Opening up that can of worms, again

[Steve F]

First-Arrival vs. Power Response

Let’s revisit—just for fun—a decades-long, still-unresolved controversy among speaker designers:

The question of first-arrival sound vs. far-field power response.

Boiled down into simple terms, the first-arrival crowd feels that the on-axis, first-arrival “anechoic” frequency response is what the ear grabs onto first and that sound is the primary determinant of a speaker’s essential sound/tonal quality. They feel that the room reflections, driver interaction, phase considerations, etc—however closely in milliseconds those reflections and artifacts may follow the ‘first arrival’—are separated out by the brain/ear system and what you fixate on is that first arrival.

Balderdash, says the far-field power response crowd. The room reflections etc are so close in time to the first arrival that they swamp the first arrival completely, leaving the overall far-field energy output of the speaker—the so-called “power response”—as what you hear to determine the tonal quality of that speaker. Only if you sit very close to the speaker—say, within 2 or 3 feet, on axis—does the first arrival response really count. Once you’re back about 6-8 feet, all first arrival sound is simple folded into the far-field power response.

Of course, the two are not mutually-exclusive. As a speaker designer, you can reduce near-field diffraction, correctly align your drivers, etc, do all kinds of things to optimize the near-field first-arrival sound without hurting or affecting the far-field power output. That is what most good speaker designers do—they optimize the near-field response “just in case” it makes a difference.

(It doesn’t, really. Double blind listening tests have proven it over and over. It’s highly doubtful that a blindfolded listener could reliably identify whether they were listening to AR-11’s placed horizontally or vertically from 10 feet away. But “just in case,” there’s no harm to design a speaker that also has “good” near-field characteristics.)

Test it Yourself

There’s a great test you can perform yourself at home that perfectly demonstrates the difference between near-field and far-field response. You don’t need any expensive test equipment. As a matter of fact, you don’t even need a Radio Shack SPL meter.

You need a pillow. Preferably, a sofa pillow, maybe 12” x 12” or so. That’s it.

Here’s what you do: sit close to the speaker, about 2 feet away. Play some full-range music that is rich in upper-midrange and treble content. (Jazz is good for this, because it has lots of cymbals, “reedy”-sounding saxophones, and piano upper harmonics.)

Listen to the music. Now hold the pillow up in front of your face so that your line of sight to the speaker is blocked.

Lower the pillow away from your face. Note the difference in the sound depending on whether the pillow is raised or lowered. When the pillow is raised in front of your face, the speaker’s sound becomes noticeably duller and muffled. This is because the majority of the high frequencies reaching your ear are the result of the speaker’s direct radiation, right from the tweeter to your ear, on a straight line. So when the pillow is raised, it blocks and absorbs those highs, and you can’t hear them. This is the “near field,” and what the pillow is interfering with is the so-called “first-arrival” sound.

Now move back to your normal listening position, about 8-10 feet away from the speakers. Repeat the test: Play the music and listen with the pillow sitting on your lap. Now play the same music and raise the pillow in front of your face. You’ll notice something pretty surprising: the sound is almost identical whether the pillow is raised or lowered. This is because you’re now sitting in the far field, and almost all the high frequencies are reaching your ears by way of room reflections and indirect sound—not on a straight line like in the near-field example.

(The wider a speaker’s high-frequency dispersion, the more this will be true, because a wide-dispersion speaker generates more room reflections than a narrow-dispersion speaker. In a generalization, one could say that a speaker with a 1” dome tweeter will generate more room reflections than a horn speaker with an intentionally limited radiation pattern.)

So all those near-field considerations like driver alignment, cabinet diffraction, phase relationships, etc. are a bit less important in real world listening conditions than their theoretical importance might suggest.

The take-away for good speaker designers is this: A smooth 1m anechoic on-axis response (“first-arrival”) is a good starting point toward developing a good sounding speaker. So are factors like lack of diffraction and intelligent driver alignment. Those things do matter, to some degree. (Note that above I said the far-field sound was “almost” identical with the pillow raised or lowered. Not “absolutely” identical, but “almost” identical.)

However, after the near-field basics are taken care of as best as is practical for that given design, the prime determinant of the speaker’s listening quality will be a smooth far-field power response.

(BTW, “smooth” far-field response, not necessarily “flat” far-field response for accurate, lifelike sound. Roy Allison had a formula that he derived (I wish I could find the article where I originally read it), something like a .5 or 1 dB per octave rolloff above 1 or 2kHz, something like that. “Flat” power tends to be too bright.)

Steve F.

[Carlspeak]

Now that Steve has opened that 'can', I think his thread marks an opportune time to re-open another can (i.e. the KITCHEN). IMO, this is where his discussion belongs because it applies to all speakers; not just AR's. Moving the thread to the kitchen with a pointer will accomplish this task. IIRC, my earlier post to reopen the kitchen was given some support by folks, who's opinions at least I respect.

C'mon Mark, Gene, anybody?, open it up!

[JKent]

The problem with the Kitchen was the lack of civility--an all too common problem in our society in general. Part of the problem is the fact that when one reads a printed post, it lacks the body language and inflection that the writer intended. The reader then reads into the post. But another problem is that "some" writers are just rude and abusive, and (I think) that's why Mark closed the door.

I agree with Carl that there are "hot" topics that could/should legitimately be in the Kitchen. But I don't think Gene should have to police it and chastise individuals who overstep the bounds of propriety.

Maybe "one more chance" to see if we can self-regulate?

Kent

[Carlspeak]

I agree with Carl that there are "hot" topics that could/should legitimately be in the Kitchen. But I don't think Gene should have to police it and chastise individuals who overstep the bounds of propriety.

Maybe "one more chance" to see if we can self-regulate?

Kent

As I pointed out in my original request to re-open the kitchen, the two chief protagonists who tended to monopolize the most heated discussions are now completely gone from CSP. One has passed away and the other no longer visits here. Hence my confidence that the kitchen will be a much 'calmer' place in the future.

[Steve F]

As I pointed out in my original request to re-open the kitchen, the two chief protagonists who tended to monopolize the most heated discussions are now completely gone from CSP. One has passed away and the other no longer visits here. Hence my confidence that the kitchen will be a much 'calmer' place in the future.

As I posted a while back in agreement with Carl on the re-opening of the Kitchen, the remaining main posters all have a very good track record of civility and proper decorum. We've consistently demonstrated the ability to disagree without crossing the bounds of rudeness and inappropriate behavior.

But if Gene won't budge.....what do all of you think of this ol' can o'worms? It's my contention that good, well-behaved, low-diffraction near-field response (like the 91 or 9) doesn't automatically preclude a smooth, well-controlled far-field response. A good designer can--and should--achieve both.

Steve F.

[Carlspeak]

.....what do all of you think of this ol' can o'worms? It's my contention that good, well-behaved, low-diffraction near-field response (like the 91 or 9) doesn't automatically preclude a smooth, well-controlled far-field response. A good designer can--and should--achieve both.

Steve F.

I think they are not mutually exclusive. Achieving both should bring about an overall nice performing loudspeaker.

If you are anyone reading this has a copy of Toole's 'Sound Reproduction..." tome, I refer you'all to Fig. 18.17 on pg. 399 and the acompanying discussion. The figure shows polar and directivity plots of a $16K cone and dome loudspeaker and spearately, a $11K cone woofer/horn mid/high speaker. Note the difference in directivity. The former produces a wide dispersion and the latter more directional - which one would expect of these. Nice comparison.

Toole says listener ratings of these speakers were basiclly a toss up as to which was liked more.

[Steve F]

I think they are not mutually exclusive. Achieving both should bring about an overall nice performing loudspeaker.

If you are anyone reading this has a copy of Toole's 'Sound Reproduction..." tome, I refer you'all to Fig. 18.17 on pg. 399 and the acompanying discussion. The figure shows polar and directivity plots of a $16K cone and dome loudspeaker and spearately, a $11K cone woofer/horn mid/high speaker. Note the difference in directivity. The former produces a wide dispersion and the latter more directional - which one would expect of these. Nice comparison.

Toole says listener ratings of these speakers were basiclly a toss up as to which was liked more.

Toole (along with many others) is a strong believer in axial response being the primary factor in speaker preference and he has conducted tests that "prove" his point. Many people can contrive a set of circumstances or preference questions to come to a pre-conceived, desired conclusion. I'm not saying that Toole's preference tests were contrived or disingenuous, only pointing out that they are not definitive, regardless of what he or his adherents might think.

The "pillow" test shows the fallacy of ascribing too much faith in one extreme position or another.

Much depends on whether the listener is fully immersed in the reverberant field, close up in the direct field, or at the "critical distance" where the two influences overlap.

In any event, given that good near-field and good power response can be gotten with one design, it seems counter-intuitive not to strive for both.

Steve F.

[soundminded]

First-Arrival vs. Power Response

Let’s revisit—just for fun—a decades-long, still-unresolved controversy among speaker designers:

The question of first-arrival sound vs. far-field power response.

Boiled down into simple terms, the first-arrival crowd feels that the on-axis, first-arrival “anechoic” frequency response is what the ear grabs onto first and that sound is the primary determinant of a speaker’s essential sound/tonal quality. They feel that the room reflections, driver interaction, phase considerations, etc—however closely in milliseconds those reflections and artifacts may follow the ‘first arrival’—are separated out by the brain/ear system and what you fixate on is that first arrival.

Balderdash, says the far-field power response crowd. The room reflections etc are so close in time to the first arrival that they swamp the first arrival completely, leaving the overall far-field energy output of the speaker—the so-called “power response”—as what you hear to determine the tonal quality of that speaker. Only if you sit very close to the speaker—say, within 2 or 3 feet, on axis—does the first arrival response really count. Once you’re back about 6-8 feet, all first arrival sound is simple folded into the far-field power response.

Of course, the two are not mutually-exclusive. As a speaker designer, you can reduce near-field diffraction, correctly align your drivers, etc, do all kinds of things to optimize the near-field first-arrival sound without hurting or affecting the far-field power output. That is what most good speaker designers do—they optimize the near-field response “just in case” it makes a difference.

(It doesn’t, really. Double blind listening tests have proven it over and over. It’s highly doubtful that a blindfolded listener could reliably identify whether they were listening to AR-11’s placed horizontally or vertically from 10 feet away. But “just in case,” there’s no harm to design a speaker that also has “good” near-field characteristics.)

Test it Yourself

There’s a great test you can perform yourself at home that perfectly demonstrates the difference between near-field and far-field response. You don’t need any expensive test equipment. As a matter of fact, you don’t even need a Radio Shack SPL meter.

You need a pillow. Preferably, a sofa pillow, maybe 12” x 12” or so. That’s it.

Here’s what you do: sit close to the speaker, about 2 feet away. Play some full-range music that is rich in upper-midrange and treble content. (Jazz is good for this, because it has lots of cymbals, “reedy”-sounding saxophones, and piano upper harmonics.)

Listen to the music. Now hold the pillow up in front of your face so that your line of sight to the speaker is blocked.

Lower the pillow away from your face. Note the difference in the sound depending on whether the pillow is raised or lowered. When the pillow is raised in front of your face, the speaker’s sound becomes noticeably duller and muffled. This is because the majority of the high frequencies reaching your ear are the result of the speaker’s direct radiation, right from the tweeter to your ear, on a straight line. So when the pillow is raised, it blocks and absorbs those highs, and you can’t hear them. This is the “near field,” and what the pillow is interfering with is the so-called “first-arrival” sound.

Now move back to your normal listening position, about 8-10 feet away from the speakers. Repeat the test: Play the music and listen with the pillow sitting on your lap. Now play the same music and raise the pillow in front of your face. You’ll notice something pretty surprising: the sound is almost identical whether the pillow is raised or lowered. This is because you’re now sitting in the far field, and almost all the high frequencies are reaching your ears by way of room reflections and indirect sound—not on a straight line like in the near-field example.

(The wider a speaker’s high-frequency dispersion, the more this will be true, because a wide-dispersion speaker generates more room reflections than a narrow-dispersion speaker. In a generalization, one could say that a speaker with a 1” dome tweeter will generate more room reflections than a horn speaker with an intentionally limited radiation pattern.)

So all those near-field considerations like driver alignment, cabinet diffraction, phase relationships, etc. are a bit less important in real world listening conditions than their theoretical importance might suggest.

The take-away for good speaker designers is this: A smooth 1m anechoic on-axis response (“first-arrival”) is a good starting point toward developing a good sounding speaker. So are factors like lack of diffraction and intelligent driver alignment. Those things do matter, to some degree. (Note that above I said the far-field sound was “almost” identical with the pillow raised or lowered. Not “absolutely” identical, but “almost” identical.)

However, after the near-field basics are taken care of as best as is practical for that given design, the prime determinant of the speaker’s listening quality will be a smooth far-field power response.

(BTW, “smooth” far-field response, not necessarily “flat” far-field response for accurate, lifelike sound. Roy Allison had a formula that he derived (I wish I could find the article where I originally read it), something like a .5 or 1 dB per octave rolloff above 1 or 2kHz, something like that. “Flat” power tends to be too bright.)

Steve F.

I am going to say that IMO both are wrong. First thing to understand from my posting on the other thread is that for accurate sound reproduction of acoustic instruments, acoustics of the venue of performance has an effect on tone so reproducing that tone without also reproducing the acoustics can't be done. That's because they are different aspects of the same phenomenon. What's left is reproducing the sound of musical instruments as they would be heard if they were in your room. To achieve that, both the directly arriving AND sound reflected off the walls must be flat...all the way back to the recording microphones. Commercial loudspeaker systems fail this particular criteria for at least two reasons (there are other non spectral criteria BTW.) First the speakers don't radiate sound indirectly with the same spectral transfer function (off axis FR) as they do in the forward direction. Even AR 3a does not radiate flat power and neither does AR LST. This is easily seen by the fact that tweeter output falls off away from axis as frequency increases although with this particular tweeter to a lesser degree than others. The second reason is that even if radiated power were truly flat, the room boundaries reflect sound to a different degree as a function of frequency due to the acoustic properties of the material they're made of. It's analogous to shining a white light at a pink tinted mirror, the reflections will be pink when they arrive where you see them. The speaker is not engineered to control the reverberent and direct field spectra independently resulting in an arriving spectrum such that both will be flat. The phenomenon is predicted by my mathematical model and I've re-engineered many of my speakers to correct for this.

The other factor is differences in spectral balance of recordings. In my experience no two are alike. Therefore each must be compensated for individually. How is that done? The only way is by ear, by remembering what the actual instruments sound like and applying equalization individually. It is tedious and can be very frustrating. Success if it comes is after trial and error, often requiring many repeated attempts.

[genek]

It's not a Gene thing. I'll moderate any forum Mark decides he wants to have open.

[Steve F]

OK, for the sake of this discussion, let’s assume that everything Soundminded has posited is absolutely true. Actually, since he’s done so much research and has such extensive practical experience, there’s no reason to just assume it’s true “for the sake of this discussion.” There’s actually no reason to believe that it is anything other than completely true.

Nonetheless, what he’s presented has something of the “What is the meaning of life?” quality to it: an unanswerable question, a problem without a practical, repeatable solution.

In the real, imperfect, fundamentally-flawed, false-assumption-filled world in which the average person listens to reproduced music, said music is played back on loudspeakers that fail to or aren’t intended to take into account the multiple, complex issues that Soundminded raises.

We recognize that and acknowledge it. Home listening systems and commercial recordings can’t simulate real music, for all the reasons perfectly laid out in Soundminded’s explanation.

Now, let’s put all that aside.

In this real, imperfect world in which we live, real people design imperfect loudspeakers that are part of imperfect systems that play back imperfect recordings.

Are those imperfect loudspeakers more pleasing and do they do a better job or a worse job of conveying the information on those admittedly imperfect recordings if they are optimized for first-arrival response or if they are optimized for far-field power response?

Is there any compelling reason to favor narrow- or controlled-directivity loudspeakers over wide-dispersion loudspeakers?

What role do room acoustics (dead v live v balanced) and listener distance (near v far v critical) play in the speaker designer’s assumptions of how their speaker is to be used? What role does program material play (large-scale orchestral v small-scale jazz trio, etc.)?

Is a speaker that does well in both near- and far-field, like the AR-91, a better or worse speaker than one optimized for near- or far-field only, assuming as much of everything else (FR, distortion, etc.) is the same?

We know that if one believes in the Meaning of Life conditions that virtually everything I’m asking for opinions on here is irrelevant and non-sequitur. But humor me anyway.

First-arrival or Far-field and why?

Steve F.

[kirk57]

Test it Yourself

There’s a great test you can perform yourself at home that perfectly demonstrates the difference between near-field and far-field response. You don’t need any expensive test equipment. As a matter of fact, you don’t even need a Radio Shack SPL meter.

You need a pillow. Preferably, a sofa pillow, maybe 12” x 12” or so. That’s it.

Here’s what you do: sit close to the speaker, about 2 feet away. Play some full-range music that is rich in upper-midrange and treble content. (Jazz is good for this, because it has lots of cymbals, “reedy”-sounding saxophones, and piano upper harmonics.)

Listen to the music. Now hold the pillow up in front of your face so that your line of sight to the speaker is blocked.

Lower the pillow away from your face. Note the difference in the sound depending on whether the pillow is raised or lowered. When the pillow is raised in front of your face, the speaker’s sound becomes noticeably duller and muffled. This is because the majority of the high frequencies reaching your ear are the result of the speaker’s direct radiation, right from the tweeter to your ear, on a straight line. So when the pillow is raised, it blocks and absorbs those highs, and you can’t hear them. This is the “near field,” and what the pillow is interfering with is the so-called “first-arrival” sound.

Now move back to your normal listening position, about 8-10 feet away from the speakers. Repeat the test: Play the music and listen with the pillow sitting on your lap. Now play the same music and raise the pillow in front of your face. You’ll notice something pretty surprising: the sound is almost identical whether the pillow is raised or lowered. This is because you’re now sitting in the far field, and almost all the high frequencies are reaching your ears by way of room reflections and indirect sound—not on a straight line like in the near-field example.

I tried this and have to say that I quite enjoyed the sound at the farfield position with the pillow in front of me. It is somewhat similar to what I heard with my Bose 901s (that is NOT a criticism of 901s!) Also it seems my speakers have fairly wide high frequency dispersion.

[soundminded]

OK, for the sake of this discussion, let’s assume that everything Soundminded has posited is absolutely true. Actually, since he’s done so much research and has such extensive practical experience, there’s no reason to just assume it’s true “for the sake of this discussion.” There’s actually no reason to believe that it is anything other than completely true.

Nonetheless, what he’s presented has something of the “What is the meaning of life?” quality to it: an unanswerable question, a problem without a practical, repeatable solution.

In the real, imperfect, fundamentally-flawed, false-assumption-filled world in which the average person listens to reproduced music, said music is played back on loudspeakers that fail to or aren’t intended to take into account the multiple, complex issues that Soundminded raises.

We recognize that and acknowledge it. Home listening systems and commercial recordings can’t simulate real music, for all the reasons perfectly laid out in Soundminded’s explanation.

Now, let’s put all that aside.

In this real, imperfect world in which we live, real people design imperfect loudspeakers that are part of imperfect systems that play back imperfect recordings.

Are those imperfect loudspeakers more pleasing and do they do a better job or a worse job of conveying the information on those admittedly imperfect recordings if they are optimized for first-arrival response or if they are optimized for far-field power response?

Is there any compelling reason to favor narrow- or controlled-directivity loudspeakers over wide-dispersion loudspeakers?

What role do room acoustics (dead v live v balanced) and listener distance (near v far v critical) play in the speaker designer’s assumptions of how their speaker is to be used? What role does program material play (large-scale orchestral v small-scale jazz trio, etc.)?

Is a speaker that does well in both near- and far-field, like the AR-91, a better or worse speaker than one optimized for near- or far-field only, assuming as much of everything else (FR, distortion, etc.) is the same?

We know that if one believes in the Meaning of Life conditions that virtually everything I’m asking for opinions on here is irrelevant and non-sequitur. But humor me anyway.

First-arrival or Far-field and why?

Steve F.

"In the real, imperfect, fundamentally-flawed, false-assumption-filled world..."

Steve that's a compelling argument. I'll bet it was the same argument used about 120 years ago when some engineers and tinkerers proposed that land vehicles could one day be propelled by nothing more than harnassing the energy of contained explosions of petroleum distillates but here we are and 7 billion people use that as almost the exclusive means of land transportation.

"real people design imperfect loudspeakers that are part of imperfect systems that play back imperfect recordings."

Progress comes when people who hit a brick wall with the old ideas that only take them so far rethink the whole problem from scratch without any preconceptions. They sometimes can and do see problems in a different light and invent whole new theories that are exploited to get much better solutions to problems. Who would have thought before the vacuum tube was obsoleted by transistors that the equivalent of millions of them would fit in the palm of your hand powered by a small battery with more computing power than all the machines in the world in that day combined?

"Are those imperfect loudspeakers more pleasing...?"

I don't understand the relevance of that statement. My aunt who started out to be a concert pianist and collected thousands of phonograph records and later cds enjoyed them enormously on a 2 watt per channel Symphonic record player with two cheap 4" table radio speakers and a Collaro record changer with a crystal turnover cartridge, later a large Sony boombox with a cd player in it. What I thought this discusion and the purpose of Acoustic Research speakers is about was to provide sound that is accurate in the sense of reproducing musical instruments. Whether or not that's more or less pleasing I think is a different discussion.

I'm not surprised that the idea of adaptive loudspeakers which achieve similar results in many rooms by compensating for their differences has not been developed yet. The whole idea of equalization has been rejected by much of what exists of the audiophile community and low cost automated versions are just now finding their way in to the consumer receiver market.

I'm also not surprised that there is no standardization for making recordings. There once was something like it. Back in the bad old days of phonograph records, recording studios made it standard practice to have the FR calibration of their monitor systems checked and their equalization adjusted weekly, this in the hope presumably that there would be consistancy from one recording to another in terms of spectral balance. That also played a role in making one record label's product not all that dissimilar to another's. For cds made in a later era when the notion of equalization to a standard is a taboo, the balance from one cd to another is all over the map. That's one possible explanation for why so many people still prefer the sound of vinyl phonograph records over cds.

My experience is that when speakers are re-engineered to achieve the objective of flat transfer of power, not just flat on axis power and certainly not flat total radiated power, the results sound to me more like real musical instruments I'm familiar with. Is that better? I'm sure the people who make expensive instruments like Steinway and Baldwin and people who still enjoy live music played on real instruments may think so.