Martin, if you're reading this, I'm curious what your impressions between using the Lowther all the way up to its rolloff point around 12~13 kHz, and the compression driver you're now using in the same range. I'd guess the mids sound more open, relaxed, and effortless, with a more spacious and "tactile" quality. But that's only a guess. Curious what your impression has been.
<<
I have come to realise, playing with ruined pianos, saxophones and Le Cleac'h horns that the fabled 'midrange' is a mythical concept when it comes to real (non sine wave) instruments and voices. Tone depends on all audible frequencies, and possibly inaudible. Lowther bass improves to my ears with the addition of a compression driver, probably because the percussive leading edges are better defined. Likewise the choice of bass reproducer probably effects teh subjective tonality of the compression driver. The full range Lowther is at ease with itself (if not turned up too high).
I was surprised to discover that my 2.3m long 50Hz Lowther bass horn sounds wonderful crossed over at 800Hz to the 425 horn. I thought 'line of sight' would be essential over for the crucial 2 octaves above 200Hz.
Where does good tonality come from? So much about the construction and room / system interactions comes into play. Good imaging comes easy - given controlled directivity I guess. Compression drivers image better, but getting the tonality right I think much depends on the bass driver, and probably the tweeter too, if used.
martin
<<
I have come to realise, playing with ruined pianos, saxophones and Le Cleac'h horns that the fabled 'midrange' is a mythical concept when it comes to real (non sine wave) instruments and voices. Tone depends on all audible frequencies, and possibly inaudible. Lowther bass improves to my ears with the addition of a compression driver, probably because the percussive leading edges are better defined. Likewise the choice of bass reproducer probably effects teh subjective tonality of the compression driver. The full range Lowther is at ease with itself (if not turned up too high).
I was surprised to discover that my 2.3m long 50Hz Lowther bass horn sounds wonderful crossed over at 800Hz to the 425 horn. I thought 'line of sight' would be essential over for the crucial 2 octaves above 200Hz.
Where does good tonality come from? So much about the construction and room / system interactions comes into play. Good imaging comes easy - given controlled directivity I guess. Compression drivers image better, but getting the tonality right I think much depends on the bass driver, and probably the tweeter too, if used.
martin
Much appreciated, Martin. Sigh. I've done many speakers that measured really well during the development process, on and off-axis, but the tone just wasn't there. Sometimes doing really subtle things with the crossover can bring them back to life - that happened several times with the Ariel - but as often as not the drivers just aren't suitable for each other, and no amount of crossover trickery can rescue them. Sometimes the whole design concept just plain doesn't work, regardless of which drivers are chosen.
It's been my experience that if the drivers don't sound natural and have an appealing sound quality, the crossover usually can't rescue them. You can use notch filters to get rid of obnoxious peaks at the top of the band, or even worse, in the passband, but there's a price to be paid for doing that. Back at Audionics in the Seventies, I used notch filters on all of my speakers, so they measured as flat as anything on the market at that time, but they tended to sound drab and colorless. The less screwing around I did, the better they sounded - but the trick was to find drivers that had a pleasant sound without "help" from the crossover.
I've flirted briefly with minimalist one-cap crossovers (with and without shunting resistors), but those haven't satisfied either - too severe a price in dynamic compression and odd driver interactions near the crossover point. So I've ended up using low-Q second-order electrical filters, and twiddling around with finding out the perceptual threshold of crossover coloration. The more transparent the drivers and amplification, the lower this threshold is.
There's a fine balance between over-engineering (and chasing out all the colorations) and having a lively-sounding speaker system (but with persistent colorations that become annoying over time). It's taken me a long time to realize that you have to leave the basic character of the driver alone; if you don't like a Lowther, for example, equalizing it won't help. An equalized Lowther has none of the life and sparkle of a "natural" Lowther while still having more coloration than a generic audiophile driver. Similarly, if you mess around with one of the those carbon-fiber audiophile drivers, you're never going to really successfully clean up the HF breakup region, no matter how tricky the crossover is. You have to come to terms with the sound of the drivers as-is, then do a little clean-up (mostly excursion control) on the margins.
With the driver line-up I have, I'm hoping it retains the basic Altec A5 character but without the severe colorations of the old-school horns. There's a certain quality to the LeCleac'h horn that is hard to pin down but sounds very open and spacious to me; it's not the dispersion, but something to do with the tonality and a certain wide-open, effortless quality to the sound.
I think part of your experience with the "midrange" finding itself all over the band - in places you don't expect - is the concept of drivers having reserve bandwidth. A lot of modern designers write off the drivers an octave (or less) outside the intended frequency range, and rely on the crossover filters to dump the unwanted parts of the driver's response. I think this is a mistake; if the drivers have good-quality sound well outside the "intended" bandwidth, the system sounds more relaxed, more natural, and more tuneful, and less like a big machine that's clanking along trying to fake music.
That's my big gripe about the sonic fashion in the modern high-end; the sound just isn't very natural. These overcooked speakers sound OK on the sparse ultra-fi CDs that are popular with the hifi show crowd, but really fall down on classic recordings made in the Fifties and Sixties. If a hifi system only sounds good - or even listenable - on a handful of demo-quality (digital) recordings, something is terribly, terribly wrong.
This is where reserve bandwidth and reserve headroom come in; if the drivers have an appealing quality on their own (don't require heavy EQ), and there is ample headroom in the frequency and dynamic domains, then the drivers (and amplifiers) never enter a high-stress region, or even come close. Amplifiers are at their worst when parts of the circuit approach slewing, excess current demand, or an on-off switching transition; similarly, drivers are at their worst when too much excursion, or too much energy being directed to a HF breakup region, is asked of them. The task of the crossover, along with the overall system design, is to gently guide the energy away from these high-stress regions.
Strictly a wild guess, but I suspect you have plenty of reserve bandwidth in the two horn systems, both the LF Lowther and the large-format AH-425.
It's been my experience that if the drivers don't sound natural and have an appealing sound quality, the crossover usually can't rescue them. You can use notch filters to get rid of obnoxious peaks at the top of the band, or even worse, in the passband, but there's a price to be paid for doing that. Back at Audionics in the Seventies, I used notch filters on all of my speakers, so they measured as flat as anything on the market at that time, but they tended to sound drab and colorless. The less screwing around I did, the better they sounded - but the trick was to find drivers that had a pleasant sound without "help" from the crossover.
I've flirted briefly with minimalist one-cap crossovers (with and without shunting resistors), but those haven't satisfied either - too severe a price in dynamic compression and odd driver interactions near the crossover point. So I've ended up using low-Q second-order electrical filters, and twiddling around with finding out the perceptual threshold of crossover coloration. The more transparent the drivers and amplification, the lower this threshold is.
There's a fine balance between over-engineering (and chasing out all the colorations) and having a lively-sounding speaker system (but with persistent colorations that become annoying over time). It's taken me a long time to realize that you have to leave the basic character of the driver alone; if you don't like a Lowther, for example, equalizing it won't help. An equalized Lowther has none of the life and sparkle of a "natural" Lowther while still having more coloration than a generic audiophile driver. Similarly, if you mess around with one of the those carbon-fiber audiophile drivers, you're never going to really successfully clean up the HF breakup region, no matter how tricky the crossover is. You have to come to terms with the sound of the drivers as-is, then do a little clean-up (mostly excursion control) on the margins.
With the driver line-up I have, I'm hoping it retains the basic Altec A5 character but without the severe colorations of the old-school horns. There's a certain quality to the LeCleac'h horn that is hard to pin down but sounds very open and spacious to me; it's not the dispersion, but something to do with the tonality and a certain wide-open, effortless quality to the sound.
I think part of your experience with the "midrange" finding itself all over the band - in places you don't expect - is the concept of drivers having reserve bandwidth. A lot of modern designers write off the drivers an octave (or less) outside the intended frequency range, and rely on the crossover filters to dump the unwanted parts of the driver's response. I think this is a mistake; if the drivers have good-quality sound well outside the "intended" bandwidth, the system sounds more relaxed, more natural, and more tuneful, and less like a big machine that's clanking along trying to fake music.
That's my big gripe about the sonic fashion in the modern high-end; the sound just isn't very natural. These overcooked speakers sound OK on the sparse ultra-fi CDs that are popular with the hifi show crowd, but really fall down on classic recordings made in the Fifties and Sixties. If a hifi system only sounds good - or even listenable - on a handful of demo-quality (digital) recordings, something is terribly, terribly wrong.
This is where reserve bandwidth and reserve headroom come in; if the drivers have an appealing quality on their own (don't require heavy EQ), and there is ample headroom in the frequency and dynamic domains, then the drivers (and amplifiers) never enter a high-stress region, or even come close. Amplifiers are at their worst when parts of the circuit approach slewing, excess current demand, or an on-off switching transition; similarly, drivers are at their worst when too much excursion, or too much energy being directed to a HF breakup region, is asked of them. The task of the crossover, along with the overall system design, is to gently guide the energy away from these high-stress regions.
Strictly a wild guess, but I suspect you have plenty of reserve bandwidth in the two horn systems, both the LF Lowther and the large-format AH-425.
Hi Lynn / all, I am curious what lies behind the choice of the mid-high variant of the RAAL ribbon tweeter (140-15D) as opposed to the high (70-10). Given that the intention is to cross over at about 5K Hz (if I understand correctly) would not the 70-10 be a better choice?
I hooked up some ribbon tweeters of my own (Fountek jp3.0) today and searched for the best sounding crossover point and slopes (using a dbx driverack pa). I ended up with 6K Hz and 12dB Butterworth. And yes, it makes a good deal of difference once one gets it right. The ribbons are heaps better than the rather nice looking old Yamaha alnico 1" dome tweeters (JA0505) I was using. The ribbons got better when I took off the plastic baffle plate as RAAL advise. Question now is of course - how much better will the RAAL be? The fact that I get such good results with the cheap (reletively) Fountek encourages me a lot to fork out the dosh for the RAAL. But I cant really see the point of the mid-high variant given the 6K cross over point. And it may actually be worse given one probably doesnt really want the extra ribbon length and the foam lens trick for the high end response.
This leads to a further question related to diffraction - what determines the optimum 'baffle' size and shape for a 5K - ???K transducer? And what do the saw shaped RAAL edges seek to achieve? Easy enough to experiment with (and I mean to) - but what is the theory?
I have to say this is all sounding superb - even the wife stayed to listen and said it sounded 'crisp'. High praise I can tell you.
Best rgds,
Martin
I hooked up some ribbon tweeters of my own (Fountek jp3.0) today and searched for the best sounding crossover point and slopes (using a dbx driverack pa). I ended up with 6K Hz and 12dB Butterworth. And yes, it makes a good deal of difference once one gets it right. The ribbons are heaps better than the rather nice looking old Yamaha alnico 1" dome tweeters (JA0505) I was using. The ribbons got better when I took off the plastic baffle plate as RAAL advise. Question now is of course - how much better will the RAAL be? The fact that I get such good results with the cheap (reletively) Fountek encourages me a lot to fork out the dosh for the RAAL. But I cant really see the point of the mid-high variant given the 6K cross over point. And it may actually be worse given one probably doesnt really want the extra ribbon length and the foam lens trick for the high end response.
This leads to a further question related to diffraction - what determines the optimum 'baffle' size and shape for a 5K - ???K transducer? And what do the saw shaped RAAL edges seek to achieve? Easy enough to experiment with (and I mean to) - but what is the theory?
I have to say this is all sounding superb - even the wife stayed to listen and said it sounded 'crisp'. High praise I can tell you.
Best rgds,
Martin
Hi Lynn
You say this like a colorless speaker is a bad thing. I have always felt that the speaker should be colorless.
And again the implication that having some coloration is desired. This seems to be a very odd position to me.
I'm proud to say that my speakers are "colorless" and perhaps even dull or "drab". Thats what I am after. It may take some getting used to, many say this, but in the end, after you do, everything else is just too colored to be pleasant. The color is in the music, not in the speaker.
Lynn Olson said:
You say this like a colorless speaker is a bad thing. I have always felt that the speaker should be colorless.
And again the implication that having some coloration is desired. This seems to be a very odd position to me.
I'm proud to say that my speakers are "colorless" and perhaps even dull or "drab". Thats what I am after. It may take some getting used to, many say this, but in the end, after you do, everything else is just too colored to be pleasant. The color is in the music, not in the speaker.
Hello,
I guess that most probably (but he'll confirm or not) Lynn will agree that a loudspeaker that doesn't add its own coloration is probably a good thing.
How I interpreted the sentence you quoted ,is that Lynn meant that a loudspeaker which one pulls out the life from music rendering the music less vivid or even "grey" (without colour) is not so desireable.
Best regards from Paris, France
Jean-Michel Le Cléac'h
I guess that most probably (but he'll confirm or not) Lynn will agree that a loudspeaker that doesn't add its own coloration is probably a good thing.
How I interpreted the sentence you quoted ,is that Lynn meant that a loudspeaker which one pulls out the life from music rendering the music less vivid or even "grey" (without colour) is not so desireable.
Best regards from Paris, France
Jean-Michel Le Cléac'h
gedlee said:
Hello Earl,
I envy you, people like me have often some pain listening to such systems that "pull out the life from music".
But I think that, today, we don't know everything in audio and specially about our perceptions and for the moment we are just unable to explain the complex process beneath that feeling (the pull out...). May be one of those days we will know.
Does this situation forbids us to speak about what we feel as a "pull out of the life of music" even if we cannot explain it. I don't think so.
Best regards from Paris, France
Jean-Michel Le Cléac'h
I envy you, people like me have often some pain listening to such systems that "pull out the life from music".
But I think that, today, we don't know everything in audio and specially about our perceptions and for the moment we are just unable to explain the complex process beneath that feeling (the pull out...). May be one of those days we will know.
Does this situation forbids us to speak about what we feel as a "pull out of the life of music" even if we cannot explain it. I don't think so.
Best regards from Paris, France
Jean-Michel Le Cléac'h
gedlee said:
Colour
Lynn's post 3640 from page 146 of this thread nicely covers the topic of 'colouration' and rendition of 'tonal colour' in audio systems .
I know because I just spent about 3 hours going through the thread to see exactly what prompted Lynn's transition from 390 to 288 compression driver....still haven't found it !!
MJ
Lynn's post 3640 from page 146 of this thread nicely covers the topic of 'colouration' and rendition of 'tonal colour' in audio systems .
I know because I just spent about 3 hours going through the thread to see exactly what prompted Lynn's transition from 390 to 288 compression driver....still haven't found it !!
MJ
Looks like posts 3891 and 3899, I answered my own question .
Lynn, if I order some 8-ohm ones, do I need to consider "the lower power 16-ohm diaphragm" ?
My deliberations on the 288 vs. other options have come because the price has gone up ( Alnico tax !? ) and the very timely and useful drop in value of the pound ...
Still looks a good choice though ; despite Romy's assertions the 288 series are colourless .
MJ
Lynn, if I order some 8-ohm ones, do I need to consider "the lower power 16-ohm diaphragm" ?
My deliberations on the 288 vs. other options have come because the price has gone up ( Alnico tax !? ) and the very timely and useful drop in value of the pound ...
Still looks a good choice though ; despite Romy's assertions the 288 series are colourless .
MJ
Jmmlc said:
Jean-Michel
When one simply takes your position they don't seek to understand perception in objective and quantifiable terms. I don't accept this position and I do seek scientific aproaches to what we hear. We are much better at it than you assume and you seem to be unwilling to move in the direction of science, instead holding on to the ideals of subjectivism.
Subjectivists tend to hide behind this view, that "we are just unable to explain the complex process beneath that feeling". It makes life easy to not have to attempt to explain or quantify what we perceive - "it sounds good to me and thats all I need to know." What could be easier.
I prefer to take the harder road of science where things need useful and well defined terminology and quantifiable objective measures. Lifes tough.
I appreciate the previous posts - these are good points that need expansion and discussion. A big part of the problem are the limitations of the English language in describing subtle aspects of sound and the emotional reactions of the listener.
I would be the first to agree that an overall coloration is extremely undesirable in any component that is represented as "high fidelity". To switch domains, in cooking, it would similar to a sauce slathered on every dish regardless of taste, or those poor people who dump salt on their plate before they even taste what's being served. In photography, it would be like a color cast, like the murky green-brown shadows areas of Ektachrome, or the purple skies of Kodachrome.
These dominant colorations are fairly easy to track down, appearing as peaks or elevated regions in the frequency response, long-persisting resonances in the CSD, or chaotic time decay in the impulse response. In electronics, slewing artifacts, Class AB switching artifacts, power-supply hash, RFI breakthrough, and other low-level problems add a sense of "grit" and "grain" to the musical presentation. With some effort, colorations like these can be chased out of the system - systems that are free of obvious (and measurable) coloration are unfortunately fairly rare in the high-end mainstream.
But there is another aspect of coloration that bears more discussion. I don't like using the subjective word "transparency" - that conveys an ABC sense of "count the instruments" or "oh, you can hear the singer breathing". This, to me, is nothing more than an obtrusive special effect, an artifact of close-miking, and something you don't notice in a live musical performance. I'd feel like an idiot if my first impression of a live performance was "wow, what transparency!"
A word that's closer to what I'm thinking of is "texture". This has a photographic analog: a 35mm camera cannot reproduce fine textures as well as a large-format (4x5 or larger) camera can. Back when I was doing my own darkroom work, I had a Schneider Componon enlarging lens, a Bogen enlarger, a grain-focuser, and knew my way around the Pentax Spotmatic (bought new in Hong Kong in 1964). Although I was working close to the limits of what 35mm could do, I found out the hard way that photographs of sand looked like - concrete. It took a real close-up, getting within a few inches of driftwood resting on sand, or switching to medium or large-format to render the sand with any sense of realism. The 35mm medium has impact, but it doesn't render textures that well. It's taken 10~14 megapixel DSLR's and careful post-processing in Photoshop to approach the texture-rendition of medium-format cameras.
The same applies to cooking. Fresh spices have hundreds of flavor components, and these degrade differentially on exposure to air, heat, and ultraviolet light. A spice that's been sitting on the shelf for several years is much flatter and at the same time harsher-tasting than a fresh, high-quality one. Cooking with worn-out spices narrows the flavor gamut and coarsens the overall result.
When you listen to live acoustic (unamplified) music, one of the most striking qualities of the sound is the sheer vividness of the tonal colors, the way you can easily hear what the instruments are made of, and the expressiveness of the playing. The vividness of the sound gives it a physical impact that just isn't there in most reproduced sound - this isn't a matter of loudness. I've been moved to tears by the sound of the 16th-Century harpsichords in the Deutsche Museum in Munich; the sound was was like no recording I've ever heard, and the sheer beauty of the sound had me rooted to the spot for an hour. There is a quality of tactility - of physicality - to real, acoustic sound that is very, very rare in reproduced sound.
The inverse of this is the monotonous artificiality of MP3 and other forms of lossy digital compression. No reverb, no sense of acoustics or presence, just a loud hammering quality that sounds thoroughly electronic and "processed" - the audio equivalent of eating McDonald's hamburgers day after day. Flat, cartoon-like, and not even close to the real thing.
Most hifi systems sound somewhere in between these two extremes. On the occasional "demo" recording with sparse instrumentation, there's a fleeting illusion of reality, which collapses abruptly when more complex symphonic or choral material is played. This is what I hear when I go to hifi shows - the demo material sounds occasionally impressive, but when I put on the kind of music I enjoy, the sound quality goes way down, with gross and obvious colorations apparent. Occasionally I hear systems that are reasonably free of obvious artifacts, but the music still sounds a long way from the vividness of the real thing.
It was working with direct-heated triode amplifiers that led to an interest in this "live", vivid quality in loudspeakers. With electronics, it's easy to flip a switch between circuit topologies and do a A/B/A audition, and then measure in some depth what the electrical change was (change in harmonic distortion spectra, freedom from slewing, absence or presence of complex power-supply interactions). If there was no audible difference, so what, no matter what the measurements say. On the other hand, there can be immediately audible differences - Gary Pimm and I considered these the ones you heard before you return to your seat - that could be very hard to track down. Power-supply improvements tended to fall into this category, along with improving linear current delivery in critical parts of the circuit (typically in the driver stage, for both transistor and tube amps).
Some of the changes resulted in no obvious "coloration" differences, but immediately obvious differences in "realism" and a sense of "presence". These were repeatable, easily audible to unsuspecting visiting non-audiophiles, and available at a flip of a switch.
Working with the Amity, Aurora, and Karna amplifiers sensitized me to these kind of "vivid" differences that didn't correlate with any obvious tonal coloration. After a while, I started noticing some speakers had more of this quality than others - electrostats and horns in particular. By contrast, large-area magnetic-planars and complex low-efficiency direct-radiator speakers seemed to be most deficient in this quality. There seems to be some correlation with IM distortion, but it's a lot weaker than in amplifiers. Very generous headroom, high efficiency, and freedom from clutter (stored energy) in the time domain seems to have an influence as well.
P.S. Hi Martin, sorry about the long digression. Don't know if it clarified or muddied the waters. Regarding the choice of the RAAL, it came down to two things: matching efficiency and the above-mentioned headroom. The 70-10 is 3 dB or so less efficient, and is not designed to work down into the midrange. It really is a supertweeter. By contrast, the 140-15D is a real tweeter, and Alexander considers it suitable down to 2 kHz (although not necessarily at its best). The biggest problem with (true) ribbon tweeters is excursion - the linear region in the gap is very, very small. Adequate headroom, in terms of more bandwidth than you really need, is the best way to avoid tweeter stress, or getting out of sync with the enormous headroom of the compression driver.
In this case, there is lots of overlap - the 140-15D works down to 2 kHz, and is not working hard at all at 5 kHz (according to Alexander, who ought to know). The 288 with the Tangerine phase plug is reasonably well behaved up to 12~15 kHz. Thus, the 5~7 kHz crossover. This is above the critical 1~3 kHz region, where 1/2 dB changes are readily audible, and below the upper HF region where compression drivers are not at their best (nonuniform wavefronts from the phase plug and the first diaphragm breakup modes). The overall philosophy is keeping all drivers well away from the high-stress regions in the spectral and power domains.
IslandPink, yes, get the "16-ohm low-power diaphragm". This is the classic 288 diaphragm, not the later Pascalite variant. Remember, "high power" in this context means MORE power than a movie theater driver! The 288 had plenty of power for 500 to 2000-seat theaters from 1945 to 1975 - it was the advent of the Crown DC300 and Phase Linear 700 that spelled the end of the 288 diaphragm and the ascendency of the JBL titanium drivers. I don't think you are planning on clipping a 200-watt transistor amp into your horns, are you?
I would be the first to agree that an overall coloration is extremely undesirable in any component that is represented as "high fidelity". To switch domains, in cooking, it would similar to a sauce slathered on every dish regardless of taste, or those poor people who dump salt on their plate before they even taste what's being served. In photography, it would be like a color cast, like the murky green-brown shadows areas of Ektachrome, or the purple skies of Kodachrome.
These dominant colorations are fairly easy to track down, appearing as peaks or elevated regions in the frequency response, long-persisting resonances in the CSD, or chaotic time decay in the impulse response. In electronics, slewing artifacts, Class AB switching artifacts, power-supply hash, RFI breakthrough, and other low-level problems add a sense of "grit" and "grain" to the musical presentation. With some effort, colorations like these can be chased out of the system - systems that are free of obvious (and measurable) coloration are unfortunately fairly rare in the high-end mainstream.
But there is another aspect of coloration that bears more discussion. I don't like using the subjective word "transparency" - that conveys an ABC sense of "count the instruments" or "oh, you can hear the singer breathing". This, to me, is nothing more than an obtrusive special effect, an artifact of close-miking, and something you don't notice in a live musical performance. I'd feel like an idiot if my first impression of a live performance was "wow, what transparency!"
A word that's closer to what I'm thinking of is "texture". This has a photographic analog: a 35mm camera cannot reproduce fine textures as well as a large-format (4x5 or larger) camera can. Back when I was doing my own darkroom work, I had a Schneider Componon enlarging lens, a Bogen enlarger, a grain-focuser, and knew my way around the Pentax Spotmatic (bought new in Hong Kong in 1964). Although I was working close to the limits of what 35mm could do, I found out the hard way that photographs of sand looked like - concrete. It took a real close-up, getting within a few inches of driftwood resting on sand, or switching to medium or large-format to render the sand with any sense of realism. The 35mm medium has impact, but it doesn't render textures that well. It's taken 10~14 megapixel DSLR's and careful post-processing in Photoshop to approach the texture-rendition of medium-format cameras.
The same applies to cooking. Fresh spices have hundreds of flavor components, and these degrade differentially on exposure to air, heat, and ultraviolet light. A spice that's been sitting on the shelf for several years is much flatter and at the same time harsher-tasting than a fresh, high-quality one. Cooking with worn-out spices narrows the flavor gamut and coarsens the overall result.
When you listen to live acoustic (unamplified) music, one of the most striking qualities of the sound is the sheer vividness of the tonal colors, the way you can easily hear what the instruments are made of, and the expressiveness of the playing. The vividness of the sound gives it a physical impact that just isn't there in most reproduced sound - this isn't a matter of loudness. I've been moved to tears by the sound of the 16th-Century harpsichords in the Deutsche Museum in Munich; the sound was was like no recording I've ever heard, and the sheer beauty of the sound had me rooted to the spot for an hour. There is a quality of tactility - of physicality - to real, acoustic sound that is very, very rare in reproduced sound.
The inverse of this is the monotonous artificiality of MP3 and other forms of lossy digital compression. No reverb, no sense of acoustics or presence, just a loud hammering quality that sounds thoroughly electronic and "processed" - the audio equivalent of eating McDonald's hamburgers day after day. Flat, cartoon-like, and not even close to the real thing.
Most hifi systems sound somewhere in between these two extremes. On the occasional "demo" recording with sparse instrumentation, there's a fleeting illusion of reality, which collapses abruptly when more complex symphonic or choral material is played. This is what I hear when I go to hifi shows - the demo material sounds occasionally impressive, but when I put on the kind of music I enjoy, the sound quality goes way down, with gross and obvious colorations apparent. Occasionally I hear systems that are reasonably free of obvious artifacts, but the music still sounds a long way from the vividness of the real thing.
It was working with direct-heated triode amplifiers that led to an interest in this "live", vivid quality in loudspeakers. With electronics, it's easy to flip a switch between circuit topologies and do a A/B/A audition, and then measure in some depth what the electrical change was (change in harmonic distortion spectra, freedom from slewing, absence or presence of complex power-supply interactions). If there was no audible difference, so what, no matter what the measurements say. On the other hand, there can be immediately audible differences - Gary Pimm and I considered these the ones you heard before you return to your seat - that could be very hard to track down. Power-supply improvements tended to fall into this category, along with improving linear current delivery in critical parts of the circuit (typically in the driver stage, for both transistor and tube amps).
Some of the changes resulted in no obvious "coloration" differences, but immediately obvious differences in "realism" and a sense of "presence". These were repeatable, easily audible to unsuspecting visiting non-audiophiles, and available at a flip of a switch.
Working with the Amity, Aurora, and Karna amplifiers sensitized me to these kind of "vivid" differences that didn't correlate with any obvious tonal coloration. After a while, I started noticing some speakers had more of this quality than others - electrostats and horns in particular. By contrast, large-area magnetic-planars and complex low-efficiency direct-radiator speakers seemed to be most deficient in this quality. There seems to be some correlation with IM distortion, but it's a lot weaker than in amplifiers. Very generous headroom, high efficiency, and freedom from clutter (stored energy) in the time domain seems to have an influence as well.
P.S. Hi Martin, sorry about the long digression. Don't know if it clarified or muddied the waters. Regarding the choice of the RAAL, it came down to two things: matching efficiency and the above-mentioned headroom. The 70-10 is 3 dB or so less efficient, and is not designed to work down into the midrange. It really is a supertweeter. By contrast, the 140-15D is a real tweeter, and Alexander considers it suitable down to 2 kHz (although not necessarily at its best). The biggest problem with (true) ribbon tweeters is excursion - the linear region in the gap is very, very small. Adequate headroom, in terms of more bandwidth than you really need, is the best way to avoid tweeter stress, or getting out of sync with the enormous headroom of the compression driver.
In this case, there is lots of overlap - the 140-15D works down to 2 kHz, and is not working hard at all at 5 kHz (according to Alexander, who ought to know). The 288 with the Tangerine phase plug is reasonably well behaved up to 12~15 kHz. Thus, the 5~7 kHz crossover. This is above the critical 1~3 kHz region, where 1/2 dB changes are readily audible, and below the upper HF region where compression drivers are not at their best (nonuniform wavefronts from the phase plug and the first diaphragm breakup modes). The overall philosophy is keeping all drivers well away from the high-stress regions in the spectral and power domains.
IslandPink, yes, get the "16-ohm low-power diaphragm". This is the classic 288 diaphragm, not the later Pascalite variant. Remember, "high power" in this context means MORE power than a movie theater driver! The 288 had plenty of power for 500 to 2000-seat theaters from 1945 to 1975 - it was the advent of the Crown DC300 and Phase Linear 700 that spelled the end of the 288 diaphragm and the ascendency of the JBL titanium drivers. I don't think you are planning on clipping a 200-watt transistor amp into your horns, are you?
Lynn,
I share your views on highfi. Just a few days ago, we were listening to a new amp layout for an active speaker. Although we were comparing only one channel from the same source, the depth of image could vary clearly be compared. I don't think we did much, just the basic engineering stuff doing a reasonable relayout of the board and flattening out the load impedance.
Like taking health foods, some go with the real expensive stuff, some go with the cheap suff. My dad is 90, and he takes care of himself. Shops, cooks, cleans, and keeps track of what comes out of physical/health checkup.
Coffee and tea are also have supporters of various taste. Some like the most natural possible without adding anything in the drying process, some like it straight. Then the weather and soil also effects the taste.
Back to audio, whenever people refer to "life sucked out", it could be one of two things: either the listener has a very strong taste of coloration, or the system actually lacks capability to resolve detail. Some cases, it's both. However, the cause can be very different. There are people here where I live that even put various cups, bowls, tubes, in their listening room so that the additional resonance brings a soothing feel when they listen. It's certainly interesting to see the discussion going on even it's not my cup of tea.
I share your views on highfi. Just a few days ago, we were listening to a new amp layout for an active speaker. Although we were comparing only one channel from the same source, the depth of image could vary clearly be compared. I don't think we did much, just the basic engineering stuff doing a reasonable relayout of the board and flattening out the load impedance.
Like taking health foods, some go with the real expensive stuff, some go with the cheap suff. My dad is 90, and he takes care of himself. Shops, cooks, cleans, and keeps track of what comes out of physical/health checkup.
Coffee and tea are also have supporters of various taste. Some like the most natural possible without adding anything in the drying process, some like it straight. Then the weather and soil also effects the taste.
Back to audio, whenever people refer to "life sucked out", it could be one of two things: either the listener has a very strong taste of coloration, or the system actually lacks capability to resolve detail. Some cases, it's both. However, the cause can be very different. There are people here where I live that even put various cups, bowls, tubes, in their listening room so that the additional resonance brings a soothing feel when they listen. It's certainly interesting to see the discussion going on even it's not my cup of tea.
Lynn,
Your response to query from DIYers especially your response, post no: 4833 seems very pertinent and relevant to good speakers design. I have these thoughts for some time and your response confirms that.
I like to suggest you may like to put some of your responses on your website to assist potential novice DIY speakers designers like myself.
Cheers.
Your response to query from DIYers especially your response, post no: 4833 seems very pertinent and relevant to good speakers design. I have these thoughts for some time and your response confirms that.
I like to suggest you may like to put some of your responses on your website to assist potential novice DIY speakers designers like myself.
Cheers.
