Lynn Olson said:
You got me, I'm really a CIA operative working in Thialand. Ai is all a front. OOPS I'd better be carefull!
Another reason for the diplomatic passport thing for the kids is that children are an assets greatest liability. This has been used over and over. So it is critical that an undercovers kids be squicky clean.
Its been a most interesting story, but I would warn that not all "coincidences" are not "coincidences". Accidents do happen and "out-of-the-blue" situation are often just that. I would say that your fathers promotion to the bank was probably ligit. The bank would cherish his credentials and contacts, even if he were incompetent (not that he was). Seems to me that the later job that he had was probably on the up-and-up. But the OSS to Foreign service is just too typical a path to be ignored.
Yes, no way of knowing if he was NOC or not - and you're right, he was very well-connected with the movers and shakers in Hong Kong, since he was the go-to guy when US and HK businesses wanted to have successful dealings with each other. He was also in charge of a lengthy series of trade negotiations between HK and US textile manufacturers - sounds boring, but there was a lot of money involved.
Both of us were mystified by the Vietnam War - the US certainly didn't care if the French lost their empire, and there was no real strategic interest that justified the enormous costs of propping up the South Vietnam government. With a little effort, the US could have made Uncle Ho into the Asian version of Tito, skillfully playing the Chinese against the Russians, driving both nuts, and frustrating Chinese plans in the region.
Our best guess was the Red Scare purge of the McCarthy years had removed all of the Old China Hands from the upper levels of the Federal government, thus there was no-one left at State or other parts of the US Govt that had any experience in the region.
Probably more significant, there was all the domestic consequences left over from the "Who Lost China" propaganda campaign of the late Forties - Clare Booth Luce and all of Chiang Kai-Shek's friends in DC pushed that one along, making both Democrat and Republican Presidents afraid of looking "weak" in Asia.
Both of us were mystified by the Vietnam War - the US certainly didn't care if the French lost their empire, and there was no real strategic interest that justified the enormous costs of propping up the South Vietnam government. With a little effort, the US could have made Uncle Ho into the Asian version of Tito, skillfully playing the Chinese against the Russians, driving both nuts, and frustrating Chinese plans in the region.
Our best guess was the Red Scare purge of the McCarthy years had removed all of the Old China Hands from the upper levels of the Federal government, thus there was no-one left at State or other parts of the US Govt that had any experience in the region.
Probably more significant, there was all the domestic consequences left over from the "Who Lost China" propaganda campaign of the late Forties - Clare Booth Luce and all of Chiang Kai-Shek's friends in DC pushed that one along, making both Democrat and Republican Presidents afraid of looking "weak" in Asia.
Lynn Olson said:
I've read about this period a lot and your assesment is pretty much right on.
Did you ever read about General Stillwell? He was pro Mao and very anti Chiang Kai-Shek. Roosevelt was starting to listen, but Truman would have none of it. He saw Mao as a patsy of Stalin, and, rightly so, hated Stalin. I think that Mao used Stalin for what he could get, but, as events would show, had no real respect for him.
Hi
Yes, belief desire and fear is the most solid soil of "reality" ( not to speak of human zombie-stupidity ) .
Look at your statement from that perspective and put it upside down once you ( we ) want to change that current reality.
Basically oppose " utopia " against " realism " - and always keep eyes and heart open ...
... but to whom am I talkin about ...
Dan Brown couldn't have done better ( though its no fiction here ) !
Greetings
Michael
Yes, belief desire and fear is the most solid soil of "reality" ( not to speak of human zombie-stupidity ) .
Look at your statement from that perspective and put it upside down once you ( we ) want to change that current reality.
Basically oppose " utopia " against " realism " - and always keep eyes and heart open ...
... but to whom am I talkin about ...
Dan Brown couldn't have done better ( though its no fiction here ) !
Greetings
Michael
Hey John K -
I've been thinking about this part of your post for a few days.
Now that's an interesting thought! Usually we are told not to make a baffle symmetrical, a circle being the worst - the resulting FR is supposed to be very bad. But as you say on axis all points around the circle would be equidistant, they would appear as one source. At least in the time domain. It should sound like a bigger version of the driver itself.
But would those secondary sources cause comb filtering? Would they interfere with each other?
I ran some simulations of a round baffle yesterday in "The Edge". An 8" driver on a 16" diameter round baffle. (Actually a 200mm driver and 400mm baffle).
Yes, there were ripples in the on axis FR, but not big nasty ones. Moving the driver toward the top of the baffle a bit smoothed those out. The mic need to be 2 or 3 meters from the baffle to integrate well.
But off axis was another matter! Move the mic off axis by a meter or 2 and all hell breaks loose. It's the same for other baffle shapes, tho.
This line of thought has got me wanting to try a "Lollipop" baffle. A round baffle above the floor with a thin support. And of course it would have BudP's funny little blocks all around the edges.
Any thoughts?
I've been thinking about this part of your post for a few days.
john k... said:
Now that's an interesting thought! Usually we are told not to make a baffle symmetrical, a circle being the worst - the resulting FR is supposed to be very bad. But as you say on axis all points around the circle would be equidistant, they would appear as one source. At least in the time domain. It should sound like a bigger version of the driver itself.
But would those secondary sources cause comb filtering? Would they interfere with each other?
I ran some simulations of a round baffle yesterday in "The Edge". An 8" driver on a 16" diameter round baffle. (Actually a 200mm driver and 400mm baffle).
Yes, there were ripples in the on axis FR, but not big nasty ones. Moving the driver toward the top of the baffle a bit smoothed those out. The mic need to be 2 or 3 meters from the baffle to integrate well.
But off axis was another matter! Move the mic off axis by a meter or 2 and all hell breaks loose. It's the same for other baffle shapes, tho.
This line of thought has got me wanting to try a "Lollipop" baffle. A round baffle above the floor with a thin support. And of course it would have BudP's funny little blocks all around the edges.
Any thoughts?
Hi John K,
Re your post#1514 explanation >>
Suppose that you have a circular baffle. If there are no slots then the rear wave can be though of as a ring radiator at the baffle edge. On axis this reduces to a single source at the observation point. That is, the delay from all points around the circumference of the baffle to a point on the dipole axis will be the same. <<
This ties in exactly with JohninCR's stated observations much earlier in this thread, where he warned that tilting a baffle also tilts the sound 'stage'.
This would suggest time aligning the drivers wrt the baffle, though of course rear waves would be different then also.
Seems we are always challenged when we move away from the single Fullrange concept.
Cheers ......... Graham.
Re your post#1514 explanation >>
Suppose that you have a circular baffle. If there are no slots then the rear wave can be though of as a ring radiator at the baffle edge. On axis this reduces to a single source at the observation point. That is, the delay from all points around the circumference of the baffle to a point on the dipole axis will be the same. <<
This ties in exactly with JohninCR's stated observations much earlier in this thread, where he warned that tilting a baffle also tilts the sound 'stage'.
This would suggest time aligning the drivers wrt the baffle, though of course rear waves would be different then also.
Seems we are always challenged when we move away from the single Fullrange concept.
Cheers ......... Graham.
panomaniac said:
Except that the "ring" would be inverted polasrity.
panomaniac said:
This is my point. Start with the circular baffle with point sources and the on axis response is exactly what you get with a true dipole (two point sources separated by some distance radiating with inverted polarity). Yes there would be serious comb filtering above the first dipole peak, but that is the characteristic of a dipole and not a consequence of baffle diffraction. Certainly the rear wave is "diffracted” around the baffle edge (just as the front wave is diffracted around the edge to the rear). But on axis this has no impact on the response (for a very thin baffle). If you consider driver directionality the severity of the comb filtering above the first peak will tend to be lessened, depending on baffle size.
Off axis, what happens depends on the baffle shape since the time delays from the baffle edge to the observation point will now vary around the baffle edge. But even then, the resulting response is just the sum of the direct radiation form the front side plus the radiation from the rear source distributed around the baffle edge as a ring radiator. This can be likened to conventional edge diffraction with the edge sources twice as strong for a dipole than for the conventional speaker, but the if you decompose the response at the edge into the incident wave (from the driver) and the diffracted wave you see that the diffracted wave contribution from the front and rear effectively cancel out for thin baffles which is exactly why the strength of the sources around the edge are twice as strong for a dipole. Collectively you can call this whole thing diffraction, but you can also note that the conventional scattering of the wave incident upon the baffle edge from the front source (as well as that from the rear) don't contribute to the response. The reason I make this distinction is because with a conventional speaker we can apply edge treatment to attenuate the contribution from the edge scattered wave but if we do that with a dipole then we are messing with the dipole response. While the 90 degree off axis response will still be null (assuming symmetric front and rear amplitude), on axis the front response will become more and more independent of the rear response. That may be a good thing and is a reasonable objective, but it’s more like trying to emulate an infinite baffle than a dipole. This is the case for both conventional and OB system. But of course, the degree of success will typically decrease with frequency. This problem exacerbated with OBs by the use of large baffles where the driver is fairly omnidirectional at frequencies above the dipole peak. Large baffles seem like a good thing since they increase low frequency sensitivity. Choosing the correct baffle size and shape is a trade off between low frequency sensitivity, driver directionality and how high in the frequency range you want to retain true dipole response.
I'm discussing much this on the web page I'm working on. It's nearly complete but I'll probable need to edit it a little before posting it (or not). The delay is that I also want to discuss some of the trade offs between using large baffle for improved low frequency sensitivity vs smaller baffle for minimizing response problems above the first dipole peak, as I started to here.
The other point I want to make is that I think there need to be a clear cut distinction between dipoles and OB speakers. Most dipoles are OB but very few OB speakers are true dipoles.
What about the following combination for a high sensitivity dipole: Beyma Pro 6 series (high Qts, overlooked easily because in the car section)
http://car.beyma.com/ENGLISH/listaprod.php?id_familia=13
and large Mundorf AMTs (similar sensitivity, also 4 Ohm)
http://www.mundorf.com/english/speakerschassis/Air-Motion-Transformer.pdf
?
http://car.beyma.com/ENGLISH/listaprod.php?id_familia=13
and large Mundorf AMTs (similar sensitivity, also 4 Ohm)
http://www.mundorf.com/english/speakerschassis/Air-Motion-Transformer.pdf
?
panomaniac said:Thanks John K.
Still not sure about the whole thing. Circular baffle = good. Or circular baffle = bad? I'm baffled.
Maybe it would be a good use for some of those old 16" transcription records. Nice and thin and stiff. Has a musical look to it, too.
Ok, bottom line for the application most people seem to want, smoothest response above the initial dipole peak, a circular baffle = bad.
The equivalency to a ring radiator with a single-point radiator in the center is apt - and the radiation from the ring radiator is of course polarity-inverted, time-delayed and not flat in response, unlike the radiation from the center driver.
What bothers me is the extreme sensitivity to angle. It's one thing to look at 0 and 90 degrees off-axis, but both are special cases, and do not correspond to real listening conditions. At the minimum, the listener is always listening a few degrees off-axis, simply because the left and right ears are in different lateral positions, and if the speaker is aimed directly at one ear, the other will be off by a few degrees.
If there are very sharp directional artifacts (a few degrees wide) this will interfere with stereo imaging, since the ear/brain is trying to use interaural & pinna processing to localize the sound. Narrow-width artifacts interfere with this, since left and right ears receive different spectral and time images before head and pinna diffraction and delays occur. This is why "West Coast Sound" 3-way speakers that weren't mirror-imaged (think JBL L100) had such vague and unstable localization.
I've found that speakers with narrow-width directional artifacts have quite poor image quality and odd, hard-to-pin-down colorations. These are evident on measurement as ripples that zip up and down the spectrum as the microphone is slowly swept in an arc across the frontal listening plane.
These artifacts are also visible in the time domain - I really wish there were more published measurements showing a 3D time vs bearing angle display. I'm trying to dig up an old AES paper by the JVC R&D team where they had a terrific visualization that showed what looked like a circular tsunami coming out of the driver - the differences between a conventional horn, a horn-ribbon, a dome tweeter, and a cone tweeter were quite startling and made the point about narrow-width artifacts.
At any rate, I view much of the baffle discussion in terms of antenna design, although we're working with compression instead of transverse waves. A strong goal of this system is a clean time response with smooth, well-behaved impulse response across the frontal arc, and most especially, free of narrow-width wrinkles only a few degrees wide. I am not too picky whether this is achieved with a compression driver with a modern waveguide design or high-quality ribbon tweeters.
These fine-grained impulse vs angle measurements should also reveal what's going on with BudP's EnABL pattern, as well as the various edge treatments under discussion.
I am partitioning the whole question into two parts - fine-grained behaviour over a frontal arc of 45 to 60 degrees (covering the most likely listening positions), and more coarse-grained behaviour into the listening room as a whole. The two parts are:
1) On and near-axis and ...
2) Total power emission into a sphere.
I am not as intererested in each individual room reflection as I am about the collective response of all of them, and that in turn has a lower weighting than the direct-arrival sound, which is the first, primary, most-important priority for me.
I'm aware this is sort of an old-school viewpoint, but well, there are "traditional" aspects to the sound of my speakers, including the overall balance of the sound.
What bothers me is the extreme sensitivity to angle. It's one thing to look at 0 and 90 degrees off-axis, but both are special cases, and do not correspond to real listening conditions. At the minimum, the listener is always listening a few degrees off-axis, simply because the left and right ears are in different lateral positions, and if the speaker is aimed directly at one ear, the other will be off by a few degrees.
If there are very sharp directional artifacts (a few degrees wide) this will interfere with stereo imaging, since the ear/brain is trying to use interaural & pinna processing to localize the sound. Narrow-width artifacts interfere with this, since left and right ears receive different spectral and time images before head and pinna diffraction and delays occur. This is why "West Coast Sound" 3-way speakers that weren't mirror-imaged (think JBL L100) had such vague and unstable localization.
I've found that speakers with narrow-width directional artifacts have quite poor image quality and odd, hard-to-pin-down colorations. These are evident on measurement as ripples that zip up and down the spectrum as the microphone is slowly swept in an arc across the frontal listening plane.
These artifacts are also visible in the time domain - I really wish there were more published measurements showing a 3D time vs bearing angle display. I'm trying to dig up an old AES paper by the JVC R&D team where they had a terrific visualization that showed what looked like a circular tsunami coming out of the driver - the differences between a conventional horn, a horn-ribbon, a dome tweeter, and a cone tweeter were quite startling and made the point about narrow-width artifacts.
At any rate, I view much of the baffle discussion in terms of antenna design, although we're working with compression instead of transverse waves. A strong goal of this system is a clean time response with smooth, well-behaved impulse response across the frontal arc, and most especially, free of narrow-width wrinkles only a few degrees wide. I am not too picky whether this is achieved with a compression driver with a modern waveguide design or high-quality ribbon tweeters.
These fine-grained impulse vs angle measurements should also reveal what's going on with BudP's EnABL pattern, as well as the various edge treatments under discussion.
I am partitioning the whole question into two parts - fine-grained behaviour over a frontal arc of 45 to 60 degrees (covering the most likely listening positions), and more coarse-grained behaviour into the listening room as a whole. The two parts are:
1) On and near-axis and ...
2) Total power emission into a sphere.
I am not as intererested in each individual room reflection as I am about the collective response of all of them, and that in turn has a lower weighting than the direct-arrival sound, which is the first, primary, most-important priority for me.
I'm aware this is sort of an old-school viewpoint, but well, there are "traditional" aspects to the sound of my speakers, including the overall balance of the sound.
Hi
JohnK, are the pic's below the illustration of what you are pointing out?
If so I hardly can't wait your page being finished to have a understandable explanation on that behaviour including the baffle effect.
The pictures above show two omnidirectial sources slightly set apart with the rear one inverted and equal in strength and 1dB / 3dB / 6dB / 12dB attenuated.
This reflects real world OB's due to the motor assembly atttenuation at higher frequencies – except for that there of course directivity starts to come into play with increasing frequency as well (sadly not modelled in ShwoArray ).
Its interresting to see that smooth transition from dipole to omni, no ?
Greetings
Michael
JohnK, are the pic's below the illustration of what you are pointing out?
If so I hardly can't wait your page being finished to have a understandable explanation on that behaviour including the baffle effect.
The pictures above show two omnidirectial sources slightly set apart with the rear one inverted and equal in strength and 1dB / 3dB / 6dB / 12dB attenuated.
This reflects real world OB's due to the motor assembly atttenuation at higher frequencies – except for that there of course directivity starts to come into play with increasing frequency as well (sadly not modelled in ShwoArray ).
Its interresting to see that smooth transition from dipole to omni, no ?
Greetings
Michael
Simple Cousin
Lynn, what about a simple one using Hemp FR8C and say a 15inch Eminence near the floor? Some trapezoid baffle, offset Hemp, some multi drilled edges, and a passive circa 500hz XO? Would not that be also a showcase of the principle? Good for 25sqm spaces and lighter program maybe.
Lynn, what about a simple one using Hemp FR8C and say a 15inch Eminence near the floor? Some trapezoid baffle, offset Hemp, some multi drilled edges, and a passive circa 500hz XO? Would not that be also a showcase of the principle? Good for 25sqm spaces and lighter program maybe.
Re: Simple Cousin
Funny you should say that. That's what I'm working on now. Except with 2x 6.5" long stroke woofers near the floor (per side). Probably a lower x-over point. Small BR box and plate amp.
I might try my Selenium 15s with them, just to hear. OB bass needs a mighty boost, tho.
Just waiting for the drivers to cross the wide Pacific.
salas said:
Funny you should say that. That's what I'm working on now. Except with 2x 6.5" long stroke woofers near the floor (per side). Probably a lower x-over point. Small BR box and plate amp.
I might try my Selenium 15s with them, just to hear. OB bass needs a mighty boost, tho.
Just waiting for the drivers to cross the wide Pacific.
Hi Michael,
Last year I concluded that the best edge for a open baffle was a large tube either side of a driver on a baffle board.
You mention bass efficiency being a problem.
I have tackled this with my own current project by way of a line having internal length of 8ft and folded at floor level, with two drivers at 4ft height for genuine FR working, one forward facing and one rear for bipole too, the forward one having pressure relief holes drilled close to it to linearise its (primary wave) transduction.
If this were turned upside down and made from pipe with an arching curve around the top of a baffle (like the old Wurltizer Jukebox), with a LF driver at floor level, then maybe some stand-alone OB requirements could be solved.
With a tweeter in the very top of curve, a FR just below, the next augmentation LF driver leading into FR territory being below it and thus closer to the floor, then the final bass driver at the pipe end closest to and facing listeners also at floor level, the other end of the pipe venting into the room corner. This would have super LF capabilities to an easy 25Hz.
Now that makes me wonder what I could do with my own cabinet by doing an upside down version with surrounded FR driver in the fold region, for it is pipe length which gains the LF.
Cheers .......... Graham.
Last year I concluded that the best edge for a open baffle was a large tube either side of a driver on a baffle board.
You mention bass efficiency being a problem.
I have tackled this with my own current project by way of a line having internal length of 8ft and folded at floor level, with two drivers at 4ft height for genuine FR working, one forward facing and one rear for bipole too, the forward one having pressure relief holes drilled close to it to linearise its (primary wave) transduction.
If this were turned upside down and made from pipe with an arching curve around the top of a baffle (like the old Wurltizer Jukebox), with a LF driver at floor level, then maybe some stand-alone OB requirements could be solved.
With a tweeter in the very top of curve, a FR just below, the next augmentation LF driver leading into FR territory being below it and thus closer to the floor, then the final bass driver at the pipe end closest to and facing listeners also at floor level, the other end of the pipe venting into the room corner. This would have super LF capabilities to an easy 25Hz.
Now that makes me wonder what I could do with my own cabinet by doing an upside down version with surrounded FR driver in the fold region, for it is pipe length which gains the LF.
Cheers .......... Graham.
Hi Michael,
The plots you posted aren't exactly what I was referrign to. They just show whan happens when the strength of the rear source is reduced. That is very different than what happens is we assume the wave propagation along the baffle surface is attenuated.
After some of the more recentr discusstion I have decided to reformat the web page I was working on. I think there are a lot more important issues regarding the selections of baffle size/shape, etc than just worrying about what happens at the edge and what to call it.
The plots you posted aren't exactly what I was referrign to. They just show whan happens when the strength of the rear source is reduced. That is very different than what happens is we assume the wave propagation along the baffle surface is attenuated.
After some of the more recentr discusstion I have decided to reformat the web page I was working on. I think there are a lot more important issues regarding the selections of baffle size/shape, etc than just worrying about what happens at the edge and what to call it.