I don't care about "accusations" based on beliefs, you mentioned that twice -you seem to worry too much about other peoples' opinions.
Pushing for evolution is never redundant, just because you think that what we have today is enough and adequate. That's what pushes evolution in the first place (and you ignored the reasons I gave you above), otherwise we could have stayed in the caves enjoying the thrill of hunting for food and escaping wild animals -which was "adequate" to survive.
It reminds me the complaints about the "unnatural silence" and the "distracting" lack of hiss when the first CDs were published "what those engineers were thinking and wanted absolute silence?"
Can you back that claim? You might have read that somewhere or heard that somwhere. With stuff like that if you don't have the reference to the scientific study, claims turn to gospel.
Electrostatic headphone do not have more than 100 times less distortion of best dynamic ones, look at the innerfidelity measurements. In fact they do not have too many advantages over the best dynamic and planar headphones to justify the price, plus have have a rather poor sub bass response and require special amplification.
Even the commercial manufactures are struggling so hard with the technology to get a good and reliable electrostatic headphone driver that is within reasonable manufacturing budget. DIYing might be fun, but you will get poor results that perform worse mid range dynamic drivers.
Can you back that claim?
There are not many amplifiers out there that measure spectacularly good from 20Hz-20kHz(THD, IMD...), they are not easy to design. Why would you try to build something based on unfounded beliefs that would be so much worse than we already have?
What do you mean by that? If you mean that what we think we hear is strongly influenced by what we see, what we know, what others say and do, color, price, shape of equipment, yes that is true and old hat. Is that what you mean?
Jan
I can't answer for the OP, but ...
There is some evidence that maybe ... it's not well understood and it may not be true, but also it may be ... that we can detect intermodulation effects of frequencies created by frequencies both below and above the limit imposed by those little hair-like detectors in our ears.
We do know that we can detect frequencies below 50 Hz, yet there are not any hair cells at the organ of Corti that respond to frequencies below 50 Hz. The use of a 20~20KHz range dates back to the 1930's and the work of Fletcher-Munson. It's been useful, but maybe its time to have another look, here in the 21st century.
It is proposed by some that with regard to high frequencies, again there are specific receptors to individual frequencies at the upper limit but when two or more (so to speak) are stimulated, a different response is created in the brain than what would correspond to the specific frequencies they should respond to.
Or something like that ;-) My language might be a bit clumsy.
The answer, then, is maybe. Sorry for the off-topic post and I don't want to open a can of worms. Just something to think about, nothing more.
Part of the problem with auditory research as it's normally conducted is they don't care about audio frequencies in the way audio people do. Almost exclusively the research is focused on speech recognition, and there is a need for peer review and repeatability. In order to foster peer review, the equipment used is often what we would consider Mid-Fi or even worse, and there is a standard that is used where frequency is limited to a 200~8KHz bandwidth. This is of no use to us, for the most part. Mercifully that does not describe all auditory research but what is being done outside of those criteria is very limited and I would say incomplete.
You bring up a good point with regard to sighted / hidden sources. There is a thing, again part of the brain's interpretation of sound, where we seem to be able to hear more acutely if we have something to look at that represents the source of a sound. Well configured and constructed opaque (to light) screens no larger than necessary to hide all the DUTs and maybe more vertical in height than absolutely necessary (in other words, for a stereo system, at least two individual screens if not three, or perhaps just anything but a curtain) might be preferred to some methods used in tests (double-blind, or longer term subjective, doesn't matter), because it would give us something to look at, which seems important to hearing acuity.
Now that reads way off topic. Apologies once more.
Last edited:
True, but it is not only that we hear more, it is also the way the brain decides what we hear. That decision is formed based on several inputs, of which the signals from the ears form only one part. The visual senses also feed in the sound perception system, as well as internal states like remembered opinion, aesthetic circuits judging shape and colors, and a whole slew of others. In that sense you can 'hear' what you see. (And vice versa, for all other senses).
Jan
Their THD+noise plots should be taken with a big grain of salt. That's the conclusion I arrived after checking all their open-type THD measurements after your suggestion, then I did some reading on their methodology and I read the above phrase on their website -they admit it themselves!
Unfortunately they use 44.1Khz sampling rate which limits their THD measurements at 7khz (!) as the 3d harmonic goes up to 21khz and their equipment's limit is 22.05Khz (half the sampling freq). High-end precision measurements with only up to 7khz, really?
Also I noticed two high spikes of THD at 200hz and 2Khz in all their measurements and I guessed it was an artifact of their system, then I found out they admit that too.
And finally, they use an expensive test-head with microphones inside synthetic ears, to measure the sound with a distorted response which they equalize later.
That's a big mistake IMO because first, every head is different with different ears, hair etc, second, what's the point in distorting the sound and freq. response with a dubious synthetic imitation, with the overoptimistic unproven claim that the inside microphone resembles the human hearing membrane, only to (again) dubiously cancel all that later?
Add to this, that the head costs $20,000, yet the company doesn't bother to mention its own performance graphs and numbers on the datasheet -except some for the synthetic ...mouth. Obviously it is NOT intended for any precision measurements. Bottom line, the whole system is flawed.
So I cannot accept as reliable any of their data -until someone cares to provide a proper precision measurement methodology along with proof. Their methodology might be distorting all their measurements, closing the gap between large differences, hence the "big grain of salt".
Common sense is enough to figure out that the (practically) zero-mass diaphragm on electrostatics is incomparably more advantageous and promising in reaching low distortion at a flat and expanded response, than any dynamic one could ever dream of. It's simply and by far, the best we have.
Not really (if there was the slightest chance for that, I wouldn't bother in the first place). What makes the whole difference between a DIY construction and a commercial one, is research and development (R&D), special materials and special processing. If special processing and materials are not needed -or can be substituted successfully, then if a hobbyist did more R&D, he would end up with a much better final "product". Period.
Only DIYselfers who don't bother with R&D make "worse" constructions when they try something new.
That's a combination of things I've read in the past and personal experience, I can't provide a link to scientific evidence, but I might be able to back that claim after the project is finished.
"So much worse"? That's a very pessimistic assumption, based on what, on statistics? Statistics can only capture the average, not the unexpected. And "not easy" doesn't equal impossible. "Not easy" is just more interesting and more challenging. You'll see proof in the end, don't worry.
"Common sense" is a neat way of measuring the frequency response and distortion of an electromechanical system. Could you share how you do it? Do you look at the headphones, then go and sit in a dark room until the relevant figures appear in your head? Certainly beats all that expensive test gear which others have to use!MrMagic said:
That's your claim, not mine. Re-read above to understand what has been written. Or keep trolling...
Only if you think that diaphragm mass is the thing determining distortion and response. This shows how 'common sense' can lead you astray massively.
Jan
It's one of the many factors, but with an orders of magnitude heavier, highly resonating mass, you can't go far and definitely never as far. That's what I'm saying. With proper development you can minimize the rest of the factors and reach far better performance.
Most of the distortion comes from the frame, the reflections inside, the shape, the support (the way you hold and stretch the diaphragm), the mass of the frame, the other components' resonances like the stators, their stiffness, or flexibility (linear or not), the resonance of the holes, their shape, the air noise at the holes, etc.
If you do similar optimizations on a magnetic one, you are still left with a heavy diaphragm, cone, etc. Then you have to apply much higher energy to make that heavier mass respond which will create another circle of problems by causing stronger resonances, noise, etc.
Let alone the most important of all: You can never achieve the same amount of evenly applied force on the moving surface, with an electromagnetic field.
Bottom line, like I said, the electrostatics are far more promising to achieve the best performance possible. They are already awesome, most reviewers agree that they are top, and they can be made far better IMO if more R&D was done.
Even according to the flawed measurements mentioned previously, only one dynamic headset comes close to the best (SR009), the planar Audeze LCD-X (at $1700), but if you look closer, you'll see that it's response in the higher frequencies is about 6db lower than the electrostatic. So no matter how hard you try with a heavy diaphragm, cone, etc, you can't go as far.
End of the "common-sense" seminar.
Last edited:
It's one of the many factors, but with an orders of magnitude heavier, highly resonating mass, you can't go far and definitely never as far. That's what I'm saying. With proper development you can minimize the rest of the factors and reach far better performance.
Most of the distortion comes from the frame, the reflections inside, the shape, the support (the way you hold and stretch the diaphragm), the mass of the frame, the other components' resonances like the stators, their stiffness, or flexibility (linear or not), the resonance of the holes, their shape, the air noise at the holes, etc.
If you do similar optimizations on a magnetic one, you are still left with a heavy diaphragm, cone, etc. Then you have to apply much higher energy to make that heavier mass respond which will create another circle of problems by causing stronger resonances, noise, etc.
Let alone the most important of all: You can never achieve the same amount of evenly applied force on the moving surface, with an electromagnetic field.
Bottom line, like I said, the electrostatics are far more promising to achieve the best performance possible. They are already awesome, most reviewers agree that they are top, and they can be made far better IMO if more R&D was done.
Even according to the flawed measurements mentioned previously, only one dynamic headset comes close to the best (SR009), the planar Audeze LCD-X (at $1700), but if you look closer, you'll see that it's response in the higher frequencies is about 6db lower than the electrostatic. So no matter how hard you try with a heavy diaphragm, cone, etc, you can't go as far.
End of the "common-sense" seminar.
Good. Can we go back to facts and figures of reality now?
What is the dominant mechanism for distortion in an ESL/ESHP?
Are you aware of Walker's work on that?
Jan
Engineers do not "abandon" or "drop" products, companies do. You are judging a specific technology by its sales, confusing marketing with technology.
I've read a lot in the past, I can't recall the name. The major sources of distortion and imperfect response, are by far of mechanical type, like the examples I gave you above, for a properly biased ESL/ESHP. Are you referring to the variations of the charge during motion?
Since there are complaints about being off-topic, you could open a relevant thread, to stop hijacking this one, and we could continue there.
Last edited:
Well that says it all. The guy who wrote the book on ESLs and you can't recall the name.
I don't see the need for a specific thread on this, thanks.
Jan
Walt's 1979, two-part article on SID is good if somewhat laborious to read through.
It is really very, very simple. The electrical property that resists change in voltage is capacitance. Amplifiers contain capacitance and may be driving a capacitive load. To make the output voltage change you have to charge the relevant capacitances. Current supply is always limited in each part of the circuit and therefore there is a limit to how fast those capacitances can be charged. That's it in a nutshell.
The slew rate is, for example, often limited by the LTP CCS driving the Miller capacitance. A typical LTP tends to get quite distorted as it swings beyond 50% of maximum current.
A perfectly functioning amp with undistorted CD music source and low noise should not need to produce more than 0.125V/us per peak V. For an LTP/Miller limited amp you can double this to give the LTP linearity margin, 0.25v/us/Vpk or about 10V/us for 100W average into 8 ohms.
It's really not worth ruminating about because the vast majority of home-use, discrete amplifier circuits can easily meet the slew rate spec. Many, especially older, op-amps do not and that is what Walt wrote about in 1979.
It is really very, very simple. The electrical property that resists change in voltage is capacitance. Amplifiers contain capacitance and may be driving a capacitive load. To make the output voltage change you have to charge the relevant capacitances. Current supply is always limited in each part of the circuit and therefore there is a limit to how fast those capacitances can be charged. That's it in a nutshell.
The slew rate is, for example, often limited by the LTP CCS driving the Miller capacitance. A typical LTP tends to get quite distorted as it swings beyond 50% of maximum current.
A perfectly functioning amp with undistorted CD music source and low noise should not need to produce more than 0.125V/us per peak V. For an LTP/Miller limited amp you can double this to give the LTP linearity margin, 0.25v/us/Vpk or about 10V/us for 100W average into 8 ohms.
It's really not worth ruminating about because the vast majority of home-use, discrete amplifier circuits can easily meet the slew rate spec. Many, especially older, op-amps do not and that is what Walt wrote about in 1979.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.