sound ---> ears ---> brain
The point I was and am trying to make is that in this path of sound ---> ears ---> brain the last part is doing an active job of "decoding" raw information.
That's why the piece of music that I cited is so interesting.
The hit on the music stand is there, but the brain only decodes it on rare occasions. It's there all the time in the recording and the reproduction on my system does not change moment to moment or day to day. Not enough to cause this bit of sonic detail to appear and disappear.
Well, there are tools and parts that do appear and disappear on my workbench, and I theorize that they go elsewhere - like to a workbench in Asia or maybe somewhere in Ohio for a brief time - and then reappear...
But I don't think this happens in my system...
So, what is happening?
I'm here to say that the way you and your brain decodes sound is not a constant at all. The decoding of sound is not an incremental process, nor is it exactly a continuous process either... it does seem to happen in some sort of overlapped time segments... a bit like reading one of those scrolling signs, like the one in Times Square NYC... but unlike that sign where a section of text is present all at the same time, with sound only an instantaneous bit of energy is available at any discrete moment in time. The brain is doing the integration over time.
So, here in the most basic sense the variables logically present themselves. For example: what is the integration time? How long? Does it vary? On what basis? Does it vary depending on the complexity of information in any given time frame? How so? What happens when the time frame is very long, when it is very short? How does this effect the perception? How are the time frames strung together? Are they strung together to form a greater whole? What about when there are other sounds in the environment - including sub sonics and ultrasonics, low level vibrations, broadband noises (A/C, air vents, etc...)?? What effect does stress have upon how the brain does this job?
It's my view that the better the source of the sound - ie. the system's presentation and baseline abilites - the less work the brain has to do per unit time/time frame to make sense of what is being presented. Remember, the key to hearing is making sense of the sound! The less work the brain has to do to make the basic decisions, decipher the cues, decode the sound, the more time it has to work on the more subtle cues and sounds!
A good example of how this happens, so you can either recall it happening to you, or make note of it the next time, is when you tune in to a radio program playing music in the middle of a song and you do not recognize the music at all until a bit of time passes. In fact it sounds completely unique and unrecognizable for a bit of time.
Now, what qualities have to be conveyed to the listener in order for the brain to have the easiest job of decoding is really what the discussion is about.
That means that the true answers do not lie in just a single driver, or which driver or even speaker is "best" at all. That's too narrow a definition, and doesn't truly address the ultimate issues or problem. What is needed is to somehow figure out how best to convey the information through the air to the ear, and then that is in light of what the ears need to receive in order to most readily do that decoding job.
The discussions on things like "Ambiphonics" and "Stereolith" show us that what the ears "need" is not quite so obvious, and the means to convey it are neither simple nor pat.
So whaddya think?
_-_-bear
The point I was and am trying to make is that in this path of sound ---> ears ---> brain the last part is doing an active job of "decoding" raw information.
That's why the piece of music that I cited is so interesting.
The hit on the music stand is there, but the brain only decodes it on rare occasions. It's there all the time in the recording and the reproduction on my system does not change moment to moment or day to day. Not enough to cause this bit of sonic detail to appear and disappear.
Well, there are tools and parts that do appear and disappear on my workbench, and I theorize that they go elsewhere - like to a workbench in Asia or maybe somewhere in Ohio for a brief time - and then reappear...
But I don't think this happens in my system...
So, what is happening?
I'm here to say that the way you and your brain decodes sound is not a constant at all. The decoding of sound is not an incremental process, nor is it exactly a continuous process either... it does seem to happen in some sort of overlapped time segments... a bit like reading one of those scrolling signs, like the one in Times Square NYC... but unlike that sign where a section of text is present all at the same time, with sound only an instantaneous bit of energy is available at any discrete moment in time. The brain is doing the integration over time.
So, here in the most basic sense the variables logically present themselves. For example: what is the integration time? How long? Does it vary? On what basis? Does it vary depending on the complexity of information in any given time frame? How so? What happens when the time frame is very long, when it is very short? How does this effect the perception? How are the time frames strung together? Are they strung together to form a greater whole? What about when there are other sounds in the environment - including sub sonics and ultrasonics, low level vibrations, broadband noises (A/C, air vents, etc...)?? What effect does stress have upon how the brain does this job?
It's my view that the better the source of the sound - ie. the system's presentation and baseline abilites - the less work the brain has to do per unit time/time frame to make sense of what is being presented. Remember, the key to hearing is making sense of the sound! The less work the brain has to do to make the basic decisions, decipher the cues, decode the sound, the more time it has to work on the more subtle cues and sounds!
A good example of how this happens, so you can either recall it happening to you, or make note of it the next time, is when you tune in to a radio program playing music in the middle of a song and you do not recognize the music at all until a bit of time passes. In fact it sounds completely unique and unrecognizable for a bit of time.
Now, what qualities have to be conveyed to the listener in order for the brain to have the easiest job of decoding is really what the discussion is about.
That means that the true answers do not lie in just a single driver, or which driver or even speaker is "best" at all. That's too narrow a definition, and doesn't truly address the ultimate issues or problem. What is needed is to somehow figure out how best to convey the information through the air to the ear, and then that is in light of what the ears need to receive in order to most readily do that decoding job.
The discussions on things like "Ambiphonics" and "Stereolith" show us that what the ears "need" is not quite so obvious, and the means to convey it are neither simple nor pat.
So whaddya think?
_-_-bear
MLSSA sort of measures this. They provide a measure of any output of the DUT that is not correlated with the input signal. Well beyond THD, IM or noise, its a measure of all the "junk" that shouldn't be there, plotted in amplitude across frequency.
I think they originally intended it as a sort of "rub and buzz" test for industrial applications.
Dave, also
Isn't it very obvious that human can be so deeply attracted by sound system ability to reproduce believable sound-stage?
Yes, there are many "kinds" of sound-stage. Just pick one that suits your taste and you're done
(the rest is just cosmetics)I would throw in a vote for the Eton line of drivers - used in Avalon and other first rate commercial speakers. I have personally built systems using the 7/380 and the sonic textures are pretty amazing once you get it dialed in.
One thing I would note is that the quality of the cross over (for any driver) has a massive impact on the acoustic results. A poor crossover could easily explain one person loving the detail a driver produces and another thinking the same driver was a mid-fi rip off.
One thing I would note is that the quality of the cross over (for any driver) has a massive impact on the acoustic results. A poor crossover could easily explain one person loving the detail a driver produces and another thinking the same driver was a mid-fi rip off.
I suspect that with electrostats, two things are responsible. First they are driven over the entire radiating area. Second, they have very low mass per unit of radiating area. A good stat has an almost perfect watefall plot. So do some conventional speakers. These are usually the ones with most detail. So in short, transient response, transient response, transient response. Music is a series of transient so............ Kindest regards
They are more directional vertically and horizontally than point source type designs. That gives them a higher direct to reverb ratio, and less masking of detail by room contribution.
Hi DDF,
that may be true, but they are also dipoles and radiating backwards this increases indirect sound. The horizontal dispersion can also be quite good (on par with dynamic speakers if properly designed.
The only thing what really acounts is that it is highly directive in the vertical direction, so eliminating floor reflection to a large degree. I don't know if this explains the whole thing.
Stax headphones are electrostatics and are also highly rated for their detail
that may be true, but they are also dipoles and radiating backwards this increases indirect sound. The horizontal dispersion can also be quite good (on par with dynamic speakers if properly designed.
The only thing what really acounts is that it is highly directive in the vertical direction, so eliminating floor reflection to a large degree. I don't know if this explains the whole thing.
Stax headphones are electrostatics and are also highly rated for their detail
Hi MJ,
The dipoles have a higher directivity index than standard 2 or 3 ways, so they radiate less indirect sound.
I have an old pair of Stax headphones. Only a sample of one (and old) but they aren't that detailed compared even to my D2000s
I second that high acoustic resistance low mass. That is also the case with horn-drivers.
But a dipole has its special radiation behavior. And a horn its diffenrent path length problem.
Some models of Stax (like mine) do have a direct drive amplifier so they lack a audio-transformer (which possibly degrades the sound).
I also listened to Stax with a conventional transformer and to me the sound quality wass less so I bought the current direct drive one.
Regarding loudspeakers, could Doppler distortion be a factor?
Electrostatics do have a large active area which reduces excursion and doppler distortion. Unfortunately from my search on the internet there is no consensus whether this doppler distortion is relevant enough or not.
I also read that speaker parameters fluctuate large with amplitude, caused by magnetic field, surround shifting parameters etc. Again, electrostatic speakers my circumvent this problems as they work quite different.
I also listened to Stax with a conventional transformer and to me the sound quality wass less so I bought the current direct drive one.
Regarding loudspeakers, could Doppler distortion be a factor?
Electrostatics do have a large active area which reduces excursion and doppler distortion. Unfortunately from my search on the internet there is no consensus whether this doppler distortion is relevant enough or not.
I also read that speaker parameters fluctuate large with amplitude, caused by magnetic field, surround shifting parameters etc. Again, electrostatic speakers my circumvent this problems as they work quite different.
I'm confused - you express that enclosure design is important, yet you're not in favor of CLD? Could you please explain?
Yes, measuring a speaker is rather like "measuring" a glass of wine. Measurements are useful for determining crossover points, adding filters, estimating box size, finding problems in the enclosure, whatnot... But once the speaker is "done" or nearly done, measurements will not tell you if the speaker will sound great. I imagine wine makers measure things also, but the taste at the end is what should* matter most. (*I say should, but really, it's marketing that's more important than taste, I bet)
Yup, but I believe there will be time when a new kind of measurement (or combination of measurements) be found to tell that a speaker will sound "great" (there will be time also when the definition of "great" will be streamlined). But not until those who knows (or will know) are willing to share the findings.
I don't know how the word "not" squeezed its way into that.
And to that I ask,
"why"?
It's truly a matter of what you're measuring. If your speaker doesn't sound great at the end of it all, then there's something you didn't measure. Either that, or your measurements are revealing something which is wrong with the speaker anyways.
As for preference, if one prefers a poor speaker, you're right, measurements don't do much bout that. If the goal is an accurate speaker then preference has little to do with it. Design the entire system (Room, Speaker, Listening Position, Amp) to measure well and you'll get a great speaker.
Are there still individual choices to be made in speaker design? Some. But that threshold of preference is unlike "taste in wine". Wine has no frame of reference. Audio's reference is difficult to say definitively, but there's still a "point" that we aim for which is tangible. If good measurements are not getting us very near that point, then something's clearly wrong in what you're measuring.
FWIW, I haven't heard a speaker that measures great and sounds bad, nor have a gear a speaker that measures bad and sounds great on more than a handful of albums. Also speakers that measure very similar tend to sound very similar. I'd say this is not likely coincidence b/c it happens time and time again on all sorts of source material. I'd actually say that measurements tell us more than we will likely hear and we already know that measurements correlate to preference. If your measurements bare no resemblance to what you are hearing, something is wrong.
Dan
Dan
Last edited: