Nice for the price.
Designing up a magnetic fixture. I originally was thinking just a stupid loop using 1 inch threaded rod and angle iron fixed to a piece of plywood. May still as a precursor, but have started chomping on something that can handle a 3 or 4 inch dia voice coil. I was surprised that 4 inch diameter bog cold rolled steel was 18 dollars an inch. 1 by 2 rectangular wa much better.
May have to look local, shipping from online metals would be more than the steel.
Jn
Designing up a magnetic fixture. I originally was thinking just a stupid loop using 1 inch threaded rod and angle iron fixed to a piece of plywood. May still as a precursor, but have started chomping on something that can handle a 3 or 4 inch dia voice coil. I was surprised that 4 inch diameter bog cold rolled steel was 18 dollars an inch. 1 by 2 rectangular wa much better.
May have to look local, shipping from online metals would be more than the steel.
Jn
I bought some previously, and they were so good (at the price) I bought
more this Xmas and gave them away as presents.....
If they have your blessing, then I think £64 + whatever shipping is alright. Do you think your H2 generator will make them even better?
Zung, do you understand what I am complaining about with my $28,000 set of loudspeakers? It is the fact that the midrange driver is wired inverted in polarity, and this, while seeming to improve frequency response, compromises the time response. Now when it comes to cheap speakers, thanks for the recommendation.
PMA, you are right when it comes to classical music. The Sequerra Met7's are not optimum, because they lack both extreme lows, extreme highs, etc. BUT their TIME RESPONSE, is pretty darn good! And I hear it.
Now, I appreciate classical music, but I was not raised with it like many Europeans, and only in college did I begin to listen to it in any real way. Now, I once married a classical violinist, worked at the Institute for Advanced Musical Studies (IHEM) in Switzerland for a couple of years, and had many dozens of classical musicians as close friends, back then, but now I just tend to listen to female voice, some old rock and roll (my youth), or my new found Blues channel at this time. IF I wanted to listen to classical music, I would indeed turn on my Wilson Sashas, as they would do the job better. Even better in some ways, for classical, was my old K-horns that my associate now owns. I used them for about 15years, and when I could, I turned them up to realistic levels, to listen to Finlandia, ballet music, (I once studied ballet) etc that required such a speaker to sound right. However, the K-horns had an even more severe TIME RESPONSE problem than the Sashas do today, so after much controversy, I sold them to replace them with LS3-5A's with a subwoofer, for several years. The A's just sounded more natural on human voice. Darn! Today, I find the Sequerra Met 7's to sound pretty good also with human voice, and as long as I do not play them too hard, the distortion seems reasonable to me. I use them mostly for voice anyway. What I wanted to discuss with you, PMA, was the difference in time response between the Sashas and the Sequerra's and how it could change the sound. I suspect that I asked too much, and you don't think it that important. Oh well.
PMA, you are right when it comes to classical music. The Sequerra Met7's are not optimum, because they lack both extreme lows, extreme highs, etc. BUT their TIME RESPONSE, is pretty darn good! And I hear it.
Now, I appreciate classical music, but I was not raised with it like many Europeans, and only in college did I begin to listen to it in any real way. Now, I once married a classical violinist, worked at the Institute for Advanced Musical Studies (IHEM) in Switzerland for a couple of years, and had many dozens of classical musicians as close friends, back then, but now I just tend to listen to female voice, some old rock and roll (my youth), or my new found Blues channel at this time. IF I wanted to listen to classical music, I would indeed turn on my Wilson Sashas, as they would do the job better. Even better in some ways, for classical, was my old K-horns that my associate now owns. I used them for about 15years, and when I could, I turned them up to realistic levels, to listen to Finlandia, ballet music, (I once studied ballet) etc that required such a speaker to sound right. However, the K-horns had an even more severe TIME RESPONSE problem than the Sashas do today, so after much controversy, I sold them to replace them with LS3-5A's with a subwoofer, for several years. The A's just sounded more natural on human voice. Darn! Today, I find the Sequerra Met 7's to sound pretty good also with human voice, and as long as I do not play them too hard, the distortion seems reasonable to me. I use them mostly for voice anyway. What I wanted to discuss with you, PMA, was the difference in time response between the Sashas and the Sequerra's and how it could change the sound. I suspect that I asked too much, and you don't think it that important. Oh well.
Thank you for your interest in the method.
The method is not based on wavelets, but rather on linear systems theory and system identification. Instead of analyzing a response of a linear system by Fourier-transforming its output, this method tries to find the transfer function of a candidate linear system that produces the signal at hand from white noise at its input. The transfer function obtained is then further analyzed.
I know this sounds complicated, so I'm enclosing the paper I presented at a conference three years ago, which contains an outline of the underlying theory. The paper is in public domain, but not easy to obtain. It does not deal with audio, but with measurements on rotating machinery; and in the concrete case the object was a ship's Diesel engine. However, the method is universally applicable; I learned it from the literature cited in the paper.
Note that there is an error in the paper in that Eq. (3-8) should be in terms of the cosine function instead of the sine as it stands. My apologies.
I did this work back in 2003, but the company I worked for only allowed me to publish it three years ago.
The problem to be solved at the time was to somehow obtain the instantaneous angular velocity of the engine crankshaft, which was needed for the data analysis, but was missing in the measured data set (and a very expensive one at that). The source for this information was found in the pressure pulsations measured in the exhaust channel of the engine, which are closely related to the crankshaft rotation.
The rest is in the paper.
Feel free to ask questions; I'll do my best to answer them.
Regards,
Braca
I'm totally with you. Don't you think the tweeter is out of phase in the graph I got from the Stereophile? Or is it some sereve undershoot from the tweeter?
Last edited:
I do not think that the Met 7 step response is good. It is unipolar, but that's not everything. The response reflects all those irregularities seen in frequency response. Former Thiel speakers also emphasized time response. Similarly, they had another issues. I know several owners that were very disappointed with them and sold them very quickly, loosing a lot of money. I also am not a big fan of that female voice test. Or maybe yes, but then it would be Cecilia Bartoli together with orchestra and both, together, has to sound perfect. We have different views of evaluation of "sound quality", let's leave it as it is.
Tweeter "out of phase" is usually an inevitable result of crossover design. You cannot just reverse it, it would be simply silly and would result in a deep dip on a crossover frequency. Crossover order, filter characteristics and driver's responses define how to set the polarity.
All that all-pass audibility is almost always a hoax.
All that all-pass audibility is almost always a hoax.
Sorry but it's not time response you're hearing. The phase changes all over the place on the drivers, having one's polarity reversed isn't going to do much of anything for "time alignment".
The reason they would have likely done it is that at the crossover point they had comb filtering due to the distance between the tweeter and mid. Changing the polarity can fix that, or most of it. Wilson is known for not placing the tweeters and mids particularly close...
i'm well aware that my mastery of English language is... unorthodox.
I count on your indulgence.
i wanted to say "to linearize the delay curve". And the lost specific semantic meaning was to suppose that, may-be, the absence of accident in it could be more important that its absolute horizontality.
When I was messing around with tuning engines, once affordable wideband 02 sensors were available it was interesting to look at how the mixture varied over a cycle. Talking to some friends who had worked on F1 engine management they did say that they worked on second derivative of crank speed as things were accelerating too fast. Never found corroboration of that to see if true. But I always found engine management to be fascinating stuff when you are pushing the limits and I've only scratched the surface.
Thanks for all the measurements PMA.
Magnetic nonlinearities act on the cone through the same mechanism as the input signal, as force on the voicecoil. So magnetic distortion spectral components should not be influenced differently by cone geometry than the main signal. It is the mechanical nonlinearities that are modified due to where they originate in the mechanical system. So I think it is okay to reference these measurements as far as the magnetics are concerned.
As for 3rd harmonics, I have been laboring under the assumption that there was an eddy current mechanism for creating them. I think some people took this to mean eddy currents causing saturation of the steel, which I didn't understand because the eddy currents are actually opposing the flux which causes saturation. Furthermore, the steel in the speaker is biased by the magnetic field from the magnet, so it would be predominantly 2nd harmonic unlike George's PDFs where steels are analyzed based on their odd harmonics.
If like George says the steel is heavily saturated by the magnetic field, this means it is either deep in saturation or it is on a sharp part of the saturation curve, which I would think would tend to produce high order harmonics. I haven't implemented B saturation yet but it is on the list.
I will try again ignoring odd harmonics and see if just the even harmonics make a good fit.
Magnetic nonlinearities act on the cone through the same mechanism as the input signal, as force on the voicecoil. So magnetic distortion spectral components should not be influenced differently by cone geometry than the main signal. It is the mechanical nonlinearities that are modified due to where they originate in the mechanical system. So I think it is okay to reference these measurements as far as the magnetics are concerned.
As for 3rd harmonics, I have been laboring under the assumption that there was an eddy current mechanism for creating them. I think some people took this to mean eddy currents causing saturation of the steel, which I didn't understand because the eddy currents are actually opposing the flux which causes saturation. Furthermore, the steel in the speaker is biased by the magnetic field from the magnet, so it would be predominantly 2nd harmonic unlike George's PDFs where steels are analyzed based on their odd harmonics.
If like George says the steel is heavily saturated by the magnetic field, this means it is either deep in saturation or it is on a sharp part of the saturation curve, which I would think would tend to produce high order harmonics. I haven't implemented B saturation yet but it is on the list.
I will try again ignoring odd harmonics and see if just the even harmonics make a good fit.
i'll stay corrected. Thanks J.C.
Next time, I'll use "perhaps".
(In French, is is "peut-être").
To answer your question, it seems obvious, which is why I treated it with "mokusatsu" and leave again this so friendly place.
Oh, Lord, we have a proverb that says "La musique adoucit les moeurs.".
Last edited:
- Status
- Not open for further replies.