What is the correlation b/w measurements and good sound?
- Thread starter caesar
- Start date
I can measure impulse response, phase response and frequency response and all can be "fixed" [I use DiracLive]. Once those are addressed and using a target curve that is known to satisfy, I am a happy guy. But if anything is obvious, measurements not withstanding, it is 98% about "preference".
Back in the days of John Dunlavy and when I was a dealer for him, I spent a lot of time in his factory in Colorado Springs. He measured EVERYTHING (phase response, step response, impulse response, frequency response, radiation patterns, etc). He had obtained a number of well known speakers that he was able to test in his anechoic chamber and for which he had the measurements in a book. One of the worst measuring speakers BUT one that I LOVED (and still do), were the Maggies.
It is all about "preference"!!!
Back in the days of John Dunlavy and when I was a dealer for him, I spent a lot of time in his factory in Colorado Springs. He measured EVERYTHING (phase response, step response, impulse response, frequency response, radiation patterns, etc). He had obtained a number of well known speakers that he was able to test in his anechoic chamber and for which he had the measurements in a book. One of the worst measuring speakers BUT one that I LOVED (and still do), were the Maggies.
It is all about "preference"!!!
Hi
Poking at a favorite hornet nest here... If you pull the tail of a tiger ....
The questions should be more precise about what kind of measurments are we referring to. Earl gEddes has demonstrated that we tolerate a level of THD that is far higher than what was previously thought .. THD of the order of 10% for example is deemed quite acceptable by many, in the bass we tolerate and come to accept far more. So the question should truly be the correlation between what are we measuring and our preferences not "measurements" as a whole.
It is sooo easy and convenient to dismiss measurements. This has become a favorite past time here on the WBF. Question : Does anyone actually believe that those who design and manufacturer your <insert favorite manufacturer> components do not measure them so that they can at least repeat the performances? And once they have a good set of what they and/or like make sure that these components meet those requirements .. err.. measure well within the parameters they have found to correlate with those desiderata? Our hobby is based on technology. Without getting too philosophical here; a mixture of the scientific and pragmatism. There are things we know how to measure and many we don't, it isn't because they aren't measurable. However much we haul rocks against the works of Harman luminaries, they did come up with an indisputable fact: Preferred speakers (in this example) share some common measurable traits...err. measurements... We don't know it all but we know at least that. This is a good place to start ...
Poking at a favorite hornet nest here... If you pull the tail of a tiger ....
The questions should be more precise about what kind of measurments are we referring to. Earl gEddes has demonstrated that we tolerate a level of THD that is far higher than what was previously thought .. THD of the order of 10% for example is deemed quite acceptable by many, in the bass we tolerate and come to accept far more. So the question should truly be the correlation between what are we measuring and our preferences not "measurements" as a whole.
It is sooo easy and convenient to dismiss measurements. This has become a favorite past time here on the WBF. Question : Does anyone actually believe that those who design and manufacturer your <insert favorite manufacturer> components do not measure them so that they can at least repeat the performances? And once they have a good set of what they and/or like make sure that these components meet those requirements .. err.. measure well within the parameters they have found to correlate with those desiderata? Our hobby is based on technology. Without getting too philosophical here; a mixture of the scientific and pragmatism. There are things we know how to measure and many we don't, it isn't because they aren't measurable. However much we haul rocks against the works of Harman luminaries, they did come up with an indisputable fact: Preferred speakers (in this example) share some common measurable traits...err. measurements... We don't know it all but we know at least that. This is a good place to start ...
I'm thinking more of us as individuals using measurements for our own use. In that context it makes perfect sense to listen, look and repeat until looking at the measurement and correlating that with the sound becomes reliable to a useful extent. I certainly am not at a point where I can just look and imagine what that sounds like. There's a lot of dissonance between the senses going on but it surely can be learned to a degree where it can be helpful. It's not all that different from using the instrumentation in our cars using say the speedometer and matrixing that reading to an oncoming curve then timing braking and turn in accordingly. You can always fly by the seat of your pants and feel your way through but if you're after lap time consistency the speedo and tach come in mighty handy. Still, I'm not saying it is an absolute necessity. If you happen to have a tool as simple as say a Phonic on a tripod, and a good memory it can help you keep track of the series of changes you've been experimenting with. The biggest benefit to me is that it helps me to "zero out the console" so to speak or in writing terms go back to the original draft. I don't enjoy finding I've in effect walked in a circle in a forest so I mark the waypoints.
Speaker manufacturers do microphone measurements. For electronics that's not necessary or accurate. Aside from THD, the measurements going into electronics are more to make sure the unit is perfectly stable and problem free. The topology could be win or lose when it's hooked up to speakers however. You can have two amps that on paper have the same measurements as far as you can tell, but subjectively one might be a total flop.
Not sure I understand that point ... You believe that in electronics one matches, match say tubes and or transistor by ear alone? Reistors? Caps? In electronics you measure and trim to reach stated, design goals ...That way more measurements than "aside THD" ...
Without going to an absurd example..If they sound different.. In some parameters they must measure differently.
What I'm getting at is no, the measurements for overall output don't tell you anything about how an amplifier will sound (competent designs). And also no, no one matches tubes and capacitors by ear, that doesn't make any sense. However selection is very much by ear for capacitors since they measure, within about all practical reasons, the same when they have the same values. You can measure differences between them in isolated situations designed for it, but those measurements are difference and not better or worse by any intelligible reason of measurement. As in you can't conclude which is better by looking at the minor and hard to accomplish measurement, unless you know something about what people say about the sound.
To reach design goals any part with the same basic specifications will work. That doesn't mean they'll sound remotely the same. You can't get a significant difference in measurement from using 1% metal film resistors and using Dale RN resistors, but you can sure as hell hear it. Also the design goals are calculated, not guessed and checked. I suppose you could call measuring DC offset a "measurement" but it can't tell you anything about the sound, it's just an indicator for trimming the circuit so you don't heat up your speakers. That's how it goes inside the electronics.
You could get a difference in sound by adjusting the bias of output transistors. But you wouldn't really measure it because you already know increasing it will lower 3rd harmonic distortion. The trouble is you're not sure if that sounds better or not. Some may prefer either way, and the amount of 3HD might be so minor you wouldn't think it's notable; and depending on design it may not be.
If designing electronics was as easy as setting measurement goals then we'd all have Samsung and Panasonic electronics. They've got plenty of engineers more than capable of doing that, but the reality is they may know nothing at all about what actually sounds good.
This isn't belief. I'm telling you from personal experience as well as reading on and conversing with designers of praised audio products.
If you're buying a $50k amplifier do you really want it to be based on simple measurements a monkey can accomplish, or would you prefer that it was designed by someone that is intimate with every part in it and was able to make a cohesive unit with knowledge that can't be had at the average electronics school but rather earned by truly caring about what it takes to accomplish high end sound?
To reach design goals any part with the same basic specifications will work. That doesn't mean they'll sound remotely the same. You can't get a significant difference in measurement from using 1% metal film resistors and using Dale RN resistors, but you can sure as hell hear it. Also the design goals are calculated, not guessed and checked. I suppose you could call measuring DC offset a "measurement" but it can't tell you anything about the sound, it's just an indicator for trimming the circuit so you don't heat up your speakers. That's how it goes inside the electronics.
You could get a difference in sound by adjusting the bias of output transistors. But you wouldn't really measure it because you already know increasing it will lower 3rd harmonic distortion. The trouble is you're not sure if that sounds better or not. Some may prefer either way, and the amount of 3HD might be so minor you wouldn't think it's notable; and depending on design it may not be.
If designing electronics was as easy as setting measurement goals then we'd all have Samsung and Panasonic electronics. They've got plenty of engineers more than capable of doing that, but the reality is they may know nothing at all about what actually sounds good.
This isn't belief. I'm telling you from personal experience as well as reading on and conversing with designers of praised audio products.
If you're buying a $50k amplifier do you really want it to be based on simple measurements a monkey can accomplish, or would you prefer that it was designed by someone that is intimate with every part in it and was able to make a cohesive unit with knowledge that can't be had at the average electronics school but rather earned by truly caring about what it takes to accomplish high end sound?
Qualifying : "Simple".. You made my points. Measuring is not an automatic act to be performed by , as you put it, "monkeys". It implies knowledge as opposed to simply knowing how to use an instrument. Now the point about what one learns (anywhere) and how it is applied, is not endemic to Audio designs nor price: Given the same set of notes some produce masterpieces and some ...
Phelonious, Got ya. But Bryston, for example, has great measurements and has great bass, while Sanders amp is an exercise in engineering and has lean bass. Ditto for Magico and YG... As an experienced marketing professional, let me throw out this thought at you: people like having the luxury of good sound in their life. At the same time, there is an unquenchable demand among consumers for good information to make solid decisions when spending their hard earned $$$$. Yet in this hobby, people are in the dark. Useful information is hard to come by. Most (not all!!!) reviewers are disgusting and just act as marketing departments for manufacturers and have created a Fukc the Audio Fan culture because they get very expensive audio gear as free loans, or buy at a great discount so they can later resell.
Audio fans are dying to get closer the knowledge of the experience with an audio product and everyone would love to know the true value of an audio product. After all, if you want a laptop, it's so easy to check the size, weight, cpu, memory, software, etc. , regardless if it says Dell, HP, Lenovo, or whatever. But the problem is that audio is an experience good, which can't be evaluated like that. So here come these companies and position themselves as "designing based on science" (as if everyone is not!). So people in a market segment for whom science-based decision making is an important criteria can then have a piece of mind and write a big check.
Yet, checking a typical measurements section in Stereophile, they publish 6 figures: impedance/ phase vs. frequency, spectral decay, crossover response, lateral response, vertical response, etc. From a quick novice glance, the measurements don't look very different from a "dirt cheap" speaker to one that costs as much as a house. (An acquaintance of mine has 7 children and actually lives in a home that costs less than a Magico q7 part 2.)
So can someone please explain to me, as clearly and patiently as you would explain to your doting grandfather, what these 6 stereophile measurements mean and whether they shed light on the what the speaker will sound like? And are there any other important measurements that Stereophile may be missing?
.... just throwing this out there ....
The greatest loss in realism between the live performance and listening to the stereo in your living room ....
.... is the RECORDING
The greatest loss in realism between the live performance and listening to the stereo in your living room ....
.... is the RECORDING
Caesar
Those are your opinions/impressions of the gears you mentioned. That is not the case for many including I. Please show us how close do you really think those measurements under the same conditions are, for the record we can hear (depending on frequencies) deviations as low as 1 dB ... and truthfully the set of measurements you alluded to are not complete. a trivial example: It is evident that the square footage of a room doesn't suffice to describe it... Even its height is not enough. Those however can give you some notion on the size of the room , the same way that looking at some graphs of FR or impedance can help you infer of the sound of a speaker. Now for those who are not DSP-adverse , we know that speakers even when EQ-ed similarly as to provide similar to close FR do not sound the same. Nothing startling: the FR is one of many parameters that would help in "describing" for the lack of a better word the sound of a speaker. We persist to want measurements of what make speakers sound different to ne impossible to perform . Repeating myself ad infinitum.. Do we know everything? no.. Do we know enough .... no! DO we have some places to start ? Absolutely... Difficult? Likely! Impossible ? NOPE!
IME ... nope !!!
I have a killer "bad" recording, Ike and Tina Turner, back in the good ol' days, some tracks recorded on a cheap tape unit in a club setting - on a normal system the glare and hotness of these recordings would drive a man insane,
. I picked it up in op shop for a dollar, the previous owner obviously couldn't take it ...But, an amazing transformation takes place when the system is in good enough shape - all screechiness falls away, and you're truly aware of Tina just powering on the vocals, and all the subtleties of the backing musicians' efforts fall into place - it has become like listening to a live performance, and one gets the full emotional charge from the experience ... these are the reasons I find it worthwhile pulling out all the stops on occasion, to get the best sound.
I would say that a simple rule applies to measurement and perceived sound quaility: the less optimised the complete system is, as a gestalt, the more conventional measurements count in terms of what one hears, one can easily "hear" the poorer measuring component compared to another; at the other end of this spectrum, relatively weakly measuring gear very carefully refined will deliver the high quality listening experience - you make ya choices ...
IME ... nope !!!
I have a killer "bad" recording, Ike and Tina Turner, back in the good ol' days, some tracks recorded on a cheap tape unit in a club setting - on a normal system the glare and hotness of these recordings would drive a man insane,
But, an amazing transformation takes place when the system is in good enough shape - all screechiness falls away, and you're truly aware of Tina just powering on the vocals, and all the subtleties of the backing musicians' efforts fall into place - it has become like listening to a live performance, and one gets the full emotional charge from the experience ... these are the reasons I find it worthwhile pulling out all the stops on occasion, to get the best sound.
Sorry fAs but I couldnt disagree more. A bad recording sounds bad on the best system. If you have a "screechy" recording as you put it, Ill interpret that as bright, what are you doing to your system to stop it from sounding bright?
Making it work better!
It will still be bright, by measurement, but the trick is to stop the brain focusing on that - which means that the musical message takes centre stage, this comes through clearly enough that any "deficiencies" in the sound are filtered out by the brain.
Once one has done this a few times, one realises what's going on ...
 | Steve Williams Site Founder | Site Owner | Administrator |  | Ron Resnick Site Co-Owner | Administrator |  | Julian (The Fixer) Website Build | Marketing Managersing |