John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
What I mean, subconscious reactions that happen before we realize that the sound goes from speakers.
The first time a musician, talking to an other during a pause, had turned his head in direction of the speakers as i called his name from the cabin," i saw that is was good".
I don"t understand why all the others were laughing after that.
have to be hammered hard for a hour or two at least with serious pub rock or equivalent to come to life
Did-you mean two hours of Spinal Tap http://en.wikipedia.org/wiki/This_Is_Spinal_Tap music with your volume at 11 each time you want to listen at " A Little Night Music " ?
 
Wavebourn is on the right track to making 'progress' in our field. Many here are pretty good engineers or at least techs. There are only so many ways to do something, and many paths have been taken already, perhaps many times. It is important to note the 'usefulness' or "success" of each path and in the pursuit of audio quality, choose the 'best' paths.
It is not always easy to take a particular path or cheap to employ, but not ALL paths lead to the 'best' in audio quality. At the same time, there is not ONE path that stands above all others as being the BEST path possible.
So, whether we employ horns, direct radiators, Manger speakers, electrostatics, ribbons, or some other more exotic speaker system, the PROOF is in the listening. IF we can fool ourselves or others, (for example, I was SURE at one time, that Dick Sequerra had a live violinist in his listening room) and THAT is a major point of the 'exercise' of making high quality audio.
Lack of generating 'listener fatigue' is another.
There may be many other 'virtues' we can put in, as well but these, to me, are the two most important in my audio designs.
 
Well said. It so happens that speakers which do well in Blind Listening Tests are also the ones where I've experienced similar things to what you've described and seen it happened to other people too.

Most of my recordings are of live concerts so have cues of this sort which can startle the listener when he realises they are actually on the recording.

In fact, I'll stick my neck out and claim that speakers that don't do well in my Blind Listening Tests will NOT be able to fool the listener as you have described.

Right; and when there are some "cues" that can be "ignored" that means it's not high-end yet, even if it is very close. It is what I meant. Like when I play 11 Hz through subwoofer and increasing signal level start hearing "Rrrrr" that means I hear distortions that are cues to "reproduction".
 
The first time a musician, talking to an other during a pause, had turned his head in direction of the speakers as i called his name from the cabin," i saw that is was good".
I don"t understand why all the others were laughing after that.

Yes, I had similar cases too. :)
Once when Sergey Zadvorny was alive, after one home concert we sat around the table drinking tea with cookies. I switched on the record I made on the concert, we continued conversation, and after some phrase Sergey said on the record his concertmaster turned the head to Sergey at the table and answered. :D
 
Maybe we need to add a switch to audio systems that does what the TV's do; "torch mode" to sell them and "ISF" mode to listen to them.

I respectfully disagree Demian.

When I worked on my home system (I always use myself as a guinea pig), my wife kept asking, "What are you still working on? It already sounds great!". But when later I took her to some demonstrations of known brands in the mall, she understood well what I am working on. Now she says, I spoiled her tastes, and she can't stand the sound from stock sound systems.

One more example, people saw prototypes of my foldable line arrays during last BAF, when I used them for PA with large diaphragm condenser mikes and 31-band EQ. This system, but with prototype of my vacuum tube Pyramid amp, and with my class-A console, was still in my Armada after one Saturday concert when we come to flea market in Walnut Creek one Sunday. I saw Diego Jameau singing there, and told him that I have my system with me. We set it up quickly, and he started singing smiling from ear to ear. Soon he was surrounded by people, including sellers who left their stuff. One of them said her brother is musician and she wanted to buy such system for him. When I said that it is a prototype and how much it costs, her face was sad.

My experience is, people prefer naturally sounding systems from the moment they heard them. It work as if an ear opening. Like popping of corks from ears. Like threshold of expectations change, and what used to be "sounds good" is no good anymore.
 
Lack of generating 'listener fatigue' is another.
Interesting question raised often on the forums.
I'm always very fatigued, walking out of a concert. More the concert was good, more i'm fatigued, like after a night of love (as far as i can remember).
Same thing when it is about listening to music with a good system. More the system is involving, more it catch your attention with details, more it brings fatigue: Transients, high dynamic, high levels, all along with grove and beauty of the played music kills you .
I do not power on very often my big system to listen music, and cannot afford-it for ambiance music in background. A listening session require a very special approach from me, like going to concert, a lot of attention, that can not really exceed much of a couple of hours.
How about you ?
 
Last edited:
diyAudio Member RIP
Joined 2005
Interesting question raised often on the forums.
I'm always very fatigued, walking out of a concert. More the concert was good, more i'm fatigued, like after a night of love.
Same thing when it is about listening to music with a good system. More the system is involving, more it catch your attention with details, more it brings fatigue: Transients, high dynamic, high levels, all along with grove and beauty of the played music kills you .
I do not power on my big system to listen music, and cannot afford-it for ambiance music in background. A listening session require a very special approach from me, like going to concert, a lot of attention, that can not really exceed much of a couple of hours.
How about you ?
Concerts are usually too long, in my opinion.

After a spectacular performance of Also Sprach Zarathustra at Disney Hall, I turned to Floyd Toole and imitated a radio host: "And it looks like we'll have an encore! Ein Heldenleben! :eek:
 
Hi jneutron, all
Ah, I see .
4W /sq/in acoustic power density would be about 158dB I think (you can do acoustic levitation with that intensity). This would be a pressure of about 1588 Pascals out of 101,325 so the pressure would be about .22psi (you can work through the gas law and temperature effect on speed haha).

The more difficult part to quantify is the nonlinearity of air is an accumulating effect. The effect goes up with intensity (because of the larger difference in the instantaneous pressure and rarefied temperatures) AND it goes up the larger the number of wavelengths it travels at that intensity making a slowly expanding horn more problematic.

In a horn, the sound is traveling through an expanding passageway and so the same power spread over a larger area means the pressure falls all the way from the diaphragm to the mouth. In an exponential shape horn, the pressure falls more slowly and so has a greater susceptibility of air distortion, at the opposite extreme, a conical horn has a very rapid expansion but also does not load lower frequency end of the same driver well. The impedance transformation part of horn operation has a “high pass” effect to it based on how fast the horn area expands. So many contradicting issues but a lot of fun to work on.

John is quite correct when he said it is almost impossible to pre-distort properly and even at that, one can only fix some things. With driver distortion, one needs a 3d response model of what it does at every frequency at every level and even then, some aspects of what it does depend on what it was been doing an instant, a moment and minutes before.

With the levitation sources, these were narrow band devices, that was the only way I could make 160dB with 20 Watts so one had little freedom with waveshape, they produced a sine wave, how much what you got was a saw tooth depended on how loud it was and how far away one was..
One can levitate with one source and a reflector but two sources producing an interference pattern allow a higher intensity where the levitation is because it doesn’t have to go to a reflector and bounce back to make an interference pattern.
A cool part not shown in the video is that if you roll the phases of both sources, you can move the samples smoothly up down (axially). At 22KHz, 360 degrees is about ½ inch motion

Regarding absolute phase and the importance of, generally loudspeakers do not preserve the phase of the input signal and spreads harmonics of a single event out in time (negating the idea of an absolute).

The real problem is that we don’t hear like we measure, a measurement microphone can precisely capture the pressure at a given point in space. WE hear through two ears that have a far from flat response, who’s pina response makes in the ear pressure measurements of our ears look broken, big changes depending on the sounds incoming angle and so on. Our magnificent ear / Brain / learned pattern system takes what makes measurements make look “broken” and instead of “broken” that is how we can tell where sound is coming from, how high it is even how far away it is. Our processing system automatically seeks to “hear through” noise and problems to get what we expect and want so we often don’t hear giant problems in a room etc, UNTIL we change something.

If you want to hear how accurate a loudspeaker is, make a generation loss recording with a measurement mic in a semi anechoic condition (outdoors on a quiet day). Like any part of the chain, the more passes it will tolerate before transmogrifying the signal, the more accurate it is.
At work, we used this at work developing our loudspeakers (large scale sound) and it was eye opening many times.. It is surprising how few passes even some good loudspeakers will last, each generation an increasing caricature of what is wrong.
Best,
Tom Danley
Danley Sound Labs
 
Hi jneutron, all
Ah, I see .
4W /sq/in acoustic power density would be about 158dB I think (you can do acoustic levitation with that intensity). This would be a pressure of about 1588 Pascals out of 101,325 so the pressure would be about .22psi (you can work through the gas law and temperature effect on speed haha).

Yup..major stuff.

Gay-Lussac's law states P1/T1 = P2/T2, temp in absolute.

P1 = 101325
P2 = 102913
T1 = 300 K

T2 = (T1 * P2)/P1

304.7 K, 4.7 degree C rise in temp (peak)

at 300 K, sound velocity is 347.202 meters per second
at 304.7 K, velocity is 349.9 meters per second.

A column of air 1 meter long will be traversed in 2.88 milliseconds at 300k
at 304.7, it will be traversed in 2.85 milliseconds.

A shift of 30 microseconds.

A bit over 1 order of magnitude larger than human sensitivity to interaural delays for localization of an image. At the least, it will provide dithering of the hf content, which has also been demonstrated as helping human localization.

Note that 30 is slightly larger than 21. I had to borrow a colleague's fingers to do this math..
jn

ps. Note that I also neglected the fundamental distortion caused by the temperature variation. Let it not be said that I am a nitpicker..
pps. Watch out for the ideal gas law found on the web. In general, they consider the spl as low enough that there is no temperature variation, that is not what this discussion is about.
 
Last edited:
diyAudio Member RIP
Joined 2005
John is quite correct when he said it is almost impossible to pre-distort properly and even at that, one can only fix some things. With driver distortion, one needs a 3d response model of what it does at every frequency at every level and even then, some aspects of what it does depend on what it was been doing an instant, a moment and minutes before.

With the levitation sources, these were narrow band devices, that was the only way I could make 160dB with 20 Watts so one had little freedom with waveshape, they produced a sine wave, how much what you got was a saw tooth depended on how loud it was and how far away one was..
One can levitate with one source and a reflector but two sources producing an interference pattern allow a higher intensity where the levitation is because it doesn’t have to go to a reflector and bounce back to make an interference pattern.

A cool part not shown in the video is that if you roll the phases of both sources, you can move the samples smoothly up down (axially). At 22KHz, 360 degrees is about ½ inch motion

Best,
Tom Danley
Danley Sound Labs

By the by, I was only proposing predistortion as applied to the narrowband levitation problem, to alleviate the sawtooth behavior, emphatically NOT for wideband audio use.
 
I'm not quite sure what all the fuss is about the horn throat region of the horn and compression driver combination is. I see the conversation about temperature rise in the throat due to frequency and volume area but this area would seem to be the least of the problem in this regards. The initial loading of the majority of compression drivers, the rarification area lets call it, from the diaphragm to the initial slot opening is on the order of 10 to 1. This would be the area of highest temperature rise and also to any real perturbations that would affect the frequency response curves. After this point in the combination anything else is of an order of magnitude lower than the initial compression area. So why are we focusing on this area rather than the compression driver itself?

I agree with the statement that a speaker that fools you into thinking that something actually just happened in a room is a definitive plus in whether a loudspeaker combination is going to sound true to the original source. I have found myself more than once turning around to look for someone saying something on a recording before realizing that it was in the recorded playback. When something is so real sounding that it can make you jump up out of your seat we have hit the main goal of sound reproduction. How many times I have had one of my kids or a friend say I have never heard that before on a track they have heard many times before. That is also one of the things I look for. Articulation and clarity goes hand in hand with creating that magic we all call music.
 
I'm not quite sure what all the fuss is about the horn throat region of the horn and compression driver combination is. I see the conversation about temperature rise in the throat due to frequency and volume area but this area would seem to be the least of the problem in this regards. The initial loading of the majority of compression drivers, the rarification area lets call it, from the diaphragm to the initial slot opening is on the order of 10 to 1. This would be the area of highest temperature rise and also to any real perturbations that would affect the frequency response curves. After this point in the combination anything else is of an order of magnitude lower than the initial compression area. So why are we focusing on this area rather than the compression driver itself?
The diaphram to slot ratio problem is not quite the same (however I certainly acknowledge the issue you raise as an important one. I'm speaking only of the energy density modulation of the velocity along a long throat.

At the slot, there are a few more issues. The impedance discontinuity is just downright nasty. Higher temperature rise would be defined by the orifice edges and reflection at the impedance step. The bulk of the cavity will not have higher temperatures than the throat at the slot, but I'm sure peaks will.

Might as well start the process of filing the intake and exhaust contours. Honestly, I never inspected my horns or drivers for that. Didn't matter to me.

Needless to say, I'm not gonna get any kind of numbers using ten fingers and ten toes to properly discuss slot to diaphram ratio loading and contouring..

jn
 
Status
Not open for further replies.