JMFahey,
I tried zipping, it doesn't do much. The mp3 codec is technically lossy in that it throws out lots of info that is really there, true. But, from a psychoacoustic standpoint, it throws out stuff that your ears probably won't miss if you sample at the higher bit rates. The research that went into developing the mp3 algorithm at The Fraunhoffer institute showed that the human ear/brain is essentially a 32 freq (channels) real time frequency analyzer. That is if you have more than 32 Fast Fourier Transform coeficients, the resulting audio signal is indiscernable to the average ear. This is how mp3 is able to achieve such dramatic compression ratios that for the most part are quite acceptable for most listeners. I agree that when I take out my actual CD's and play them vs my iPod, I hear differences but some of that could be the DAC and preamp stage. I will put together a Howto for recording and uploading sound samples as suggested earlier by another member.
Nice looking kids by the way, hope they don't lose their hearing....
I tried zipping, it doesn't do much. The mp3 codec is technically lossy in that it throws out lots of info that is really there, true. But, from a psychoacoustic standpoint, it throws out stuff that your ears probably won't miss if you sample at the higher bit rates. The research that went into developing the mp3 algorithm at The Fraunhoffer institute showed that the human ear/brain is essentially a 32 freq (channels) real time frequency analyzer. That is if you have more than 32 Fast Fourier Transform coeficients, the resulting audio signal is indiscernable to the average ear. This is how mp3 is able to achieve such dramatic compression ratios that for the most part are quite acceptable for most listeners. I agree that when I take out my actual CD's and play them vs my iPod, I hear differences but some of that could be the DAC and preamp stage. I will put together a Howto for recording and uploading sound samples as suggested earlier by another member.
Nice looking kids by the way, hope they don't lose their hearing....
True, average listeners may not care and so it works for them.
For me, mp3's (even higher bit-rate) don't do a good job on certain parts of acoustic music. For example, cymbals can become swooshy. Very long decays can taper off sooner than they should.
Just as a simple experiment, take an acoustic song (as WAV), convert to mp3, then convert to WAV, then re-compress to mp3. The shortcomings start to become more noticeable (often, and on certain tracks anyway).
For me, mp3's (even higher bit-rate) don't do a good job on certain parts of acoustic music. For example, cymbals can become swooshy. Very long decays can taper off sooner than they should.
Just as a simple experiment, take an acoustic song (as WAV), convert to mp3, then convert to WAV, then re-compress to mp3. The shortcomings start to become more noticeable (often, and on certain tracks anyway).
The "swishing" is typically due to the use of jstereo (joint stereo) rather than true stereo. Jstereo allows better compressibility at the lower bitrates, but many mp3 encoders have this mode set to default, regardless of the bitrate selected by the user. At 256k and up, true stereo can be used and this should basically eliminate any swishing.
Another thing that goes wrong with mp3 conversion is "normalization." Often this is also set to a default position of "on" in mp3 encoders and it can result in levels that are way too high, reduction of dynamic range, clipping in the source, etc. Sort of like all that badly "remastered" crap out there these days. Levels cranked to clipping never sounds good, and sounds even worse when compressed to jstereo at 128k!
I prefer FLAC, and I try to get the 96/24 stuff when I can find it, but I can certainly tolerate mp3 if they are proper mp3. 256 or 320 in true stereo is easily tolerable, especially pushed through a good DAC and the B-283.
One has to be very careful comparing one set of measured T/S to another taken on a differnt system. You cannot be sure where on the T/S curves the system measures.
My measures across over 100 stock drivers puts Fs at ~74 Hz. Treatment lowers Fs a bit.
That my measure is so close to factory indicates that Mark's drivers have a close to horizontal set of T/S curves.
For example factory Fs on FE127 was 70 Hz, i typically measures 90 Hz. You can see a specific example of a set of drivers i measured, then sent to Mark and he measured.
Anyone who thinks break-in is imaginary has not heard the right demonstration and is denying the effect on a mechanical system of being used from freahly made. Magnitude & direction will depend on the design & the skill of the designer.
Anyone who thinks that Cogitech is deluding himeself are mistaken. What he did is an anecdotal experience which he found meaningful -- and he said so upfront, take it for what it is. A member havng an experience he found meaningful.
dave
Attachments
+1 except i wouldn't qualify it.
The idea of taking the signal out of a system, capturing it with a mic (even a hi-quality calibrated unit), multilating it into an MP3 and then listeing to it on a different system, and thinking that it tells you anything realistic is highly optimistic.
IMO, does more harm than good.
dave
Indeed. The ensuing argument would be never-ending, which wouldn't change things in the least with respect to the question of break-in. It would only take the argument in a non-productive direction, criticizing the methods of recording, processing/compressing, and playing back of the samples. Even if all those things went right, there would still be people out there who simply say "I can't hear the difference, so it proves that break-in is imaginary". And the whole cycle would begin again.
I don't agree that it does more harm than good. C'mon... It's like saying we shouldn't post digital photos of our speakers because it doesn't capture what our eye really sees because it is only 24 bit RGB, white balanced at such and such, and jpeg compression has artifacts.... A picture is worth a thousand words... And a sound byte is maybe worth 5000 words?
Your example in the Cornu thread serves a purpose. In this particular case, I don't think it would change the direction or volume of argument.
Well, since we're listening for differences which are claimed to be large, even a mediocre recording should show discernible differences, right?
Compared to resistor, capacitor, and wire nonsense, this is at least plausible. But I'd sure like to see high resolution before-and-after impedance measurements...
Compared to resistor, capacitor, and wire nonsense, this is at least plausible. But I'd sure like to see high resolution before-and-after impedance measurements...
Sy,
I think that what you are looking at is a change in compliance of the suspension and not a change in impedance. I can see the impedance change after running a speaker at high current loading as the voicecoil and motor assembly heat up. This would be power compression due to a rise in electrical impedance. But this has nothing to do with the physical change of a loudspeaker over time, the heating is only a transitory change dependent on power input. What you are actually looking for is the change in the spider and sometimes a change in the surround if it is doped paper or cloth. It should be fairly easy to measure the physical change in the spider by a simple displacement test. Take a speaker as new and measure the deflection of the cone with a given weight added. Now break it in by running a low frequency tone for several days at high excursion and repeat the test. Make sure the speaker is back to ambient temperature conditions and remeasure the displacement with the same weight. It will not be the same, the spring rate of the spider will have changed. it will have a lower spring rate. This is what you are seeing in the change of the fs before and after. Change the numbers in the T/S parameters for compliance and this is the break-in mode described. Now I will not say this is the only change but the most significant value. A paper cone may also go through a much slower change over time as the fibers in the cone also will soften and break over time and this will also make changes in the frequency response curve but not of the fs value. This would be much harder to measure and isolate from the change in suspension parameters. I would think that you could also test this by disassemble of a cone and measuring its deflective properties also from new. This remove the electrical measurements and requires only the addition of a fixed weight and a height gauge to measure deflection.
I think that what you are looking at is a change in compliance of the suspension and not a change in impedance. I can see the impedance change after running a speaker at high current loading as the voicecoil and motor assembly heat up. This would be power compression due to a rise in electrical impedance. But this has nothing to do with the physical change of a loudspeaker over time, the heating is only a transitory change dependent on power input. What you are actually looking for is the change in the spider and sometimes a change in the surround if it is doped paper or cloth. It should be fairly easy to measure the physical change in the spider by a simple displacement test. Take a speaker as new and measure the deflection of the cone with a given weight added. Now break it in by running a low frequency tone for several days at high excursion and repeat the test. Make sure the speaker is back to ambient temperature conditions and remeasure the displacement with the same weight. It will not be the same, the spring rate of the spider will have changed. it will have a lower spring rate. This is what you are seeing in the change of the fs before and after. Change the numbers in the T/S parameters for compliance and this is the break-in mode described. Now I will not say this is the only change but the most significant value. A paper cone may also go through a much slower change over time as the fibers in the cone also will soften and break over time and this will also make changes in the frequency response curve but not of the fs value. This would be much harder to measure and isolate from the change in suspension parameters. I would think that you could also test this by disassemble of a cone and measuring its deflective properties also from new. This remove the electrical measurements and requires only the addition of a fixed weight and a height gauge to measure deflection.
Kindhornman got it right, the primary change is in compliance of the spider and suspension and it is quite measurable, fs, vas, and qts all change to some degree as the suspension components break in. Take a look at my earlier post about my experiences at a certain mid-fi audio manufacturer, where the parametric changes over relatively large driver sample sizes over a 24 - 48 hour run in time were well documented.
Anyone who dabbles in loudspeaker design will generally run the woofers and other coned drivers in before doing box design based on measured TS parameters as they will shift appreciably in the first hours of use.
While I no longer remember all of the details the in room measured bass performance of my diy Onken bass bins changed measurably and audibly in the first 100 hours of operation. I have not observed or measured any significant changes in the years since that could not be traced to other changes in the system or changes in speaker placement.
I would have a look at what Vance Dickason has to say on this subject as well as talk to anyone who actually designs speaker systems for a living. They should have a good sense of what is real and the physics behind it.
Anyone who dabbles in loudspeaker design will generally run the woofers and other coned drivers in before doing box design based on measured TS parameters as they will shift appreciably in the first hours of use.
While I no longer remember all of the details the in room measured bass performance of my diy Onken bass bins changed measurably and audibly in the first 100 hours of operation. I have not observed or measured any significant changes in the years since that could not be traced to other changes in the system or changes in speaker placement.
I would have a look at what Vance Dickason has to say on this subject as well as talk to anyone who actually designs speaker systems for a living. They should have a good sense of what is real and the physics behind it.
I am only thinking out loud here, but wouldn't the same principle apply to rubber (or similar material) surrounds? It seems to me that repeated flexing/movement of any of these suspension materials would result in a change in the spring rate and hence the corresponding change in Fs. Not just the material itself, but whatever adhesive used to bond the suspension materials to the paper/metal/etc. cone surfaces might also "loosen" and then re-settle into a more static state which allows for "freer" total cone travel with less points of stress to hamper said movement.
Changes in the front suspension will affect things, but from postings by Mark on his drivers, the surround primarily only supplies a centering function and little of the actual suspension. For instance in the A12.2p a special jig is used for assembly because the front suspesion contributes almost nothing to the driver compliance.
dave
I have read somewhere that the spider stands for ~80% of the compliance. Might be different from driver to driver though but I think it's a pretty good generalization.
cogitech,
The change in the surround if a polymer would be much more subtle in its change. We have all seen foam surrounds that turn to dust after many years and that is one of the problems with most foamed plastic materials. The natural rubber surrounds also are affected and most of this is caused by ozone damage in the air and a hardening of the rubber over time. I use only polyurethane thermoplastic surrounds that do not have these problems. They are inert and besides a chance at a color shift really don't have a natural degradation property. They are not biodegradable in essence. I would guess that 99% of all spiders are of the cloth/phenolic construction method. What happens is that the phenolic resins used are rigid thermoset materials. The movement of the loudspeaker actually cracks the phenolic resin, the higher the displacement used the more this happens. You are actually cracking the material in the U-shaped folds of the spider, it is a very understandable phenomena if you understand the materials used. As kevinkr stated if you want to get any true T/S parameters you need to first exercise the finished loudspeaker. I do that by running a cone, no box needed, on a sine wave at or near maximum xmax to break in the device.
The change in the surround if a polymer would be much more subtle in its change. We have all seen foam surrounds that turn to dust after many years and that is one of the problems with most foamed plastic materials. The natural rubber surrounds also are affected and most of this is caused by ozone damage in the air and a hardening of the rubber over time. I use only polyurethane thermoplastic surrounds that do not have these problems. They are inert and besides a chance at a color shift really don't have a natural degradation property. They are not biodegradable in essence. I would guess that 99% of all spiders are of the cloth/phenolic construction method. What happens is that the phenolic resins used are rigid thermoset materials. The movement of the loudspeaker actually cracks the phenolic resin, the higher the displacement used the more this happens. You are actually cracking the material in the U-shaped folds of the spider, it is a very understandable phenomena if you understand the materials used. As kevinkr stated if you want to get any true T/S parameters you need to first exercise the finished loudspeaker. I do that by running a cone, no box needed, on a sine wave at or near maximum xmax to break in the device.
Very interesting. Thanks for all this info (and to Dave and others, too).
So, my next question is "what is the surround on the EL70 made of?" It sure looks like natural rubber, to me. If so, how can I prevent it from hardening over time?
Also, Dave, when you apply the eN treatment, do you mask the surround? It sure looks like you did, as the surrounds have maintained a matte appearance as opposed to the satin sheen of the cone and bezel.
EDIT: EL70 spec sheet says "proprietary rubber"
Last edited:
I thought it was obvious that any mechanical movement causes wear. Look at gasoline engines or dc motors with brushes. Friction results in wear.
The pro break-in people seem to assert that breaking in speakers results in better sound. Surround degradation would not seem to be beneficial for sound though. Are the pro break-in people concluding that suspension and surround wear is always a good thing?
It appears that nothing can be concluded until we can set up an experiment with several different identical drivers. We need to be able to test a "new" driver, a "broken-in" driver, and a "worn" driver. The null hypothesis would be that break-in has no effect. Alternative hypothesis would be that a certain amount of break-in has either a positive or negative effect on sound.
Until we can see a test like this, it doesn't appear that either side can add anything meaningful to the conversation. As far as we know, we don't know if break-in is even a process that changes sound production meaningfully.
Does anyone disagree? Should we start a drive to purchase three identical woofers for testing?
The pro break-in people seem to assert that breaking in speakers results in better sound. Surround degradation would not seem to be beneficial for sound though. Are the pro break-in people concluding that suspension and surround wear is always a good thing?
It appears that nothing can be concluded until we can set up an experiment with several different identical drivers. We need to be able to test a "new" driver, a "broken-in" driver, and a "worn" driver. The null hypothesis would be that break-in has no effect. Alternative hypothesis would be that a certain amount of break-in has either a positive or negative effect on sound.
Until we can see a test like this, it doesn't appear that either side can add anything meaningful to the conversation. As far as we know, we don't know if break-in is even a process that changes sound production meaningfully.
Does anyone disagree? Should we start a drive to purchase three identical woofers for testing?
Not masked per se, but efforts are made to ensure that the flexible part of the surround does not get any of the treatment material son it.
dave
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.