Everything you mentioned in driver technology has been well explored and documented.
Planar drivers have their issues. Electrostatic speakers are merely another type of planar speaker. Serviced them for years along with all other types.
I saw Piezos when they were first introduced. The system isn't accurate or we would see them more often. If they were better, everyone would use them. Just because something is expensive to produce doesn't mean it is better either. We are using speaker types we have now because they perform well, are reliable and can be manufactured with small variations in performance.
In the 1970s they experimented and developed all kinds of different drivers. Every flaky idea you can think of was explored, some developed (and failed). We saw it all. Speaker technology is an old study. From the 1920's (maybe earlier), we have improved just about every speaker technology you can think of. A breakthrough may be made, I don't know what the future holds. But right now with what we have, the dynamic speaker everyone is familiar with is the best technology, and it is reliable.
Always remember too. Execution means more very often than theoretical performance. Often theory points to something not possible to manufacture so it works or will last.
Planar drivers have their issues. Electrostatic speakers are merely another type of planar speaker. Serviced them for years along with all other types.
I saw Piezos when they were first introduced. The system isn't accurate or we would see them more often. If they were better, everyone would use them. Just because something is expensive to produce doesn't mean it is better either. We are using speaker types we have now because they perform well, are reliable and can be manufactured with small variations in performance.
In the 1970s they experimented and developed all kinds of different drivers. Every flaky idea you can think of was explored, some developed (and failed). We saw it all. Speaker technology is an old study. From the 1920's (maybe earlier), we have improved just about every speaker technology you can think of. A breakthrough may be made, I don't know what the future holds. But right now with what we have, the dynamic speaker everyone is familiar with is the best technology, and it is reliable.
Always remember too. Execution means more very often than theoretical performance. Often theory points to something not possible to manufacture so it works or will last.
Thanks, but that is too deep for me. I do not have the skills or knowledge for that sort of thing.
For me, a DIY amp is a board with an integrated power supply, that I have to install in a case along with connectors and switches. something like a hypex ncore board. The idea of making/modifying a PCB is about as appealing as making my own plywood by glueing together veneers. Actually less so, since I actually could make my own plywood if I had to...
Thanks for the info. I'm working on a one-off prototype build that will get a lot of testing and mods done to it. Depending on how that sounds and goes, I may look at getting it out there, somehow, for other people to enjoy as well.
There's nothing like going "full commercial" to ruin a sentimental pursuit, but if there's enough interest I'll definitely look at some options like kits, group buys, parts BOMs, if not a full product if it goes that far.
There's nothing like going "full commercial" to ruin a sentimental pursuit, but if there's enough interest I'll definitely look at some options like kits, group buys, parts BOMs, if not a full product if it goes that far.
Hi abstract,
Yes, current drive amplifiers exist for audio and industrial uses. Mostly industrial for special use cases. All they are in a basic sense is a voltage to current converter that scales up the current. Just as a normal amplifier starts and ends with a voltage, scaled up in the output.
I wonder. At the load resonance, do you run out of voltage compliance at higher powers?
Yes, current drive amplifiers exist for audio and industrial uses. Mostly industrial for special use cases. All they are in a basic sense is a voltage to current converter that scales up the current. Just as a normal amplifier starts and ends with a voltage, scaled up in the output.
I wonder. At the load resonance, do you run out of voltage compliance at higher powers?
The design I'm looking at is basically voltage drive in the bass, with minor bumps in the frequency response, depending on how it's tuned. So, if I want the damping factor to start its downward slope at 200Hz instead of 500Hz or 1kHz, that will loosen the bass. Then it's a sort-of holistic matter of assessing where the system sounds best.
Actually, I'm not sure. You have an impedance peak and rise in efficiency as a result. So this may set your maximum power output as you run out of supply voltage. How far down the level is at that peak is the question in keeping frequency response flat.
Where those things all end up, I don't know. The same load on a voltage amp and current amp pose opposite problems. Interesting.
Where those things all end up, I don't know. The same load on a voltage amp and current amp pose opposite problems. Interesting.
@anatech If the ranking of drivers is derived purely from direct scientific comparison of a speaker to a reference signal, then using a percentage value to quantify the comparison, you have a valid point.
If the human ear is used at any point and time of the rating or comparison process between the drivers, it becomes a matter of personal preference to some degree, as small of a factor or percentage as that influence may be.
This all implies the difference of listening with the human ear or purely with microphones. In the case of qualifying the acoustic performance of the speaker in question, the only objectively accurate way of quantifying the measurements are by means of a direct 1:1 data comparison. The driver with the closest output compared to the reference sound registered by calibrated microphone is in theory the most accurate.
Being that we humans listen with our ears means we rely solely on our ears to quantify and compare the "perceived" accuracy of the DUT (speaker in question). The measurement test setup can't listen for us, unless you trust the test setup to produce the same results which you'll have to accept as "truthful". Accurate sound to one person isn't going to be accurate to another, being we humans hear slightly differently from one person to another. We hear music differently at varying volume levels. When someone plays back something which sounds accurate to them, its almost guaranteed to not sound accurate to you. Who do you believe is being truthful here?
You can only rely on your own ears unless the other person has similar hearing and subliminal adaptive tastes which can greatly skew their hearing.
The loudness curve also greatly affects this. Some distortion types, combinations and spectral distribution of these can sound different from an accuracy POV. The lack of distortion can even sometimes be perceived as not sounding as true or lifelike. Hearing something and describing it as accurate is a complicated process for a human being. Nonetheless we hear what we do and you can be guaranteed the next person won't hear it the same way.
My point here is you have to involve a human being to rate how "good" or perceivably "accurate" a speaker SOUNDS. You can't use a microphone and test equipment for that, unless you're collecting raw data, which is only meaningful to a computer or other type of data anlysis. It doesn't factor in the psychology or preferences of the human being listening to the sound coming out of the speaker. The word "accurate" only applies to an absolute comparison made by test equipment. If a human uses the word "accurate", he or she can only refer to it using their own ears to another reference they remember hearing sometime in their past. This is a highly flawed and uncertain process. Only the direct referencing of measurments of a speaker made by / analyzed by / quantified by a calibrated test setup can be deemed as being scientifically accurate by a defined percentage.
I'm not trying to twist words here. Its just a matter of hearing with and trusting your ears since you don't hear with the same accuracy as test equipment can, which can't be referred to as perceived sound. The two can differ greatly.
If the human ear is used at any point and time of the rating or comparison process between the drivers, it becomes a matter of personal preference to some degree, as small of a factor or percentage as that influence may be.
This all implies the difference of listening with the human ear or purely with microphones. In the case of qualifying the acoustic performance of the speaker in question, the only objectively accurate way of quantifying the measurements are by means of a direct 1:1 data comparison. The driver with the closest output compared to the reference sound registered by calibrated microphone is in theory the most accurate.
Being that we humans listen with our ears means we rely solely on our ears to quantify and compare the "perceived" accuracy of the DUT (speaker in question). The measurement test setup can't listen for us, unless you trust the test setup to produce the same results which you'll have to accept as "truthful". Accurate sound to one person isn't going to be accurate to another, being we humans hear slightly differently from one person to another. We hear music differently at varying volume levels. When someone plays back something which sounds accurate to them, its almost guaranteed to not sound accurate to you. Who do you believe is being truthful here?
You can only rely on your own ears unless the other person has similar hearing and subliminal adaptive tastes which can greatly skew their hearing.
The loudness curve also greatly affects this. Some distortion types, combinations and spectral distribution of these can sound different from an accuracy POV. The lack of distortion can even sometimes be perceived as not sounding as true or lifelike. Hearing something and describing it as accurate is a complicated process for a human being. Nonetheless we hear what we do and you can be guaranteed the next person won't hear it the same way.
My point here is you have to involve a human being to rate how "good" or perceivably "accurate" a speaker SOUNDS. You can't use a microphone and test equipment for that, unless you're collecting raw data, which is only meaningful to a computer or other type of data anlysis. It doesn't factor in the psychology or preferences of the human being listening to the sound coming out of the speaker. The word "accurate" only applies to an absolute comparison made by test equipment. If a human uses the word "accurate", he or she can only refer to it using their own ears to another reference they remember hearing sometime in their past. This is a highly flawed and uncertain process. Only the direct referencing of measurments of a speaker made by / analyzed by / quantified by a calibrated test setup can be deemed as being scientifically accurate by a defined percentage.
I'm not trying to twist words here. Its just a matter of hearing with and trusting your ears since you don't hear with the same accuracy as test equipment can, which can't be referred to as perceived sound. The two can differ greatly.
Hi profiguy,
Actually ... you can. The very best speakers are designed with test instrumentation. They do listening tests of course.
Listen to what I said carefully. If you recreate the original sound field accurately, it will sound the same to whoever is listening no matter how they hear things. Period. Now if you don't like reality, or prefer a colouration in sound, that's cool. Cerwin Vega was designed that way. All the top speakers are designed to be accurate. PSB used the NRC, JBL hired Floyd toole and design their speakers the same way. Klipsch does too. I think anyone who wants to even pretend to make a good product will use instrumentation heavily.
Now, if you want to be an audio luminary who "designs by ear", you'll have a great story, expensive product and make something that isn't that accurate. That is the way things work these days.
Actually ... you can. The very best speakers are designed with test instrumentation. They do listening tests of course.
Listen to what I said carefully. If you recreate the original sound field accurately, it will sound the same to whoever is listening no matter how they hear things. Period. Now if you don't like reality, or prefer a colouration in sound, that's cool. Cerwin Vega was designed that way. All the top speakers are designed to be accurate. PSB used the NRC, JBL hired Floyd toole and design their speakers the same way. Klipsch does too. I think anyone who wants to even pretend to make a good product will use instrumentation heavily.
Now, if you want to be an audio luminary who "designs by ear", you'll have a great story, expensive product and make something that isn't that accurate. That is the way things work these days.
IIRC EQ is mandatory with current drive. Or motional feedback. Has been done before. In the 80’s KEF experimented with it, convinced that it -at least in theory- was superior to voltage drive. Nowadays there is a tendency to refrain to mid and high range for current drive if I’m correct. Makes sense.
Can you define the above for me please? I’ve looked and looked in my audio engineering and mastering text books and can’t seem to find a reference?
Hi mayhem13,
What is so difficult to understand? I'm talking to normal humans, not engineers.
Let me put it this way. If you, by whatever method, cause the air molecules to move the same way they did during an original acoustic event near the ears of the observer, you have accurately recreated that acoustic event. Period. It will cause exactly the same pressure variations in a hearing mechanism that the original event did. Therefore it matters not what impairments the observer may have, that observer will experience the same stimulus as they originally did.
I was separating the emotional state of the observer, as was suggested by profiguy, as the reason why measurements were ineffective. The emotional state of the observer is their problem and it is completely pointless to attempt to control that.
Now, if people want to argue about which measurements are the most effective for acoustic reproduction - have at it. But using what we do know is vastly superior to trying to do it by ear. With the application of testing and quantifying (measuring), our acoustic reproducers have quickly increased in quality. We identified problems and were able to address some.
Over my lifetime, absolutely every piece of audio equipment has been improved. The electronics improved by leaps and bounds, as did loudspeakers, headphones and phono cartridges. Even tape heads. The enabler for improvements is basically improved understanding of the process. This comes about through quantifying performance through measurements. Our measuring equipment has improved and continues to do so. That allows us to better understand - everything, including audio. One thing that has been a constant is equipment designed mostly (or only) by ear tends to be the least reliable and poorly performing. Sorry folks, we are not test instruments, and we do not have consistent performance.
Where would we be if our rulers and tape measures varied in value and linearity day by day or hour by hour? Try and build anything, then convey that information to another. This is what the design by ear crowd is advocating. Some days walking a mile or km seems more difficult than others. See my point?
What is so difficult to understand? I'm talking to normal humans, not engineers.
Let me put it this way. If you, by whatever method, cause the air molecules to move the same way they did during an original acoustic event near the ears of the observer, you have accurately recreated that acoustic event. Period. It will cause exactly the same pressure variations in a hearing mechanism that the original event did. Therefore it matters not what impairments the observer may have, that observer will experience the same stimulus as they originally did.
I was separating the emotional state of the observer, as was suggested by profiguy, as the reason why measurements were ineffective. The emotional state of the observer is their problem and it is completely pointless to attempt to control that.
Now, if people want to argue about which measurements are the most effective for acoustic reproduction - have at it. But using what we do know is vastly superior to trying to do it by ear. With the application of testing and quantifying (measuring), our acoustic reproducers have quickly increased in quality. We identified problems and were able to address some.
Over my lifetime, absolutely every piece of audio equipment has been improved. The electronics improved by leaps and bounds, as did loudspeakers, headphones and phono cartridges. Even tape heads. The enabler for improvements is basically improved understanding of the process. This comes about through quantifying performance through measurements. Our measuring equipment has improved and continues to do so. That allows us to better understand - everything, including audio. One thing that has been a constant is equipment designed mostly (or only) by ear tends to be the least reliable and poorly performing. Sorry folks, we are not test instruments, and we do not have consistent performance.
Where would we be if our rulers and tape measures varied in value and linearity day by day or hour by hour? Try and build anything, then convey that information to another. This is what the design by ear crowd is advocating. Some days walking a mile or km seems more difficult than others. See my point?
Sorry….maybe my question didn’t translate?
So let’s consider the recording, mixing and mastering process a bit first as we’re trying to replicate the ‘original sound field’. Once we’ve placed a microphone in front of any source, you’re reference is already nullified…..and on and on down the signal chain from panning, EQ, compression, reverb and other time based processing?…….i think my point is clear In that there is no point of reference.
I see and read countless posts here on horn/waveguide geometry to get the ideal response……..but the most important waveguide is the one hanging on the side of our heads…..our outer ear……and no two are alike……another point of reference eliminated. Add in the countless other physiological differences that define our individuality?
Now the acoustic space we‘re listening in…..and given the forum I think we can eliminate headphones, I don’t care how you’re DI and power response/spinorama data looks on paper…….every space is different in the point of early and first reflections as well as verb and decay.…..add in the noise floor?………another point of reference erased.
Wanna extol the virtues of a near flat frequency response and low odd order distortion….i‘m all in as suitable points of reference……but everything else?…..purely subjective and personal.
So let’s consider the recording, mixing and mastering process a bit first as we’re trying to replicate the ‘original sound field’. Once we’ve placed a microphone in front of any source, you’re reference is already nullified…..and on and on down the signal chain from panning, EQ, compression, reverb and other time based processing?…….i think my point is clear In that there is no point of reference.
I see and read countless posts here on horn/waveguide geometry to get the ideal response……..but the most important waveguide is the one hanging on the side of our heads…..our outer ear……and no two are alike……another point of reference eliminated. Add in the countless other physiological differences that define our individuality?
Now the acoustic space we‘re listening in…..and given the forum I think we can eliminate headphones, I don’t care how you’re DI and power response/spinorama data looks on paper…….every space is different in the point of early and first reflections as well as verb and decay.…..add in the noise floor?………another point of reference erased.
Wanna extol the virtues of a near flat frequency response and low odd order distortion….i‘m all in as suitable points of reference……but everything else?…..purely subjective and personal.
Last edited:
Hi mayhem13,
I see where you're coming from. However, since I've worked in recording studios there may be some misunderstanding as to what their goal is.
If you want to capture a raw performance, good luck. You can't capture the full dynamic range. But if that is your goal, you are not using a recording studio.
A recording studio produces a product. They employ effects units to modify the original sound, they balance the instruments and vocals - and they re-record over errors made. They are producing a product for the public to enjoy. That's why live performances sound nothing like "the album".
The goal of a reproduction system (your stereo) is to faithfully (hopefully) reproduce the media you are playing in it's intended form. You may not agree with the band, producer and engineer, but that is the intent. Feel free to modify that piece to your liking, but you must admit you have deviated from the intended product. If you are insanely lucky enough to have a well recorded live recording that doesn't clip, your system should equally reproduce that without deviation unless you intentionally changed something.
So let's get the intended product clear before we waste time arguing in circles. I've heard live often enough, generally speaking I'll take the produced version for day to day listening. Love the live event, but I'm certainly not always up for that. If you want to argue the studio produced version deviates from the live event - you're right. However this has exactly zero to do with what your audio reproduction system is supposed to do.
I see where you're coming from. However, since I've worked in recording studios there may be some misunderstanding as to what their goal is.
If you want to capture a raw performance, good luck. You can't capture the full dynamic range. But if that is your goal, you are not using a recording studio.
A recording studio produces a product. They employ effects units to modify the original sound, they balance the instruments and vocals - and they re-record over errors made. They are producing a product for the public to enjoy. That's why live performances sound nothing like "the album".
The goal of a reproduction system (your stereo) is to faithfully (hopefully) reproduce the media you are playing in it's intended form. You may not agree with the band, producer and engineer, but that is the intent. Feel free to modify that piece to your liking, but you must admit you have deviated from the intended product. If you are insanely lucky enough to have a well recorded live recording that doesn't clip, your system should equally reproduce that without deviation unless you intentionally changed something.
So let's get the intended product clear before we waste time arguing in circles. I've heard live often enough, generally speaking I'll take the produced version for day to day listening. Love the live event, but I'm certainly not always up for that. If you want to argue the studio produced version deviates from the live event - you're right. However this has exactly zero to do with what your audio reproduction system is supposed to do.
You can't actually reproduce the original sound field, I'm not certain that was the intended point.. but if that's what you're saying, I agree. Two pressure microphones/speakers is insufficient to define the space. However it's partially captured on the recording.. as long as the additional stereo format is good enough for your needs.. and it can be.
An image is then projected in your room according to the stereo format, not the original field. You don't necessarily want to mix up your room with the original, but rather to keep it clean. The goals of reproduction are completely different to the goals of production.
@AllenB Actually nailed what I'm trying to say.
@anatech As I said before, IN THEORY what you're suggesting is somewhat valid and I'm not disputing the need for measurement.
The big problem with your theory is you can't count on the reference being an actual reference having nothing to verify it with. You can record the "sound field" you're referring to, but its virtually impossible to recreate using loudspeakers... period. They will never duplicate the event perfectly because of their inherent flaws and inability to radiate the acoustical event the same way. Because of this issue, the next best thing to use as verification (regardless of how flawed it may be) is the human ear. Why? Because it's ultimately the only sensing instrument you have left and its the actual pathway you're using to translate what you're doing as a human being. Its the ear which you need to please, satisfy or simply confer with. You can measure and calculate all you like. The buck stops at the ears.
So yes, IN THEORY you're argument is possible and I actually 100% agree you need to measure things to see where you're at in the design process. I never ever disputed this idea. My issue is the fact you can't verify your reference to duplicate it. No calibrated microphone is capable of hearing the event the exact same way as any given person's ears and no speaker is capable of perfectly duplicating the "reference".
Ultimately what's the point of doing everything you are if it doesn't sound good? Isn't that in reality what we're all trying to accomplish? Focusing exclusively on measurments and claiming it will automatically sound good isn't valid. Some people obsess about measurements and theory, but that doesn't mean they can build a good sounding speaker using that methodology. Its like those guys in Japan who thought they could design the perfect amplifier by reducing distortion to the lowest possible amount and making other specs as perfect as they could on paper. In the end, their creation sounded worse than the amplifier they were trying to improve on.
@anatech As I said before, IN THEORY what you're suggesting is somewhat valid and I'm not disputing the need for measurement.
The big problem with your theory is you can't count on the reference being an actual reference having nothing to verify it with. You can record the "sound field" you're referring to, but its virtually impossible to recreate using loudspeakers... period. They will never duplicate the event perfectly because of their inherent flaws and inability to radiate the acoustical event the same way. Because of this issue, the next best thing to use as verification (regardless of how flawed it may be) is the human ear. Why? Because it's ultimately the only sensing instrument you have left and its the actual pathway you're using to translate what you're doing as a human being. Its the ear which you need to please, satisfy or simply confer with. You can measure and calculate all you like. The buck stops at the ears.
So yes, IN THEORY you're argument is possible and I actually 100% agree you need to measure things to see where you're at in the design process. I never ever disputed this idea. My issue is the fact you can't verify your reference to duplicate it. No calibrated microphone is capable of hearing the event the exact same way as any given person's ears and no speaker is capable of perfectly duplicating the "reference".
Ultimately what's the point of doing everything you are if it doesn't sound good? Isn't that in reality what we're all trying to accomplish? Focusing exclusively on measurments and claiming it will automatically sound good isn't valid. Some people obsess about measurements and theory, but that doesn't mean they can build a good sounding speaker using that methodology. Its like those guys in Japan who thought they could design the perfect amplifier by reducing distortion to the lowest possible amount and making other specs as perfect as they could on paper. In the end, their creation sounded worse than the amplifier they were trying to improve on.
Last edited:
Hi profiguy,
I'm talking about a goal that isn't a moving target.
This is what you posted ...
Now my personal experience over multiple decades of experience beginning in the earlier 1970s in audio sales, running through service, live concerts and studio experience has pointed to one thing. The vast majority of people prefer the more accurate reproduction of music. If nothing is good, they break up into groups that avoid shortcomings they can't deal with. But give them good options and most head that way. These better systems have always measured better and better. A steady progression forward. We didn't get there by throwing up our hands saying "everybody likes something different". We got there by making the acoustic more and more like the original product or performance. One term used might be, more lifelike.
This is all I have said. But don't even try to drag psychology or human thought processes into the mix. That is beyond control and varies greatly throughout the day for even the same observer. Never mind different people. Same for hearing impairments. The affect the original sound and reproduced sound exactly the same way, so it drops out of the mix too.
I'm talking about a goal that isn't a moving target.
This is what you posted ...
You dragged a massive variable into the discussion. One that cancels itself out anyway.
Now my personal experience over multiple decades of experience beginning in the earlier 1970s in audio sales, running through service, live concerts and studio experience has pointed to one thing. The vast majority of people prefer the more accurate reproduction of music. If nothing is good, they break up into groups that avoid shortcomings they can't deal with. But give them good options and most head that way. These better systems have always measured better and better. A steady progression forward. We didn't get there by throwing up our hands saying "everybody likes something different". We got there by making the acoustic more and more like the original product or performance. One term used might be, more lifelike.
This is all I have said. But don't even try to drag psychology or human thought processes into the mix. That is beyond control and varies greatly throughout the day for even the same observer. Never mind different people. Same for hearing impairments. The affect the original sound and reproduced sound exactly the same way, so it drops out of the mix too.