Yes to all that except if you make your own recordings, YOU choose how the peaks are clipped or not.
Probably the most famous recording made with one of my mikes is Mike Skeet's Garage Door. It took several attempts. The Mk4 Soundfield is clipped as is the Sony PCM-F1. If played at the correct level, it will clip a 1000W amp into any speakers you care to name.
But in spite of its headbanging nature, it is in fact a demonstration of stereo. On speakers with accurate imaging, you almost see the door shimmer in front of you when he unlatches it.
Things have moved on since then and today it is indeed possible to record a dynamic range of about 20b. (For a true 24b converter, ask Guru Wurcer of his experience with probably the only one ever made). Unfortunately, modern 'record producers'(??!!) will insist the final product has a dynamic range of 6dB or less.
Playback is another matter and we really haven't moved that far forward since the early 80's. eg hardly anyone takes advantage of the big gains available when you integrate speakers & amps. Even Genelec & Harman efforts are crude.
I'd be interested to hear comments from anyone who is working or has worked on supa dupa powered speakers.
Very true. A current project of mine is to get a budget pro monitor to actually work properly, mentioned in another thread. When looking at items that could be sup'd up I was quite amazed at how hopeless even the premium brands were - more expensive Dynaudio and Mackie were quite feeble, a big surprise ...
Frank
So true. They want their record to sound loud as well in radio, cars at little level etc...
And in the movie business, it is worse. The 0dB reference is at -18dBF. Tv services of verification (in France) look at the tapes with a peak meter (3ms) and do not allow more than+3dB from the reference. Means that more than 10dB of headroom is never used, means that you have to limit all your mixes to a stupid little dynamic and cut down all the short peaks.
Our time is stupid.
We fight for 24 Bits (144dB), and limit our commercial products to 14B (84dB).
We produces some of our cars for > 200km/h, and limit the allowed speed to 130km/h with cops and radars everywhere.
We throw directly to the garbage > 30% of the produced food in industrial countries , while 14% of the world's population suffers of salvation.
I'm divorced because my wife used to throw away my yogurts at midnight, looking at the expiration term dates. (smiley, here ?)
[edit]BTW, may-be it was because my own expiration term date ?
Last edited:
Hi KGRlee,
I am both surprised but also not surprised someone had done generation loss recordings with loudspeakers previously, the technique was common place in the days of evaluating analogue recording tape but with loudspeakers performing so poorly I figured that was why it never caught on with speakers.
When we were developing a single point CD horn, we started making the recordings with a 24/96 recorder because the results sounded different than normal and I figured this might be a reality check.
One interesting part was that some speakers sounded bad just hearing them through a measurement mic and good headphones, by two generations, most are unlistenable; most of remaining speakers have pretty noticeable flaws. When you hear and “learn” that “amplified” acoustic signature, often you can also then hear it when you listen to the speaker playing live.
We did them both in room and outdoors, outdoors, we did them on a tower about 15 feet in the air, the mic was a meter away and the speaker was driven at an average level of 80dB. The speaker was driven with a music recording, several snips of different tracks which was played through the 24/96 recorder that recorded the mic signal. Like normal gen loss tests, the mic signal that was recorded was then played back through the speaker and the new mic signal recorded and so on. A few times, we also recorded and played the direct music track but the loss was so small it usully wasn’t worth doing.
Out doors with no close reflections you hear / measure the loudspeaker itself, when you place it in a room, you add the room and now the system’s directivity and placement in the room plays a VERY strong role in governing how fast everything falls apart.
While we don’t do the recordings that often now days, when I really want to “hear” a speakers flaws via gen loss recordings or take unquestionable semi anechoic to below the bass crossover, I still do these on a tower.
Since a good sound card can be an amazing way to record at home and many people have measurement microphones, it is weird / sad that more people don’t hook them up. One can make very convincing recordings with ONE measurement mic (it’s capturing / reproducing stereo that hugely complicates things). The best part (I prefer good headphones for this) is that you capture sounds around your house that you are totally familiar with and at times they can make the hair stand on the back of your neck (if your also prone to that kind of response).
I used to have some recordings on our old web site using a stereo capture array thing I am working on. Nothing exciting just stuff from around the yard like a train, motorcycle and fireworks. It’s still kind of cumbersome to move around so these were “organic” sounds haha. I hope they will eventually put them back up when the webtoids are working on it, I was pleased with the “space” they captured.
A video I can link and maybe I can get some suggestions on how to describe what this customer hears. He is a fellow in Denmark that took a video of two pairs of speakers he owns, one is coherent single point source horn, the other is a horn / cone system with DSP. These have a similar measured frequency response but they are not the same so far as how they radiate as an acoustic source. How would one describe that?
Danley Sound Labs SM80 vs. d&b Q7 - YouTube
Hi Kindhornman
There is no “fuss” about temperature, I was explaining the mechanism behind the changing waveshape at high intensities also known as “throat distortion”.
Gpapag has a nice link to another view of what is happening.
Jneutron, you are geometrically proper haha.
Regarding the compression driver back volume, you make an interesting point.
The first order view of the back volume is that as seen by the motor, it is a compliance (spring) in parallel with the drivers suspension. Those springs are in parallel with the moving mass of the driver which reflected back through the motor looks like a capacitor and the springs, an inductor.
Now, you have a parallel tank circuit and the total mechanical losses appear as a parallel resistor, Rm, of a high value.
This tank circuit is driven from a voltage source by a series R (ignoring a tiny series Le), the drivers Rdc.
With this circuit, one sees a single large impedance peak. With a magic wand, we attach a perfect horn to one side of the radiator and that places a new and much smaller resistance across the moving system, in parallel with the L,C and Rm. For a bass or mid horn where you are in control, the back volume is reduced to raise the Fb up to somewhere well above the low corner (and skipping over reactance annulling).
In a perfect world, when a perfect horn is 50% efficient, that added acoustic load is a resistance equal to the motor Rdc placed in parallel with the Rm,L,C, then half the power is lost to VC heating and half is radiated as sound. From here, remembering the unloaded impedance curve, if you increase the motor strength (BL^2/Rdc), you increase the bandwidth where the radiator is free to move, it’s mobility. Above and below the mobility curve, the radiators motion is constrained either by the spring compliance down low or the moving mass up high (the parallel L or C).
That constraint is why the acoustic power from any compression driver falls above a few Khz, it is the effect of the moving mass that causes the response to fall first. What is called “CD” compensation is really a misnomer because this corrects the response of the power roll off of the driver when you have a horn with the same radiation angle over a wide bandwidth.
Anyway, the air nonlinearity is mostly a 2nd harmonic, when you add two system in push pull, like tubes or woofer, you reduce the 2nd harmonic through cancelation. On the other hand, the pressure in he rear chamber of an hf driver would be VASTLY lower than on the phase plug side BUT for a some driver where the compression ratio and pressure is much less, the rear volume may be a useful thing by also reducing the 2nd H. On the other other hand (making it three) for a low frequency horn, the pressures are so much less that the distortion is essentially entirely from the driver non-linearity, not an acoustic issue.
With the Servodrive woofers I worked on in the 80’s and 90’s, that transducer has no Xmax, no motor position related nonlinearity and with them, you could drive four with a sine wave at rated power with less than 3% max THD.
Wavebourn, you are correct, there isn’t a hifi spec for “maximum usable loudness” and nearly always any spec of “max loudness” is calculated, with some seemingly based on a lightning strike directed to the input terminals..
At work, we have an independent lab take some of our data like full spherical radiation data (for commercial sound) but they also do a “bad sound guy” test for us. Here the speaker is driver with a music spectrum shaped pink noise with a 6dB peak to average ratio. The response is measured at 1Watt and then the power is raised (I think he uses 2dB) incrementally, every 5 min the power is increased. At the power level where the response has changed anywhere by 3 dB relative to the 1W response, that is the “maximum usable loudness”. It is a harsh test but it makes sense.
Anyway, here is that company, Pat Brown is also the fellow who runs Synaudcon.
Electro-Acoustic Testing Company
Best,
Tom Danley
I am both surprised but also not surprised someone had done generation loss recordings with loudspeakers previously, the technique was common place in the days of evaluating analogue recording tape but with loudspeakers performing so poorly I figured that was why it never caught on with speakers.
When we were developing a single point CD horn, we started making the recordings with a 24/96 recorder because the results sounded different than normal and I figured this might be a reality check.
One interesting part was that some speakers sounded bad just hearing them through a measurement mic and good headphones, by two generations, most are unlistenable; most of remaining speakers have pretty noticeable flaws. When you hear and “learn” that “amplified” acoustic signature, often you can also then hear it when you listen to the speaker playing live.
We did them both in room and outdoors, outdoors, we did them on a tower about 15 feet in the air, the mic was a meter away and the speaker was driven at an average level of 80dB. The speaker was driven with a music recording, several snips of different tracks which was played through the 24/96 recorder that recorded the mic signal. Like normal gen loss tests, the mic signal that was recorded was then played back through the speaker and the new mic signal recorded and so on. A few times, we also recorded and played the direct music track but the loss was so small it usully wasn’t worth doing.
Out doors with no close reflections you hear / measure the loudspeaker itself, when you place it in a room, you add the room and now the system’s directivity and placement in the room plays a VERY strong role in governing how fast everything falls apart.
While we don’t do the recordings that often now days, when I really want to “hear” a speakers flaws via gen loss recordings or take unquestionable semi anechoic to below the bass crossover, I still do these on a tower.
Since a good sound card can be an amazing way to record at home and many people have measurement microphones, it is weird / sad that more people don’t hook them up. One can make very convincing recordings with ONE measurement mic (it’s capturing / reproducing stereo that hugely complicates things). The best part (I prefer good headphones for this) is that you capture sounds around your house that you are totally familiar with and at times they can make the hair stand on the back of your neck (if your also prone to that kind of response).
I used to have some recordings on our old web site using a stereo capture array thing I am working on. Nothing exciting just stuff from around the yard like a train, motorcycle and fireworks. It’s still kind of cumbersome to move around so these were “organic” sounds haha. I hope they will eventually put them back up when the webtoids are working on it, I was pleased with the “space” they captured.
A video I can link and maybe I can get some suggestions on how to describe what this customer hears. He is a fellow in Denmark that took a video of two pairs of speakers he owns, one is coherent single point source horn, the other is a horn / cone system with DSP. These have a similar measured frequency response but they are not the same so far as how they radiate as an acoustic source. How would one describe that?
Danley Sound Labs SM80 vs. d&b Q7 - YouTube
Hi Kindhornman
There is no “fuss” about temperature, I was explaining the mechanism behind the changing waveshape at high intensities also known as “throat distortion”.
Gpapag has a nice link to another view of what is happening.
Jneutron, you are geometrically proper haha.
Regarding the compression driver back volume, you make an interesting point.
The first order view of the back volume is that as seen by the motor, it is a compliance (spring) in parallel with the drivers suspension. Those springs are in parallel with the moving mass of the driver which reflected back through the motor looks like a capacitor and the springs, an inductor.
Now, you have a parallel tank circuit and the total mechanical losses appear as a parallel resistor, Rm, of a high value.
This tank circuit is driven from a voltage source by a series R (ignoring a tiny series Le), the drivers Rdc.
With this circuit, one sees a single large impedance peak. With a magic wand, we attach a perfect horn to one side of the radiator and that places a new and much smaller resistance across the moving system, in parallel with the L,C and Rm. For a bass or mid horn where you are in control, the back volume is reduced to raise the Fb up to somewhere well above the low corner (and skipping over reactance annulling).
In a perfect world, when a perfect horn is 50% efficient, that added acoustic load is a resistance equal to the motor Rdc placed in parallel with the Rm,L,C, then half the power is lost to VC heating and half is radiated as sound. From here, remembering the unloaded impedance curve, if you increase the motor strength (BL^2/Rdc), you increase the bandwidth where the radiator is free to move, it’s mobility. Above and below the mobility curve, the radiators motion is constrained either by the spring compliance down low or the moving mass up high (the parallel L or C).
That constraint is why the acoustic power from any compression driver falls above a few Khz, it is the effect of the moving mass that causes the response to fall first. What is called “CD” compensation is really a misnomer because this corrects the response of the power roll off of the driver when you have a horn with the same radiation angle over a wide bandwidth.
Anyway, the air nonlinearity is mostly a 2nd harmonic, when you add two system in push pull, like tubes or woofer, you reduce the 2nd harmonic through cancelation. On the other hand, the pressure in he rear chamber of an hf driver would be VASTLY lower than on the phase plug side BUT for a some driver where the compression ratio and pressure is much less, the rear volume may be a useful thing by also reducing the 2nd H. On the other other hand (making it three) for a low frequency horn, the pressures are so much less that the distortion is essentially entirely from the driver non-linearity, not an acoustic issue.
With the Servodrive woofers I worked on in the 80’s and 90’s, that transducer has no Xmax, no motor position related nonlinearity and with them, you could drive four with a sine wave at rated power with less than 3% max THD.
Wavebourn, you are correct, there isn’t a hifi spec for “maximum usable loudness” and nearly always any spec of “max loudness” is calculated, with some seemingly based on a lightning strike directed to the input terminals..
At work, we have an independent lab take some of our data like full spherical radiation data (for commercial sound) but they also do a “bad sound guy” test for us. Here the speaker is driver with a music spectrum shaped pink noise with a 6dB peak to average ratio. The response is measured at 1Watt and then the power is raised (I think he uses 2dB) incrementally, every 5 min the power is increased. At the power level where the response has changed anywhere by 3 dB relative to the 1W response, that is the “maximum usable loudness”. It is a harsh test but it makes sense.
Anyway, here is that company, Pat Brown is also the fellow who runs Synaudcon.
Electro-Acoustic Testing Company
Best,
Tom Danley
It was being done in 1970 by some MIT students who were also involved in some of the early Cambridge area audio companies.
And of course there's Alvin Lucier
I Am Sitting in a Room - Wikipedia, the free encyclopedia
Tom, you are 'a breath of fresh air' and your speaker sounds pretty good, too.
I have stayed out of a lot of the 'conversation' although there were 'tidbits' of new info, even for me, being passed around.
Mostly, I see mostly amateur opinions, abounding, some 'self-serving' like: My listening set-up is great and I made it myself', others perhaps sort of 'clueless' as to what is going on.
For example: Horn throat distortion, which I started here, with a lot of 'back-talk' and relatively little ENGINEERING input, even though it has been discussed in acoustics textbooks for more than 50 years, probably closer to 100 years. There are equations predicting it, and the amount of 2nd harmonic distortion produced can be SIGNIFICANT, if you are using horns for reenforcing sound in large spaces. Even the JBL example that I put up was tested (I am fairly sure) at about 0.25W INPUT POWER to the horn, not 25W or so that the horn can actually 'handle'. This was a shock to us, 40 years ago, BUT it fits the ENGINEERING EQUATIONS both in Harry Olsen's textbook and Beranek's textbook, from the 50's. There is no need to go deeply into the 'laws of physics', here.
Measurement or calculation from the ENGINEERING EQUATIONS gives the answers. This is what engineers do, rather than physicists, who want to look deeper into 'first principles' and not finish anything tangible.
A hint on reading the JBL app note:
First look at the TESTED SPL at 1 meter (107dB). Then note that these horns usually do 114dB with one watt input. Then note the distortion and see that it INCREASES with frequency significantly.
Now IMAGINE what the distortion would be with 20W in each horn. Enough to scare you or put you off? It did that to us, 40 years ago. And guess what, after our (and who knows how many other people's) complaints, they REDESIGNED the horn throat to lower the distortion, at least a bit.
In the last 2 days, I talked to two of my associates who helped design and run the GD PA system those 40 years ago, and we all laughed at the response we got from JBL at the time. They did NOT appreciate our input, then. Later, they discovered a way to improve things. That is the way of things. '-)
I have stayed out of a lot of the 'conversation' although there were 'tidbits' of new info, even for me, being passed around.
Mostly, I see mostly amateur opinions, abounding, some 'self-serving' like: My listening set-up is great and I made it myself', others perhaps sort of 'clueless' as to what is going on.
For example: Horn throat distortion, which I started here, with a lot of 'back-talk' and relatively little ENGINEERING input, even though it has been discussed in acoustics textbooks for more than 50 years, probably closer to 100 years. There are equations predicting it, and the amount of 2nd harmonic distortion produced can be SIGNIFICANT, if you are using horns for reenforcing sound in large spaces. Even the JBL example that I put up was tested (I am fairly sure) at about 0.25W INPUT POWER to the horn, not 25W or so that the horn can actually 'handle'. This was a shock to us, 40 years ago, BUT it fits the ENGINEERING EQUATIONS both in Harry Olsen's textbook and Beranek's textbook, from the 50's. There is no need to go deeply into the 'laws of physics', here.
Measurement or calculation from the ENGINEERING EQUATIONS gives the answers. This is what engineers do, rather than physicists, who want to look deeper into 'first principles' and not finish anything tangible.
A hint on reading the JBL app note:
First look at the TESTED SPL at 1 meter (107dB). Then note that these horns usually do 114dB with one watt input. Then note the distortion and see that it INCREASES with frequency significantly.
Now IMAGINE what the distortion would be with 20W in each horn. Enough to scare you or put you off? It did that to us, 40 years ago. And guess what, after our (and who knows how many other people's) complaints, they REDESIGNED the horn throat to lower the distortion, at least a bit.
In the last 2 days, I talked to two of my associates who helped design and run the GD PA system those 40 years ago, and we all laughed at the response we got from JBL at the time. They did NOT appreciate our input, then. Later, they discovered a way to improve things. That is the way of things. '-)
Sorry, had ta nitpick...nature of the beast..
I program and operate an 11 axis motion control platform, 7 in locked interpolated spline, 3 of them also further "depthed" in a mapped pseudo 3D space to derive two additional motions in lieu of hardware. So I'm more inclined to spot the mundane points and make a big deal about nuttin..
Actually, I'm just the messenger.. Cohen was the source.
Nice info, thanks..
jn
ps..
Aha. A Niven fan.. moties from the coal sack..
Last edited:
Very interesting video, where we can notice several things: How the acoustic level can change our perception in a closed room, how a non-blind test can help us to focus on our listening, how important are the boxes resonances, how different recordings can reveal differently any problem in an loudspeaker assembly.
Just a question, what the difference if you make the same test with a 110° couple of mikes (ORTF), or an artificial head, in order to lower the room standing waves ?
Last edited:
The way of things is, they exploited design and fashion that J.B.Lansing made common back in 1930'th when he manufactured speakers for moving pictures. He got it in turn from gramophones that were already common then so nobody questioned the decision.
It is inertia John.
They are still alive. How much of their 1970's competitors all over the world ?
Last edited:
There is also the possibility that they did not have the talent to understand what it was JC was telling them, understand it, model it, and fix it.
jn
While I was at Harman/JBL I saw firsthand the pushing aside of some really smart engineers. It was just like school clique stuff.
Or may be they had a talent that understood that for continuing the business it is better not to stop production that is still in demand for the majority of potential customers in order to satisfy one of them.
Happily, musical power level contents decrease with frequency, in the same time.
Last edited:
Statistically. But we don't listen to a pink noise, do we? While sales people on JBL don't sell stuff to individuals, they sell for masses and measure success statistically.
About JBL, why to be crossed about them ? They produced quality pro speakers, well manufactured. And even if i don't liked the sound of their 70's monitors enclosure, i have to admit they allowed to produce very good mixes, at the end. For a freelance sound engineer, it is easter to work, form a studio to an other, with the same reference. In post prod movie auditoria, in France, we use to even tune the volume to a standard same SPL for 0dB. All that explain this inertia you are referring to.
About Horn, they are only a very effective way to increase efficiency, as well to control directivity. We don't have to confuse our comments about horns with comments about the driver behind (compression Vs cone).
Adding a (good) horn to any speaker just reduce distortion because it will run at less power for the same SPL. And better damping.
About Horn, they are only a very effective way to increase efficiency, as well to control directivity. We don't have to confuse our comments about horns with comments about the driver behind (compression Vs cone).
Adding a (good) horn to any speaker just reduce distortion because it will run at less power for the same SPL. And better damping.
Last edited:
- Status
- Not open for further replies.