time domain etc....
Hello Lynn
Unfortunately, when someone does not share the same convictions of others in a forum, often follows are personal attacks . It cannot be forgotten, that audio is something to give joy and pleasure , so talking about it should never come down to a level to insult each other because of different thinking, evolved stage and knowledge.
It’s interesting while I accostume to read audio forums regularly, I never found someone giving attention to time domain. Giving a quick search at google, I found some interesting informations at this site :
http://www.eclipse-td.com/e12_technology/technology1.html
The time domain theory aims to reproduce the original audio waveform, which contains the amplitude / phase frequency characteristics, through the speaker, without change. This theory is therefore considered to be an effective approach towards the reproduction of the truly original sound.
As shown in [Figure 3], the resonant component (undesired harmonics), which extends the tail end of the audio waveform, indicates the presence of undesired harmonics that are unrelated to the original signal. The purpose behind the use of time domain theory is to bring the impulse response as close as possible to the original impulse itself, eliminating any resonant components or undesirable
The impulse response can be used as an indicator in attempting to improve time characteristics. An impulse is a signal which, in theory, possesses the following characteristics: magnitude (amplitude) = infinite; time duration = infinitesimal; and an area equal to 1. An impulse is composed of a mixture of all frequencies (amplitudes and phases). Therefore, if an impulse can be accurately reproduced, the time characteristics of the input signal will be output exactly, without change. In other words, Time Domain theory is put into practice in order to provide an impulse response that is more closely matched to the actual impulse signal input. [Figure 2]
Sound reproduction becomes more accurate the closer the waveforms of the input signal (impulse) and output signal (impulse response) resemble each other.
I don’t know better way to get more close matched impulse response, than with horns. That’s why they sound “faster” than conventional speakers.
do some travelling and audition famous horn systems.
This is something I like to do. I have listened to some really weard full horn loaded systems , like in a Tokyo suburb, for example :
http://www.flickr.com/photos/audiovoice_hornloudspeakers/page8/
and this year the magico ultimos, and a small Goto horn system in italy.
These still remain amongst the most impressive playback demo’s I heard. Conventional speakers just don’t do it similarly.
One of the most accurate and natural sounding speakers btw. I heard was the magico ultimo in the frequency range from 300hz up.
Might you want to comment what our friend has to say :
Is always said that the people who spent some Time with to upper bass open baffles develop to permanent (or long lasting) corruption of taste to reproduced music. The open baffles AT to upper bass, not to mater how large they plows and what king drivers were used, produces completely different and completely artificial bass and this bass for is VERY devastating long-term listening experience. Ace the result the open baffles bass people get converted in a sort of cult where to even talk with them about sound become virtually imposable.
Rds Angelo
Hello Lynn
Unfortunately, when someone does not share the same convictions of others in a forum, often follows are personal attacks . It cannot be forgotten, that audio is something to give joy and pleasure , so talking about it should never come down to a level to insult each other because of different thinking, evolved stage and knowledge.
It’s interesting while I accostume to read audio forums regularly, I never found someone giving attention to time domain. Giving a quick search at google, I found some interesting informations at this site :
http://www.eclipse-td.com/e12_technology/technology1.html
The time domain theory aims to reproduce the original audio waveform, which contains the amplitude / phase frequency characteristics, through the speaker, without change. This theory is therefore considered to be an effective approach towards the reproduction of the truly original sound.
As shown in [Figure 3], the resonant component (undesired harmonics), which extends the tail end of the audio waveform, indicates the presence of undesired harmonics that are unrelated to the original signal. The purpose behind the use of time domain theory is to bring the impulse response as close as possible to the original impulse itself, eliminating any resonant components or undesirable
The impulse response can be used as an indicator in attempting to improve time characteristics. An impulse is a signal which, in theory, possesses the following characteristics: magnitude (amplitude) = infinite; time duration = infinitesimal; and an area equal to 1. An impulse is composed of a mixture of all frequencies (amplitudes and phases). Therefore, if an impulse can be accurately reproduced, the time characteristics of the input signal will be output exactly, without change. In other words, Time Domain theory is put into practice in order to provide an impulse response that is more closely matched to the actual impulse signal input. [Figure 2]
Sound reproduction becomes more accurate the closer the waveforms of the input signal (impulse) and output signal (impulse response) resemble each other.
I don’t know better way to get more close matched impulse response, than with horns. That’s why they sound “faster” than conventional speakers.
do some travelling and audition famous horn systems.
This is something I like to do. I have listened to some really weard full horn loaded systems , like in a Tokyo suburb, for example :
http://www.flickr.com/photos/audiovoice_hornloudspeakers/page8/
and this year the magico ultimos, and a small Goto horn system in italy.
These still remain amongst the most impressive playback demo’s I heard. Conventional speakers just don’t do it similarly.
One of the most accurate and natural sounding speakers btw. I heard was the magico ultimo in the frequency range from 300hz up.
Might you want to comment what our friend has to say :
Is always said that the people who spent some Time with to upper bass open baffles develop to permanent (or long lasting) corruption of taste to reproduced music. The open baffles AT to upper bass, not to mater how large they plows and what king drivers were used, produces completely different and completely artificial bass and this bass for is VERY devastating long-term listening experience. Ace the result the open baffles bass people get converted in a sort of cult where to even talk with them about sound become virtually imposable.
Rds Angelo
After reading your article, I think the concept is valid, I actually got inspired to think of a simple implementation of impulse editing that can be done automatically, but think it would be much better that methods were discussed among people involved in development of software, which I think very few are in this forum.john k... said:
Hi Angelo, yes, I focus on the time domain, but I'm not in the linear-phase camp! That makes it hard to write marketing literature (and I've done my share of that, trust me) but this has been my experience.
I look at the time domain as having two parts: the initial attack, and the decay phase. To a very minor degree, the attack is set by the tweeter, but it is mostly a function of whether or not the crossover is all-pass or not. Unfortunately, unless digital DSP is used, there are very severe IM distortion tradeoffs to a linear-phase crossover - you are confronted with an ugly choice between high IM distortion & pretty square waves OR substantially lower IM distortion and weird square waves. To get both you must have digital correction, not something that's attractive to people who like analog and triodes (like me).
The decay phase is slightly intermingled with the phase distortion of the crossover, but not very much. It is caused by completely different mechanisms - multiple resonances in the drivers, standing waves in the cabinets (all cabinets have these, cabinet damping is only partially effective, and diaphragms are acoustically transparent), baffle diffraction, internal reflections in the driver, unwanted radiation from the spider, dustcap resonances, etc. etc. etc. A whole lotta stuff.
Perceptually, the decay phase is quite important because that's when pitch and harmonic recognition come into play. You can't discern pitch from a click, which is all the attack phase really is. Listen to any waveform that's a millisecond or less and that's all you'll hear - it doesn't sound like much.
I'm mostly influenced by the BBC school, which great weight on decay performance of drivers and cabinets. I'm not influenced by the Altec and JBL schools at all, and only to a minor degree by the linear-phasers. I've flipped the switch many times on the digital fixer-upper boxes, and only heard a slight change, and nothing that I'd call a major improvement. It's surprising, really, how little improvement digital time correction offers, considering the dramatic change in the waveform. This is a test, by the way, that I think everyone that's fooling around with loudspeakers owes it to themselves to try.
I look at the time domain as having two parts: the initial attack, and the decay phase. To a very minor degree, the attack is set by the tweeter, but it is mostly a function of whether or not the crossover is all-pass or not. Unfortunately, unless digital DSP is used, there are very severe IM distortion tradeoffs to a linear-phase crossover - you are confronted with an ugly choice between high IM distortion & pretty square waves OR substantially lower IM distortion and weird square waves. To get both you must have digital correction, not something that's attractive to people who like analog and triodes (like me).
The decay phase is slightly intermingled with the phase distortion of the crossover, but not very much. It is caused by completely different mechanisms - multiple resonances in the drivers, standing waves in the cabinets (all cabinets have these, cabinet damping is only partially effective, and diaphragms are acoustically transparent), baffle diffraction, internal reflections in the driver, unwanted radiation from the spider, dustcap resonances, etc. etc. etc. A whole lotta stuff.
Perceptually, the decay phase is quite important because that's when pitch and harmonic recognition come into play. You can't discern pitch from a click, which is all the attack phase really is. Listen to any waveform that's a millisecond or less and that's all you'll hear - it doesn't sound like much.
I'm mostly influenced by the BBC school, which great weight on decay performance of drivers and cabinets. I'm not influenced by the Altec and JBL schools at all, and only to a minor degree by the linear-phasers. I've flipped the switch many times on the digital fixer-upper boxes, and only heard a slight change, and nothing that I'd call a major improvement. It's surprising, really, how little improvement digital time correction offers, considering the dramatic change in the waveform. This is a test, by the way, that I think everyone that's fooling around with loudspeakers owes it to themselves to try.
I should add that training plays a role as well. Here I'll describe a pair of simple ear-training methods to detect colorations - you won't thank me, since you'll be sensitized for life once you become aware of them.
1) Tweeters have a surprising amount of distortion. To hear the tweeter distortion, play music at a level you're comfortable with, and choose selections that have a continuous line of singing. Now shut off the midbass and woofers, leaving the tweeter playing. The sound will be wispy and insubstantial, and after you listen for a little while, you'll notice that on some passages, it is actually quite distorted. Notice that the distortion comes and goes in an unpredictable and maddening way - just as you start to relax, the thing blats and sizzles like a piece of tissue paper, than the annoyance flits away again.
Now that you can recognize what your tweeter's been doing all along, restore the mids and bass. You can still hear the problem, can't you? Now that awareness will never go away.
2) Cabinets have more coloration than you think. Play a symphonic piece this time, something scored with a lot of density. No chamber music or jazz, no good for this test. Play it a moderate level, and walk up to the speaker and press your ear directly against the side of the cabinet. Repeat for the rear. Try the other side.
Now sit and listen in the usual listening position. You can still hear the cabinet sound, can't you? Yes, it's been there all along, and you didn't know it. Now you do - and you won't be able to ignore it any more!
You've had a little taste of the speaker-designer world, where speaker defects jump out at you at trade shows and the homes of hifi enthusiasts. I've described only two of the many, many defects of loudspeakers.
There are lots more - the hard part is choosing which ones to minimize, and the hardest part of all, which virtues do you want to emphasize, at the expense of certain colorations? What do you want, and what can you ignore? This is where you must know your own tastes, and firmly set aside the "advice" of the magazines and internet "experts".
You've seen how I blow off the advice of others with a joke intended to deflect the issue - well, I know my own tastes, and just how little it overlaps with most audiophiles. By now I've come to terms with that, and accept insults that accuse me of either being a complete technical ignoramus or a tasteless fool. The insults, in the long term, really say more about the person saying them in the first place.
1) Tweeters have a surprising amount of distortion. To hear the tweeter distortion, play music at a level you're comfortable with, and choose selections that have a continuous line of singing. Now shut off the midbass and woofers, leaving the tweeter playing. The sound will be wispy and insubstantial, and after you listen for a little while, you'll notice that on some passages, it is actually quite distorted. Notice that the distortion comes and goes in an unpredictable and maddening way - just as you start to relax, the thing blats and sizzles like a piece of tissue paper, than the annoyance flits away again.
Now that you can recognize what your tweeter's been doing all along, restore the mids and bass. You can still hear the problem, can't you? Now that awareness will never go away.
2) Cabinets have more coloration than you think. Play a symphonic piece this time, something scored with a lot of density. No chamber music or jazz, no good for this test. Play it a moderate level, and walk up to the speaker and press your ear directly against the side of the cabinet. Repeat for the rear. Try the other side.
Now sit and listen in the usual listening position. You can still hear the cabinet sound, can't you? Yes, it's been there all along, and you didn't know it. Now you do - and you won't be able to ignore it any more!
You've had a little taste of the speaker-designer world, where speaker defects jump out at you at trade shows and the homes of hifi enthusiasts. I've described only two of the many, many defects of loudspeakers.
There are lots more - the hard part is choosing which ones to minimize, and the hardest part of all, which virtues do you want to emphasize, at the expense of certain colorations? What do you want, and what can you ignore? This is where you must know your own tastes, and firmly set aside the "advice" of the magazines and internet "experts".
You've seen how I blow off the advice of others with a joke intended to deflect the issue - well, I know my own tastes, and just how little it overlaps with most audiophiles. By now I've come to terms with that, and accept insults that accuse me of either being a complete technical ignoramus or a tasteless fool. The insults, in the long term, really say more about the person saying them in the first place.
Sure, why not. This one tells you more about speaker directivity than any of the graphs or theories about the desired shape of polar pattern.
3) Play a symphonic piece at a moderate level with a fairly continuous line - choose a piece you like, this test won't destroy your faith as much as the first two. Shut down the other channel - you can use a mono source if you want, just choose something with a pleasing and natural-sounding tonal balance.
Walk up to the remaining speaker so you're about a meter away, then walk around it slowly, completing one circle, then keep moving slowly round and round. Stop and listen what the speaker sounds like in the rear quadrant. Most monopoles (nearly all) sound - well - "gurgly" and bass-heavy from the rear. The "gurgly" quality are cabinet colorations at close range (you are close enough so room colorations don't matter) and the dull, muffled quality is the strange, heavily diffracted sound of the tweeter.
Note the angle where the sound quality suddenly changes - and with most square-sided cabinets, it will - and pay attention to the different "personalities" the speaker has at different angles. You may notice certain angles are "critical" angles, like 45 degrees and multiples of that - listen for phantom tweeters seeming to come from cabinet edges. Listen overhead, too - that will probably be different as well. Move around the speaker and see how many different personalities it has, and what they sound like.
Try this with dipoles, and if you can get access to them, quasi-cardioid speakers (dipoles with absorption of the rear-wave). They have a completely different set of personalities, don't they? The dipoles are different than the quasi-cardioids - in fact, if you're "tuning" a quasi-cardioid, try and aim for the smoothest and least disconnected set of personalities as you walk around it. The dipoles will sound different than they do in the front - they can sound more "spiky", less smooth, a little more brash. The quasi-cardioids can sound either muffled or pretty smooth - this is up to you, or the designer.
The only ones I know that sound exactly the same as you circumnavigate the speaker are Walsh and MBL radials - everything else has multiple personalities, some of them pretty ugly-sounding. If the "pretty" or "handsome" personality is only audible over a small angle, I'd say you have trouble.
But guess what? Your listening room receives all of these personalities, all at the same time! The emissions take off in all directions, and unless the room is extremely heavily damped (and sounds terrible as a result), all of the personalities appear in the reflections, but spatially scrambled, since there are not just one, but many bounces for the reflections. What started out as a rear wave bounces off the rear wall, then caroms off the sides, ceiling, or even behind the listener before arriving at the ear.
Fortunately, the first direct-arrival wave dominates the spectral perception - the ear gives it a strong preference, stonger than a FFT or RTA measuring system. But - the room spectra matters too, since that is largely responsible for spatial impression, and if that is distorted like a funhouse mirror, there will something subtly "wrong" with the overall impression of the speaker - particularly in stereo, where spatial impression is important.
So a simple walkaround can tell you a lot about a speaker, and where it has problems. This walkaround will tell you more than any number of polar charts.
3) Play a symphonic piece at a moderate level with a fairly continuous line - choose a piece you like, this test won't destroy your faith as much as the first two. Shut down the other channel - you can use a mono source if you want, just choose something with a pleasing and natural-sounding tonal balance.
Walk up to the remaining speaker so you're about a meter away, then walk around it slowly, completing one circle, then keep moving slowly round and round. Stop and listen what the speaker sounds like in the rear quadrant. Most monopoles (nearly all) sound - well - "gurgly" and bass-heavy from the rear. The "gurgly" quality are cabinet colorations at close range (you are close enough so room colorations don't matter) and the dull, muffled quality is the strange, heavily diffracted sound of the tweeter.
Note the angle where the sound quality suddenly changes - and with most square-sided cabinets, it will - and pay attention to the different "personalities" the speaker has at different angles. You may notice certain angles are "critical" angles, like 45 degrees and multiples of that - listen for phantom tweeters seeming to come from cabinet edges. Listen overhead, too - that will probably be different as well. Move around the speaker and see how many different personalities it has, and what they sound like.
Try this with dipoles, and if you can get access to them, quasi-cardioid speakers (dipoles with absorption of the rear-wave). They have a completely different set of personalities, don't they? The dipoles are different than the quasi-cardioids - in fact, if you're "tuning" a quasi-cardioid, try and aim for the smoothest and least disconnected set of personalities as you walk around it. The dipoles will sound different than they do in the front - they can sound more "spiky", less smooth, a little more brash. The quasi-cardioids can sound either muffled or pretty smooth - this is up to you, or the designer.
The only ones I know that sound exactly the same as you circumnavigate the speaker are Walsh and MBL radials - everything else has multiple personalities, some of them pretty ugly-sounding. If the "pretty" or "handsome" personality is only audible over a small angle, I'd say you have trouble.
But guess what? Your listening room receives all of these personalities, all at the same time! The emissions take off in all directions, and unless the room is extremely heavily damped (and sounds terrible as a result), all of the personalities appear in the reflections, but spatially scrambled, since there are not just one, but many bounces for the reflections. What started out as a rear wave bounces off the rear wall, then caroms off the sides, ceiling, or even behind the listener before arriving at the ear.
Fortunately, the first direct-arrival wave dominates the spectral perception - the ear gives it a strong preference, stonger than a FFT or RTA measuring system. But - the room spectra matters too, since that is largely responsible for spatial impression, and if that is distorted like a funhouse mirror, there will something subtly "wrong" with the overall impression of the speaker - particularly in stereo, where spatial impression is important.
So a simple walkaround can tell you a lot about a speaker, and where it has problems. This walkaround will tell you more than any number of polar charts.
This one is so simple it looks really stupid. That doesn't stop me when nobody is looking.
4) Cabinets have really weird-sounding internal standing wave colorations, and especially cabinets where multiple drivers share a common chamber (line arrays, I'm looking at you). If you visit the proud audiophile or manufacturer when the drivers are pulled out, if the baffle opening is big enough, put your head inside (don't get stuck!!!), or if you want to play it safe, put your ear right at the aperture.
Listen for awhile. Note just how different it sounds than when you're several feet away in free air. The cabinet, even if is filled with Miracle Absorbing Material, has a hollow, muffled, resonant - well, boxlike quality, but with odd colorations from the damping techniques themselves. It's not really a simple "box" sound - there's other stuff, too, and hard to describe.
Speaker diaphragms are acoustically transparent, no matter what material they are made from. These weird boxy qualities are there all the time, overlaying their signature on the honest sound of the driver. This is especially severe in line arrays with a common rear chamber - the best reason for using multiple, small, isolated chambers, preferably of dissimilar dimensions.
There. Four listening techniques that will reveal a great deal of the gross and obvious defects of speakers from $100 to $100,000.
4) Cabinets have really weird-sounding internal standing wave colorations, and especially cabinets where multiple drivers share a common chamber (line arrays, I'm looking at you). If you visit the proud audiophile or manufacturer when the drivers are pulled out, if the baffle opening is big enough, put your head inside (don't get stuck!!!), or if you want to play it safe, put your ear right at the aperture.
Listen for awhile. Note just how different it sounds than when you're several feet away in free air. The cabinet, even if is filled with Miracle Absorbing Material, has a hollow, muffled, resonant - well, boxlike quality, but with odd colorations from the damping techniques themselves. It's not really a simple "box" sound - there's other stuff, too, and hard to describe.
Speaker diaphragms are acoustically transparent, no matter what material they are made from. These weird boxy qualities are there all the time, overlaying their signature on the honest sound of the driver. This is especially severe in line arrays with a common rear chamber - the best reason for using multiple, small, isolated chambers, preferably of dissimilar dimensions.
There. Four listening techniques that will reveal a great deal of the gross and obvious defects of speakers from $100 to $100,000.
Hi Lynn,
Yes, sticking your head inside a cabinet to listen reveals many different sounds with all sorts of unexpected peaks reverberating between the walls. I have noted they seem to change in tonality, becoming more intense at mid frequencies with increasing levels of reproduction, whereas hearing variation with level would be more sensitive to bass and treble with increasing level.
I wonder if these 'interferences' all relate to the cabinet itself, or whether there can be changes due to air motion over internal boundaries and edges, also around the likes of TL cabinet foldings. Maybe it is just the bass pressures increasing cabinet ringing.
And what you say about running the tweeter on its own; this can also reveal amplifier weaknesses, say in some lesser SS types and especially when these are compared to tube.
Cheers ........... Graham.
Yes, sticking your head inside a cabinet to listen reveals many different sounds with all sorts of unexpected peaks reverberating between the walls. I have noted they seem to change in tonality, becoming more intense at mid frequencies with increasing levels of reproduction, whereas hearing variation with level would be more sensitive to bass and treble with increasing level.
I wonder if these 'interferences' all relate to the cabinet itself, or whether there can be changes due to air motion over internal boundaries and edges, also around the likes of TL cabinet foldings. Maybe it is just the bass pressures increasing cabinet ringing.
And what you say about running the tweeter on its own; this can also reveal amplifier weaknesses, say in some lesser SS types and especially when these are compared to tube.
Cheers ........... Graham.
My favorite tests for box noise is to place the speaker in a dead, large or outdoor environment and then run an MLS signal, or white noise. Kill it instantaneously and listen for the "boing" that trails the end of the signal, leading to silence. 90% of venteds make nice flower pots after this test.
My friend John Atwood (former technical editor of Vacuum Tube Valley and of day jobs at Intel, Sequent, and Apple) likes to speak of the "example proof". Here's some are some examples:
There were some interesting theoretical advantages for transformer coupling in a Class A PP triode amplifier - Norm Crowhurst just barely hinted at it, and there were some interesting amplifier of the Thirties that used this method. I was intrigued and curious. Rather than follow the example of mid-Nineties SET enthusiasts - who were only interested in RC-coupled SETs, I decided to find out for myself when I designed my first amplifier. Zero support from Arthur Loesch, the NY Triode mafia guys, anyone in Oregon, all of whom thought I was nuts and at the least, wasting my time. "Everybody knows" that PP is no good at low levels, it takes a SET to get the First Watt right, blah blah blah.
Well, theory informed me that Class A PP with DHT triodes would have the lowest distortion at all levels, and Western Electric was unlikely to have chosen this topology for their most advanced 300B amplifier without a good technical reason. Further research turned up significant advantages for transformer coupling (in a non-feedback amplifier) - precisely symmetric drive for both grids under dynamic conditions, no RC-coupled "blocking" when grid current was drawn, driver distortion reduced by severalfold, and instantaneous recovery from overload. Not small advantages.
Well, when the Amity was completed, surprise surprise. It didn't sound anything like a conventional PP amplifier, nor did it sound like a SET, either. And the intrinsic distortion, with no feedback anywhere, measured lower than a SET or a conventional PP amplifier (when the loop was opened). In fact, it had lower distortion with zero feedback than a conventional Williamson or Dyna has with 20 dB of feedback!
This is an example proof. Investigate the theory, build, measure, and listen. Ignore the fashion crowd who don't have the courage to think, build, and listen for themselves. All they do is consume, and hide behind the opinions of others.
Dr. Geddes had the courage to go outside existing horn theory, develop his own theories, model them, build, measure, and listen. That's an example proof. It carries a lot more weight than any opinion on the Internet or in a magazine.
But there's a difference. Dr. Geddes is using the more rigorous Floyd Toole protocol of a double-blind listening panel to determine which acoustic parameters to optimize, and which to give less weight. I'm more like an artisan, using theory, building, measuring, and listening for my own pleasure, not the satisfaction of a listening panel. We're using similar tools, but our goals are different.
I think you can see by now why I give so little weight to anything I haven't personally auditioned, especially if there are no measurements available. As an artisan, I look for examples that sound "beautiful" - the Quad ESL57, Siegfried Linkwitz's speakers, and the Bastanis Apollo - and discover what they have in common, and the underlying design theories. If the designer is still around, so much the better, since much of the more interesting theory & practice is never written down or published. You have to meet the designers in person to find out what they really think - or if they hired somebody else to do the job.
So if some guy in Boston thinks "dipoles are for primitive tastes", so what? There are three well-known example proofs that point in the other direction. I've listened to all of them at considerable length, and don't need anyone to dictate their tastes to me. If you see yourself as an artisan, cultivate your own tastes, and learn the methodologies from others.
There were some interesting theoretical advantages for transformer coupling in a Class A PP triode amplifier - Norm Crowhurst just barely hinted at it, and there were some interesting amplifier of the Thirties that used this method. I was intrigued and curious. Rather than follow the example of mid-Nineties SET enthusiasts - who were only interested in RC-coupled SETs, I decided to find out for myself when I designed my first amplifier. Zero support from Arthur Loesch, the NY Triode mafia guys, anyone in Oregon, all of whom thought I was nuts and at the least, wasting my time. "Everybody knows" that PP is no good at low levels, it takes a SET to get the First Watt right, blah blah blah.
Well, theory informed me that Class A PP with DHT triodes would have the lowest distortion at all levels, and Western Electric was unlikely to have chosen this topology for their most advanced 300B amplifier without a good technical reason. Further research turned up significant advantages for transformer coupling (in a non-feedback amplifier) - precisely symmetric drive for both grids under dynamic conditions, no RC-coupled "blocking" when grid current was drawn, driver distortion reduced by severalfold, and instantaneous recovery from overload. Not small advantages.
Well, when the Amity was completed, surprise surprise. It didn't sound anything like a conventional PP amplifier, nor did it sound like a SET, either. And the intrinsic distortion, with no feedback anywhere, measured lower than a SET or a conventional PP amplifier (when the loop was opened). In fact, it had lower distortion with zero feedback than a conventional Williamson or Dyna has with 20 dB of feedback!
This is an example proof. Investigate the theory, build, measure, and listen. Ignore the fashion crowd who don't have the courage to think, build, and listen for themselves. All they do is consume, and hide behind the opinions of others.
Dr. Geddes had the courage to go outside existing horn theory, develop his own theories, model them, build, measure, and listen. That's an example proof. It carries a lot more weight than any opinion on the Internet or in a magazine.
But there's a difference. Dr. Geddes is using the more rigorous Floyd Toole protocol of a double-blind listening panel to determine which acoustic parameters to optimize, and which to give less weight. I'm more like an artisan, using theory, building, measuring, and listening for my own pleasure, not the satisfaction of a listening panel. We're using similar tools, but our goals are different.
I think you can see by now why I give so little weight to anything I haven't personally auditioned, especially if there are no measurements available. As an artisan, I look for examples that sound "beautiful" - the Quad ESL57, Siegfried Linkwitz's speakers, and the Bastanis Apollo - and discover what they have in common, and the underlying design theories. If the designer is still around, so much the better, since much of the more interesting theory & practice is never written down or published. You have to meet the designers in person to find out what they really think - or if they hired somebody else to do the job.
So if some guy in Boston thinks "dipoles are for primitive tastes", so what? There are three well-known example proofs that point in the other direction. I've listened to all of them at considerable length, and don't need anyone to dictate their tastes to me. If you see yourself as an artisan, cultivate your own tastes, and learn the methodologies from others.
DDF said:
True. This is a variation of the 1950's BBC chopped-sinewave technique for finding what D.E.L. Shorter called "buried resonance". Back then, they used a scope with a delayed trigger to see what interesting things happened when the sinewave was suddenly cut off - they'd slowly sweep the driver and look for high-Q "tails" on the scope.
Nowadays, we use a digital CSD to see the same thing. But interrupted-transient methods tell us a lot in a short time, both by direct audition and measurement. Direct audition of clicks, interrupted noise, steady-state noise, and broad-spectrum music (not jazz or pop music) tell us a lot when we listen for "multiple personality" dispersion and box & driver colorations. Sitting in a chair and trying to impress other audiophiles with your listening skills tells you very little.
Rattles
Guys/Lynn ...
My favourite speaker test uses this free little jewel http://www.audiotester.de/ easily hooked up to your system. It will do wonders. Frequency steps are 1/100th. Do not use an alternate sound source as you need the exact interaction between your room/house and your system. Or, even within the speaker system itself. I use an old stethoscope (the heart ones are better as they have a deep frequency side) for the speakers themselves.
You can find rattles and resonances you just thought you corrected. Sine waves may not be best. Others/Lynn should have a better grasp on that end.
It's not the rattles per say that gets you, it's the feedback it causes at the speakers.
So if you find the plaster board is rattling (that be me with old 1/2" sheet rock), that's not a real problem, either. You'll probably be painting anyway, someday. Hit the board with sheetrock screws. Stud finder <$20. There are wires in the walls. Check with contractor as to what length screws to use. Fill holes and repaint. It's not a big deal unless you make it so. Just do a little at a time. One wall in a weekend shouldn't kill almost anyone.
I love it, almost like an Easter Egg hunt. BTW, after you go thru the first round, change the volume and do again. Funner than you might think. Hell, get the Wif and kids to help.
For large items like a glass china cabinet that cannot be cured, just throw it out. Me, divorced? How did you know?
Hardware to correct problems? Minimal in most cases, system improvement, priceless.
Zene
Guys/Lynn ...
My favourite speaker test uses this free little jewel http://www.audiotester.de/ easily hooked up to your system. It will do wonders. Frequency steps are 1/100th. Do not use an alternate sound source as you need the exact interaction between your room/house and your system. Or, even within the speaker system itself. I use an old stethoscope (the heart ones are better as they have a deep frequency side) for the speakers themselves.
You can find rattles and resonances you just thought you corrected. Sine waves may not be best. Others/Lynn should have a better grasp on that end.
It's not the rattles per say that gets you, it's the feedback it causes at the speakers.
So if you find the plaster board is rattling (that be me with old 1/2" sheet rock), that's not a real problem, either. You'll probably be painting anyway, someday. Hit the board with sheetrock screws. Stud finder <$20. There are wires in the walls. Check with contractor as to what length screws to use. Fill holes and repaint. It's not a big deal unless you make it so. Just do a little at a time. One wall in a weekend shouldn't kill almost anyone.
I love it, almost like an Easter Egg hunt. BTW, after you go thru the first round, change the volume and do again. Funner than you might think. Hell, get the Wif and kids to help.
For large items like a glass china cabinet that cannot be cured, just throw it out. Me, divorced? How did you know?
Hardware to correct problems? Minimal in most cases, system improvement, priceless.
Zene
- For listening around a speaker, I must add that FR bipoles are good at it too. For anyone not familiar with FR bipoles, Planet 10 DIY audio moderator is the master bipole main man for anyone interested further. He has a site and many posts. I have made some too, and a small one, the ''TB3 Revisited'' got built by American and European members here, and it has even been used with a Hiraga Amp by a Dutch member.
- For listening in a box, I want to add a simple technique I use when choosing stuffing. I put my head and talk in to the box. Every time my voice sounds the best I can achieve by changing type and arranging the stuffing, so it does the mid bass of the final speaker.
- For listening in a box, I want to add a simple technique I use when choosing stuffing. I put my head and talk in to the box. Every time my voice sounds the best I can achieve by changing type and arranging the stuffing, so it does the mid bass of the final speaker.
I found a quick way of listening to the 'sound' of a LS enclosure is to tape a few layers of newspaper over a driver opening and tap it firmly - you hear the cabinet 'tone' without driver influence or filtering.
The sound is also very similar to what you hear when you tap the cone of an electrically disconnected driver in the cabinet, and it shows how you can hear much more from the cabinet via the driver than from the driver itself.
A good driver connected to a competent NFB amplifier with short or low resistance cables does not resonate similarly at the lower frequencies due to amp/driver damping, but it does not take much in the way of crossover inductor, cable length or even a short length of inadequate cable for the enclosure 'tones' to start becoming audible during music reproduction.
(Some of the older commercial 'Hi-Fi' LS used woefully thin internal interconnects)
(I think I have just presented a case for driver sited plate amps at LF.)
Cheers .......... Graham.
The sound is also very similar to what you hear when you tap the cone of an electrically disconnected driver in the cabinet, and it shows how you can hear much more from the cabinet via the driver than from the driver itself.
A good driver connected to a competent NFB amplifier with short or low resistance cables does not resonate similarly at the lower frequencies due to amp/driver damping, but it does not take much in the way of crossover inductor, cable length or even a short length of inadequate cable for the enclosure 'tones' to start becoming audible during music reproduction.
(Some of the older commercial 'Hi-Fi' LS used woefully thin internal interconnects)
(I think I have just presented a case for driver sited plate amps at LF.)
Cheers .......... Graham.