that they don't agree.
"The next step after measuring and controlled testing would be to try to determine a correlation between the measured changes and the perception found in the controlled testing. "
However, if an individual hears a difference, "bringing out the tools" (testing) is how one may determine the reason for the difference in sound. That's exactly what I do when I'm working on a design, the only difference is that I do this when working with a system designing a crossover. The methodology is the same, save for having a good idea ahead of time as to what changes to make in the crossover to achieve the results desired. I always verify afterwards with measurements.
The issue is centered around the tools to use, what their capabilities are and how to use them.
Dave
john k... said:
As been has said many times before... this data is confounding in that it does not jive with the magnitude & the consistently positive changes objectively observed (blind or sighted)
Indeed this conventional data would indicate that it makes things worse. Hence the conjecture that these kinds of measurements aren't showing us what is happening.
dave
dlr said:
Hey Dave,
I appreciate your thoughtful reply. You and I are on the same page.
I think a big part of the frustration in this thread comes from people who think they hear a change for the better from Bud's treatment, who want to provide the empirical proof being asked for by some, but either don't have the training, proper equipment or the ability to otherwise produce any meaningful measurements.
I don't think there is anyone following this thread who would not be delighted in having a proper set of measurements to explain what is or is not happening. It's just a matter of having the where-with-all to get it done properly.
Take care -
Bruce
planet10 said:
I don't find it confounding at all. It just doen't show the result you want it to show. Similar measurements of the types of drivers people are reporting differences in when listening might shed some light on this. It does "jibe" with the idea that added mass leads to difference response. What it doesn't jibe with is the esoteric theories of what the treatment does. If the idea is that the frequency response doesn't change then there is a problem with that idea. If the idea is that something else changes which is not revealed in the FR then before any listening test can be considered even remotely a reflection of these other changes the FR had better be equalized to match that of the untreated driver. After all, if you are looking for the effect of an unknown, then all know quantities must first be accounted for and eliminated as a cause. It's like observing a red car and a blue car in a drag race and everyone who sees it agrees that the red car is faster. Since they have been told painting the car red makes it faster the conclusion is red paint makes cars go faster. Or, in a blind case they watch in black and white and low and behold everybody still agrees that the same card is faster. And low and behold it turns out to be the red car. And then the observers say, "Dam, I guess red paint really does make the car go faster. "Well, before that can be substantiated it might be advisable to look under the hood, or check the read end ratio, or other things which are known to effect the acceleration of a car. And then there is the psychoacoustic effect. Ferraris are fast and they are painted red. There fore we are biased before hand to believing red paint make cars fast. (I always like to get a Ferrari reference in there. 🙂)
What these and other posted measurements have done is to establish that of all the possibilities of what Enable treatment might do we can at least be sure that there is something it does do: alter the FR. That is, FR alteration is not a speculative result; it is an established, real result.
John L.
I agree, the data shows a speaker that should change from reasonably accurate to pretty tough to listen to. Subjectively that wasn't the case or I would not have gone on to use the drivers so treated in a multi way system.
Those odd order distortion peaks and the coincident frequency response notch, with resultant ringing, do however, interest me. Now that I have seen them in two seperate plots of different drivers I have a surmise about their nature.
Let's assume for a moment that the initial FR depression at those points of odd order distortion masks the correlated summing of that distortion, in our subjective understanding of the speakers sound. They obviously are not gone, but, their onset is later in time and it's energy is already dissipating, when the distortion shows up on the CSD.
This same phenomena showed up in Soonsgc's final plot sequence. Where he located the inner ring on the Jordan driver to a position that dispersed this same sort of FR dip and ensuing strong node of resonance. I am going to postulate that there was another aggravated node of distortion, in odd sets, that was uncovered by applying the outer pattern ring, for the previous test sets. We do not have that data to look through, but that is OK. I suspect we will find this nodal peak in all EnABL'd drivers. It is quite likely it is also there in the original untreated drivers, as they also show a FR node at this point, but one swamped by some other factors.
Soongsc's precisely located pattern dispersed a similar nodal ringing completely. There was still ringing, but it was not collected up into a narrow band of frequencies.
I took the time to look for this node in a pair of FR8 drivers I have here for experimental treatment. Not having a test setup to use, I decided to try another method, listening for an obvious and repeatable change in sound "quality" as various materials were drug across the main cone surface, on the front and on the back side. Listening both through the cone and directly. I did find a particular diameter, that showed up over the bulk of these operations, that had this artifact of changed sonic quality.
So, I put a ring of blocks on the front side, at that point. The results were horrifying.
If you look at that plot Soonsgc provided, there is a sea of sharp points dispersed, rather than a ridge of energy at one location in the frequency band. This was exactly what the sonic quality sounded like after applying this ring. Harsh, guttural, just exceptionally nasty. I did post my results and cautioned readers not to try this at home. until I had a better idea of what was going on. It seems I may have completely uncovered what the more basic version of EnABL had crammed into a fairly narrow band width and covered over. This sonic quality was definitely not audible before application of this ring.
Having nothing smarter to do, I put a gloss coat over the portion of the cone up to and including this interim ring of blocks. The harsh guttural edge to sounds vanished to audibility, unless excited by some harsh, very sharp, edgy modern indy rock music. I continued applying the gloss material with no particular alteration in this sort of response.
It had not escaped my attention that this pattern ended up in a position that iis vertical to the cone surface, under the flange on the whizzer cone. So I deliberately mass damped the back side of the whizzer with an ever sticky acrylic paper glue and over and under bent the flange to decouple it from the rest of the whizzer. The driver no longer has the harsh response to the modern music genre. And, the midrange no longer has a "thick" and unnaturally undynamic character.
I do think these collected groupings of odd order distortions are caused by the patterns. They aren't particularly noticeable in a listening situation. Even though they are in an extremely sensitive area of the frequency band.
I will be somewhat more intrusive in getting the laptop/docking station up and running again, because I want to look at this odd order phenomena. I surmise I will find it in all EnABL'd drivers and that it will be audibly masked by the sharp FR notch and delayed onset of a diminishing resonance. It reminds me very much of something I saw here http://youtube.com/user/shermph in the excited dust test, that shows rotational energy emission events.
The part that interested me was the comment that the rising, expanding and returning "bubble" that these events arise from never ceases it's energy exchange, regardless of frequency of excitation. I strongly suspect this activity is implicated in the resonant node of FR tied to these odd order narrow bandwidth distortions.
dlr.
These tests were all conducted with a closed box. A 28 inch long closed box with an 8 by 14 inch driver mounting face. A movable internal plate, with a gap equal to the area of the driver was also included. This plate was moved for the flattest FR across the pass band up to roll off. No damping material was placed in the box.
Tests were performed in a well damped living room with well marked locations for test box, microphone and large human body. I am certain all of this insults your personal objectives for repeatable transferable testing, but it is what I had to work with. Had I found results that I though were of more than incidental interest I would have pursued tests in a better facility.
Now that John and you have indicated the importance of what I thought, in my ignorance, were changes of less note than a change in humidity might cause, I will likely resurrect testing. Especially to look at this interesting odd order distortion node.
My thanks to both of you.
Bud
I agree, the data shows a speaker that should change from reasonably accurate to pretty tough to listen to. Subjectively that wasn't the case or I would not have gone on to use the drivers so treated in a multi way system.
Those odd order distortion peaks and the coincident frequency response notch, with resultant ringing, do however, interest me. Now that I have seen them in two seperate plots of different drivers I have a surmise about their nature.
Let's assume for a moment that the initial FR depression at those points of odd order distortion masks the correlated summing of that distortion, in our subjective understanding of the speakers sound. They obviously are not gone, but, their onset is later in time and it's energy is already dissipating, when the distortion shows up on the CSD.
This same phenomena showed up in Soonsgc's final plot sequence. Where he located the inner ring on the Jordan driver to a position that dispersed this same sort of FR dip and ensuing strong node of resonance. I am going to postulate that there was another aggravated node of distortion, in odd sets, that was uncovered by applying the outer pattern ring, for the previous test sets. We do not have that data to look through, but that is OK. I suspect we will find this nodal peak in all EnABL'd drivers. It is quite likely it is also there in the original untreated drivers, as they also show a FR node at this point, but one swamped by some other factors.
Soongsc's precisely located pattern dispersed a similar nodal ringing completely. There was still ringing, but it was not collected up into a narrow band of frequencies.
I took the time to look for this node in a pair of FR8 drivers I have here for experimental treatment. Not having a test setup to use, I decided to try another method, listening for an obvious and repeatable change in sound "quality" as various materials were drug across the main cone surface, on the front and on the back side. Listening both through the cone and directly. I did find a particular diameter, that showed up over the bulk of these operations, that had this artifact of changed sonic quality.
So, I put a ring of blocks on the front side, at that point. The results were horrifying.
If you look at that plot Soonsgc provided, there is a sea of sharp points dispersed, rather than a ridge of energy at one location in the frequency band. This was exactly what the sonic quality sounded like after applying this ring. Harsh, guttural, just exceptionally nasty. I did post my results and cautioned readers not to try this at home. until I had a better idea of what was going on. It seems I may have completely uncovered what the more basic version of EnABL had crammed into a fairly narrow band width and covered over. This sonic quality was definitely not audible before application of this ring.
Having nothing smarter to do, I put a gloss coat over the portion of the cone up to and including this interim ring of blocks. The harsh guttural edge to sounds vanished to audibility, unless excited by some harsh, very sharp, edgy modern indy rock music. I continued applying the gloss material with no particular alteration in this sort of response.
It had not escaped my attention that this pattern ended up in a position that iis vertical to the cone surface, under the flange on the whizzer cone. So I deliberately mass damped the back side of the whizzer with an ever sticky acrylic paper glue and over and under bent the flange to decouple it from the rest of the whizzer. The driver no longer has the harsh response to the modern music genre. And, the midrange no longer has a "thick" and unnaturally undynamic character.
I do think these collected groupings of odd order distortions are caused by the patterns. They aren't particularly noticeable in a listening situation. Even though they are in an extremely sensitive area of the frequency band.
I will be somewhat more intrusive in getting the laptop/docking station up and running again, because I want to look at this odd order phenomena. I surmise I will find it in all EnABL'd drivers and that it will be audibly masked by the sharp FR notch and delayed onset of a diminishing resonance. It reminds me very much of something I saw here http://youtube.com/user/shermph in the excited dust test, that shows rotational energy emission events.
The part that interested me was the comment that the rising, expanding and returning "bubble" that these events arise from never ceases it's energy exchange, regardless of frequency of excitation. I strongly suspect this activity is implicated in the resonant node of FR tied to these odd order narrow bandwidth distortions.
dlr.
These tests were all conducted with a closed box. A 28 inch long closed box with an 8 by 14 inch driver mounting face. A movable internal plate, with a gap equal to the area of the driver was also included. This plate was moved for the flattest FR across the pass band up to roll off. No damping material was placed in the box.
Tests were performed in a well damped living room with well marked locations for test box, microphone and large human body. I am certain all of this insults your personal objectives for repeatable transferable testing, but it is what I had to work with. Had I found results that I though were of more than incidental interest I would have pursued tests in a better facility.
Now that John and you have indicated the importance of what I thought, in my ignorance, were changes of less note than a change in humidity might cause, I will likely resurrect testing. Especially to look at this interesting odd order distortion node.
My thanks to both of you.
Bud
john k... said:
No it doesn't. You must have missed my earlier post.
Before EnABL i was modding the FE126 & FE127. This treatment added mass in very close to the same amount & distribution as my current drivers which also incorporate EnABL. T/S measures of some 200 drivers confirms that -- within limits imposed by the weather -- the added mass is the same. The new driver have the same tonal balance of the old driver but with notably better sonics (enuff so that i strongly feel that despite the new drivers 60% higher price it is better value). Most, if not all of the improvement is in the area of providing much more low-level information that fleshes in all the textures and space of the musical performance -- ie increased downward dynamic range. The difference is not subtle, and a newly exposed listener can take a while to get a handle on what is happening. This particularily happens when someone tries to analyze it. A music lover with no care to know why or what, just says it sounds more like music and can usually be seen to express this very quickly with their body language.
dave
That U tube sight has more than just the dust devils. Look at flics named Resonance, and the same name but followed with 2,3 &4. Resonance 4 is the one that interested me.
Bud
Bud
kaan,
I did reply to your first email. Did you receive any emails from me? The direct email address was somewhat unusual, but I did not receive a message undeliverable, mailer daemon. Perhaps you should send another to me and if you have an alternate edress, I can resend to both and we will see which makes it through.
Bud
I did reply to your first email. Did you receive any emails from me? The direct email address was somewhat unusual, but I did not receive a message undeliverable, mailer daemon. Perhaps you should send another to me and if you have an alternate edress, I can resend to both and we will see which makes it through.
Bud
Nice to hear that.
My direct mail is klaus at kaan got dk
Please try again.
And no im not afraid of no spam 😉
Edit:
Never mind that, I found your mail in my spam folder. I could have sworn that it wasnt there last time i looked 😉
My direct mail is klaus at kaan got dk
Please try again.
And no im not afraid of no spam 😉
Edit:
Never mind that, I found your mail in my spam folder. I could have sworn that it wasnt there last time i looked 😉
And finally, JOHN K. I apologize for some mental aberration that keeps prompting me to type John L. I assure you it is not my only mental abberation, just as EnABL is not the only oddity I have introduced to this world. Please forgive....
Bud
Bud
Bud, in those videos what you are seeing is dust being pushed into low energy areas between the higher energy modal areas.
There is no dust in areas of high radiation. It looks different from the Chladni plates because of geometry.
There is no dust in areas of high radiation. It looks different from the Chladni plates because of geometry.
dlr,
I used the full Blackman because it showed the data and the other choices were reluctant. I do not know a Blackman window from a sash style, so I used what worked. Even this was difficult to get all of the data within. I remember spending a lot of time fighting with the program just to get it to function and then pretty short test periods for the effort expended.
This was a single test run for FR/Phase, CSD and I think the distortion data was from a single run also, not a cut and paste set of tests. I have done those when measuring a full three way system with resistance ports, so I do know the difference. As for FFT I still don't think it was a pulse test that drove the CSD. My memory from that long ago is not to be trusted for that level of detail, though.
Again, this is the DOS version of LAUD, upgraded to run in Windows 98, or maybe 95, not certain. It is still an all DOS control scheme though.
The docking station has the original ECHO sound card with reworked power supply and coupling caps to allow low frequency operation. The microphone probe/preamp box is from Liberty Instruments and runs off of a 9 volt battery. The microphone came with it and a floppy disk with correction values for that microphone. All mounted on a mic stand with a 3 foot boom that the microphone hangs from, suspended in rubber bands.
The amplifier driving the speakers under test is a DC Servo design, built for me by Bob Munger of Audisar. He considered it to be a marvel of an instrumentation amp. it sounded coarse and grainy as an audio amp. I believe it is a Borberly design from Old Colony Sound.
As for questions about LAUD details, what you see on the print outs is what I got.
Bud
I used the full Blackman because it showed the data and the other choices were reluctant. I do not know a Blackman window from a sash style, so I used what worked. Even this was difficult to get all of the data within. I remember spending a lot of time fighting with the program just to get it to function and then pretty short test periods for the effort expended.
This was a single test run for FR/Phase, CSD and I think the distortion data was from a single run also, not a cut and paste set of tests. I have done those when measuring a full three way system with resistance ports, so I do know the difference. As for FFT I still don't think it was a pulse test that drove the CSD. My memory from that long ago is not to be trusted for that level of detail, though.
Again, this is the DOS version of LAUD, upgraded to run in Windows 98, or maybe 95, not certain. It is still an all DOS control scheme though.
The docking station has the original ECHO sound card with reworked power supply and coupling caps to allow low frequency operation. The microphone probe/preamp box is from Liberty Instruments and runs off of a 9 volt battery. The microphone came with it and a floppy disk with correction values for that microphone. All mounted on a mic stand with a 3 foot boom that the microphone hangs from, suspended in rubber bands.
The amplifier driving the speakers under test is a DC Servo design, built for me by Bob Munger of Audisar. He considered it to be a marvel of an instrumentation amp. it sounded coarse and grainy as an audio amp. I believe it is a Borberly design from Old Colony Sound.
As for questions about LAUD details, what you see on the print outs is what I got.
Bud
How is the distortion represented? Is it shown in the acoustic domain as distortion in the acoustic output or is it a distortion of the mechanical vibrational modes? Can it predict the distortion imposed in the acoustic domain? Can it be translated into the acoustic domain in some way that is predictive of the acoustic impact, since that is the domain of interest in a driver? This is not being cynical, I'm curious to know the details.
Have you ever studied a signal response on an O scope? We can talk theory or we can talk results. You keep asking questions that keep expanding without realizing the root effect. its the same as my students quoting a text book without doing an actual test.
I am not on this good earth to try to impress anybody, all i do is find answers to questions. Theory is all well and good but i have never found it to replace actual results.
ron
Have you ever studied a signal response on an O scope? We can talk theory or we can talk results. You keep asking questions that keep expanding without realizing the root effect. its the same as my students quoting a text book without doing an actual test.
I am not on this good earth to try to impress anybody, all i do is find answers to questions. Theory is all well and good but i have never found it to replace actual results.
ron
- Status
- Not open for further replies.