SY,
So some places are bigger than others! That system can do 102 db "A" slow at 800 feet. But only really makes it to 8,000 hertz due to the air losses.
A voltage source amplifier with good damping factor will have some test components that do follow the music. It is the sharp jumps and dips that are the surprising results. Those are things you will hear. Try it sometime. It is an easy set up and although many amplifiers behave as reasonably expected some have quite a few surprises.
EAW KF900s were the product in question. But many pro HF drivers will actually take that kind of power WHEN PROPERLY BAND LIMITED!
One glaring flaw in our traditional measurements is that we measure at the wrong levels - "full power" testing is left over from a bygone era when amps' nonlinearities could reasonably be assumed to be monotonic. For modern amplifiers, with all their undoubted virtues, this assumption shouldn't be made automatically, it should have to be proven, or the level of imperfection measured.
A good start at an appropriate measuring level would be to define a "0VU", or maybe to pick a range of suggested 0VU's based on typical speakers' sensitivities, SPL requirements and rooms, and to specify performance at -60VU, 0VU, +20VU, etc.
The -60VU performance is especially interesting, so needs to be treated as important even though noisy.
Thanks,
Chris
A good start at an appropriate measuring level would be to define a "0VU", or maybe to pick a range of suggested 0VU's based on typical speakers' sensitivities, SPL requirements and rooms, and to specify performance at -60VU, 0VU, +20VU, etc.
The -60VU performance is especially interesting, so needs to be treated as important even though noisy.
Thanks,
Chris
To answer the first question, no. I was the only one in the room and had gone in there expecting the same awful sound as before. And no, I didn't notice the amps until after I had listened for awhile because I don't notice **** like that.
Speaking of hand-waving, head-nodding and peer pressure (the effect of which are all highly over-rated) a friend of mine relayed to me an incident which happened in the Magico room one year. The principle of the company introduced his product with much hand-waving and proclaimed a new era in audio. When he played the system my friend remarked out loud that the vocals had a hardness to them that sounded unnatural. Most of the others in the room nodded in agreement. In between tracks, the guy approached my friend and whispered in his ear, "I want you to leave this room. Now."
Anyone who attends these audio shows on a regular basis should know that there is a great deal of difference of opinion, personal preferences differ widely, and no one I know thinks that the "experts" are of any concern.
Furthermore, measurable distortion is hardly ever audible, especially in amplifying devices.
John
Great story!
During a "listening test" a few years ago, which involved people raising hands to indicate preferences, I was sitting with John Broskie. John, no dummy for sure, wryly noted the spatial and temporal pattern of the hand raises. Don't underestimate group dynamics!
Hi,
A general question. There is much debate here about the levels of harmonic distortion in Amplifiers.
As Sy so astutely observed, Amplifiers do not make any sounds (unless bad transformers are used).
Speakers make sounds.
Now, here is an offer of glory for anyone of those who wish to illustrate that traditional distortion measurements are relevant illustration of the capabilities of an Amplifier used for music reproduction.
One wonders if anyone would care to estimate the distortion of a "HiFi" speaker (let us be generous and allow 100 Liter Volume, 12" Woofer and a 3-Way system) at (say) full rated power of say 100W which we may call "analogue full scale" and at levels of -20dB (1W), -40dB (10mW) and -60dB (100uW) below "analogue full scale" for 20Hz, 50Hz, 100Hz, 1KHz, 5KHz and 20KHz...
Ciao T
A general question. There is much debate here about the levels of harmonic distortion in Amplifiers.
As Sy so astutely observed, Amplifiers do not make any sounds (unless bad transformers are used).
Speakers make sounds.
Now, here is an offer of glory for anyone of those who wish to illustrate that traditional distortion measurements are relevant illustration of the capabilities of an Amplifier used for music reproduction.
One wonders if anyone would care to estimate the distortion of a "HiFi" speaker (let us be generous and allow 100 Liter Volume, 12" Woofer and a 3-Way system) at (say) full rated power of say 100W which we may call "analogue full scale" and at levels of -20dB (1W), -40dB (10mW) and -60dB (100uW) below "analogue full scale" for 20Hz, 50Hz, 100Hz, 1KHz, 5KHz and 20KHz...
Ciao T
Hi,
Yes, this kind of thing happens a lot.
Whenever there is someone who has gained an audience and wishes to convert this audience to his way of seeing things (no matter how true or real or not), one finds that dissenting voices, no matter how accurate the findings they convey, are distinctly unwelcome and get hassled (at least) by those doing the conversion talking or their enforcers, regardless of topic, veracity of presentation and any of the like.
Of course, the end result is a monologue instead of dialogue. Clearly such suits some people extremely well...
Ciao T
Yes, this kind of thing happens a lot.
Whenever there is someone who has gained an audience and wishes to convert this audience to his way of seeing things (no matter how true or real or not), one finds that dissenting voices, no matter how accurate the findings they convey, are distinctly unwelcome and get hassled (at least) by those doing the conversion talking or their enforcers, regardless of topic, veracity of presentation and any of the like.
Of course, the end result is a monologue instead of dialogue. Clearly such suits some people extremely well...
Ciao T
Then you disagree with Nelson's article, at least the part dealing with customer preference? How would one prove that the distortion spectra is not the difference? The statement that the Boulder in and of itself sounds awful has at least one bit of contradictory evidence here, a least as anecdotally powerful as any other. Totally removing the cables, speaker, room, etc. from the picture would seem hard to do exhaustively. If you are satisfied with your results fine. If this means all amplifiers with GNF have an inherently flawed sound do to something no one knows how to quantify, you will have to work harder.
BTW I was at a Sony/Philips mini-disk rollout and thought it sounded like do-do, there were plenty of audio pros and super models there trying to convince me otherwise.
Last edited:
He is right, but there is absolutely no problem to make the "first watt" absolutely clean, speaking about non-linear distortion. And again, even at low power levels the speaker will distort 30 dB and more above good amplifier, at any distortion frequency component.
Everything I have to say about audio equipment is anecdotal, which naturally results in about half of what I have to offer ending up being contradictory.
If I had a nickel for every time I ended up eating crow, I could afford chicken.
John
Followed by the ice cream of course.
Here is a really lousy image I got from JBL. The distortion levels are raised by 20 db to make it easier to see them. Second and third are shown. They are around 45 db down from the test level of 2.83 volts input, nominal 1 watt. Measurement is at 1 meter. Level is calibrated or around 96 db spl.
Distortion is down by about 45 db re power. As amplifier distortion is often given as % of output voltage, be sure to include that! (.00316% voltage distortion would be -45 db power distortion.)
Distortion is down by about 45 db re power. As amplifier distortion is often given as % of output voltage, be sure to include that! (.00316% voltage distortion would be -45 db power distortion.)
Attachments
This test is a good one, but it is VERY important to measure the small-signal frequency response of each path first. If the damping factor of the amplifier allows the speaker load to influence frequency response by much, then the resulting difference seen in this test may just be due to frequency response as opposed to some kind of nonlinear behavior.
Although frequency response is important, it is very important to be able to distinguish between listening results that are due to frequency response differences and those that are influenced by other differences. I wish reviewers would measure in-situ system frequency response AT THE LOUDSPEAKER TERMINALS after their review.
Cheers,
Bob
Amp "sound"
(dons Kevlar suit)
Although oft-mentioned, I believe this is probably the most under-appreciated characteristic of an amplifier contributing to it's "sound."
In the early 1990s I was heavily involved in dragging car audio from being merely SPL oriented to considering sound quality as part of IASCA. The one thing you can say with absolute assurance is that a mobile sound system (excepting multi-kilowatt ones) is regularly clipping. The behavior of an amp while clipping therefore becomes a large determining factor in the overall sound quality. We were able to dramatically demonstrate this by rigging up a car with "over-easy" limiters. When switched in and out with the gain set to otherwise cause clipping, the difference was dramatic. The overall volume barely changed, but most distortion disappeared.
At any other than background listening levels, this often applies to home listening as well. If the small-signal characteristics are within generally accepted quality standards, I contend it is the large-signal and clipping behavior which lend amplifiers their "sonic signatures."
With my test equipment I can find no other good explanation why many tube and MOSFET amps have a similar "ease" when pushed hard. I dislike the high output impedance and resulting speaker frequency response modification caused by almost all tube amps, but like their sound otherwise. It has therefore been my personal choice to listen to MOSFET amps at home (old B&K at present).
Another $0.02 worth...
Howie
Howard Hoyt
CE - WXYC-FM
UNC Chapel Hill, NC
www.wxyc.org
1st on the internet
(dons Kevlar suit)
Although oft-mentioned, I believe this is probably the most under-appreciated characteristic of an amplifier contributing to it's "sound."
In the early 1990s I was heavily involved in dragging car audio from being merely SPL oriented to considering sound quality as part of IASCA. The one thing you can say with absolute assurance is that a mobile sound system (excepting multi-kilowatt ones) is regularly clipping. The behavior of an amp while clipping therefore becomes a large determining factor in the overall sound quality. We were able to dramatically demonstrate this by rigging up a car with "over-easy" limiters. When switched in and out with the gain set to otherwise cause clipping, the difference was dramatic. The overall volume barely changed, but most distortion disappeared.
At any other than background listening levels, this often applies to home listening as well. If the small-signal characteristics are within generally accepted quality standards, I contend it is the large-signal and clipping behavior which lend amplifiers their "sonic signatures."
With my test equipment I can find no other good explanation why many tube and MOSFET amps have a similar "ease" when pushed hard. I dislike the high output impedance and resulting speaker frequency response modification caused by almost all tube amps, but like their sound otherwise. It has therefore been my personal choice to listen to MOSFET amps at home (old B&K at present).
Another $0.02 worth...
Howie
Howard Hoyt
CE - WXYC-FM
UNC Chapel Hill, NC
www.wxyc.org
1st on the internet
Ed, these calculations don't look right.
By -45dB power distortion, do you mean 1 watt / 178 = 0.06 watts ?
Often, but not always. I'm running a tube amp that gives me 10dB headroom on the peaks when the average SPL at the listening position is 85dB. Plenty loud, never clips. I'd bet a number of guys in this thread have as much, or more headroom.
But I have run amps and speakers that clipped regularly. How that clipping is handled definitely make a difference as to how it sounds, no doubt.
Some amps will really hurt you when they clip, others not.
In that vein, I consider it actually a bit outdated to look at amplifier and loudspeaker driver as separate systems. Much can be gained by looking at the amplifier and the driver as an integrated system, with measurements to be applied to the whole system, not just the parts. Multiples needed for more-way loudspeakers.
I is a major source of confusion to me that high end still seems to be split in loudspeaker designers and amp designers, each doing their own thing. With the cable manufacturers in between.
It is even better to measure the actual frequency response of the whole chain, rather than just at the loudspeaker terminals. That is not where the sound comes from.
Two optimized subsystems -amp and loudspeaker - may work fine together, but there is no guarantee, and this may be the source of much real life disappointment.
Last edited:
I've done that and more (distortion) and mentioned it recently in this very thread. There was not one comment on it. Not even one. That says a lot - to me.
- Status
- Not open for further replies.