Horn vs direct bass, how is it different?

[MikeFord]

I've only listened to Khorns a few times, and really nothing else seriously that wasn't a direct radiator or ported box, and assumed that low bass, below 40hz or so would always need to come from a direct radiator. I started thinking about it last night and now I am wondering if I have missed some great sonic difference between horn bass and direct radiator.

I'm too deep to turn back on a 2x 12" direct radiator sub box, so that will be the next sub in my system, but dandy time to start pondering the next one.

1) Whats different between horn and direct radiator bass, subjective and science behind it answers desired.

2) How do they mix, like horn down to 30hz and the last octave down to 15hz a direct radiator?

[CECAA850]

I've read several places not to mix horn and direct radiating subs. I don't recall the reasoning other than phasing issues with the long horn path in the horn sub and no horn path in the direct radiator. If they both receive the same signal without accounting for the delay that the horn path creates, the horn subs sound will be slightly behind the direct radiator's sound causing a smearing of the bass.

[Thaddeus Smith]

I've read several places not to mix horn and direct radiating subs. I don't recall the reasoning other than phasing issues with the long horn path in the horn sub and no horn path in the direct radiator. If they both receive the same signal without accounting for the delay that the horn path creates, the horn subs sound will be slightly behind the direct radiator's sound causing a smearing of the bass.

...and a rip in the space-time continuum. sort of an important detail.

[Chris A]

Quote

1) What's different between horn and direct radiator bass, subjective, and science behind it...

 

This will answer the question you pose, at least from a qualitative standpoint:

https://community.klipsch.com/applications/core/interface/file/attachment.php?id=80871

When you think about it, the only difference between a Cornwall and a Khorn is the bass bin: the first being a direct radiator, and the second horn-loaded.

Another reference:

 

"[Horn-loaded bass bins, corner-loaded] provide dramatically lower bass distortion, in particular, modulation distortion, than non-corner-loaded loudspeakers (https://community.klipsch.com/applications/core/interface/file/attachment.php?id=82430). Bass modulation distortion has been found to be quite audible ."

 

Quote

2) How do they mix, like horn down to 30hz, and the last octave, down to 15hz, a direct radiator?

 

You need to correct for the added path length of the horn when integrating horn--loaded bass bins and subs into a multi-channel system, and within the loudspeaker itself using time delay filters that are available using active digital crossovers and some AV pre/pros and receivers that can actively bi-amp.

 

When you get them time-aligned with the speakers that they are crossing against, the sound image seems to coalesce and the timbre of the sound is much more natural.

 

Chris

[Quiet_Hollow]

1) Whats different between horn and direct radiator bass, subjective and science behind it answers desired.

The total sum of distortion. The horn, and most notably the front-loaded folded horn, is going to exhibit the least amount...by a couple orders of magnitude.

Subjectively, this tranlates to sounding very "real" or "clean"...so long as they're integrated properly.

2) How do they mix, like horn down to 30hz and the last octave down to 15hz a direct radiator?

Ever heard a large line array or cinema system lately? How about a Cornwall or Heresy?

Using the two together imparts a pretty steep compromise in output quality IMO, but horns and direct radiators are used in combination, in a lot of different applications for a good long while now, and quite successfully.

[ClaudeJ1]

If a cone moves, it produces distortion. If you want to REDUCE cone movement, put a HORN on it. Power input will go down, power output will go UP, distortion will go DOWN more than the power input does (very cool). Horns will play louder and more like live music because the cones don't have to move as much, so they ACCELERATE faster for more lifelike TRANSIENT response.

The greatest movement in any loudspeaker regardless of size or brand is in the woofer section. The least amount of of movement is in the tweeter section. Where is the horn needed most? In the Woofer section, of course, yet this is the LAST place people put a horn. Why? SIZE and COST. But for those who don't wish to compromise, like me, I insist on ALL HORNS in the entire sytem, except for surrounds because they are down so far in output.

This is why I use Tapped Horn Subs (18 to 24 feet long), which will give the greatest bass output into the TEENS in the smallest cabinet possible working in concert with ROOM GAIN.

Since I was just a kid out of college, I have owned Klipschorns with a LaScala in the middle using PWK's resistor box. I know this for sure because I went to his house and saw/heard for myself. I had the exact same system that PWK had, for 30 years.

I spent the last 5 years doing about 10 iterations of horns going from 5-way to 4-way (including the TH sub in this statement). My whole system is based on the K-402 with the Klipsch 1133 driver, found in over 50% of newer theaters around the world. There's a reason for this.

So if that setup produces the lowest distortion possible for a great big room like a theater with hundreds of sears, imagine what the same horns do in your home?

Talk about ridiculously LOW DISTORTION using only MILLIWATTS from little CHIP amps of about 6 Watts per channel (except for the subs), where one can measure between 0.010-0.025 W in the main channels and 0.25 (1/4 watt) in the woofers for normal listning with maybe 10X that for "cranked up demo mode."

Spoiled ROTTEN you say? You are darn RIGHT!

[Edgar]

I asked a similar question to D.B. (Don) Keele a while back. My question was posed from the standpoint of what horns do best (in my opinion); dynamics and transient response. I asked Keele whether a direct-radiator bass loudspeaker could match the dynamics and transient response of a bass horn, assuming that the combined total cone area of the direct radiator drivers was the same as the mouth area of the horn. His answer was that, since the direct radiator loudspeaker would have greater bandwidth capability than the horn (the direct radiator could go both lower in frequency and higher in frequency than the horn), the direct radiator would actually have better transient response than the horn.

I think that this needs to be qualified, however. Even under the conditions described above, the horn is likely to have greater sensitivity than the direct radiator (play louder for the same power input). This is not anything intrinsic to horns or direct radiators, but is a result of the reduction in sensitivity that occurs when drivers are designed for greater excursion, as most are today. And, as a result, the horn will be subjectively capable of greater dynamics than the direct radiator, simply because of its greater sensitivity.

Now, if one were to use some extremely high sensitivity direct radiators (with limited excursion capability), like perhaps several GPA 515-LF @ ~100 dB/W, then the dynamic differences between horns and direct radiators might just be eliminated.

[Chris A]

...His answer was that, since the direct radiator loudspeaker would have greater bandwidth capability than the horn (the direct radiator could go both lower in frequency and higher in frequency than the horn), the direct radiator would actually have better transient response than the horn...Even under the conditions described above, the horn is likely to have greater sensitivity than the direct radiator (play louder for the same power input).

That last statement implies that Keele is not implying an apples-apples comparison. Even Keele's article on horns vs. vented enclosures was not comparing sound quality, but rather relative volumes--a subject that I'm not really not that interested in (...but the "small speaker crowd" IS interested). Quality of sound usually is overlooked in these type of discussions.

The problem of modulation distortion perception, a subject of which many of today's loudspeaker designers like to ignore--especially bass modulation distortion--IS the problem. That's what apparently sets PWK apart from Keele. Relative bandwidth discussions aren't relevant, right? If bandwidth is brought up in the discussion of horns vs. direct radiators, then my response would be to first ask about the directivity control issues of direct radiators (vs. horn-loaded drivers).

Then I'd start asking about relative modulation distortion (both FM and AM) properties, and not "transient response". You can't EQ-out modulation distortion, and attempts to reduce it through in-room processing (read your JAES articles from about 15-20 years ago) show that the performance gap of simple horn-loading vs. direct radiators using esoteric post-processing approaches to reduce modulation and and other transient response distortions, is still too wide a gap to close. Horns win yet again.

Most of these arguments get down to one thing: most people don't like the size of horn-loaded bass bins. However, physics is physics: sound waves don't like to be created or directed at less than 1/4 wavelength sizes. Using direct radiators ignores the problems of modulation distortion (i.e., PWK's arguments), efficiency+economics of drivers (PWK's argument) and directivity control (Earl Geddes' argument).

Chris

[Edgar]

If bandwidth is brought up in the discussion of horns vs. direct radiators, then my response would be to first ask about the directivity control issues of direct radiators (vs. horn-loaded drivers).

To be fair, my question to Keele cited a "living room" context, where directivity, at least at bass frequencies, is really not an issue. Almost everything is in ½Pi or Pi space.

Then I'd start asking about relative modulation distortion (both FM and AM) properties, and not "transient response".

Again, it was I who posed the question in transient response terms, not he.

But pursuing the modulation distortion topic; if the direct radiator cone area equals horn mouth area, then cone excursion for the direct radiators should be comparable to cone excursion for the horn driver, so that modulation distortion should also be comparable. That was the reason that I specified cone area = mouth area, to try to put things on an apples-to-apples basis. For example, the KHorn mouth area is just under 500 square inches. To equal that, one would need 4 GPA 515-LF drivers. Those four drivers, at ~100 dB sensitivity each, would yield ~106 dB sensitivity. That's comparable to the KHorn. But that would be a very large and expensive loudspeaker system.

[Chris A]

But pursuing the modulation distortion topic; if the direct radiator cone area equals horn mouth area, then cone excursion for the direct radiators should be comparable to cone excursion for the horn driver, so that modulation distortion should also be comparable

The efficiency or "sensitivity" would be the same, but not the modulation distortion or the relative mass properties. Remember that PWK said that a 36" diameter woofer would be needed to equal the area of the Khorn bass bin placed in a good room corner. But the 36" woofer would have mass issues that the horn doesn't have. AM distortion would also be much higher for the direct radiator than the equivalent horn with like mouth size.

If you are trying to use multiple bass direct radiator drivers to equal a 36" diameter woofer, then you've got two strikes against you: the unrecoverable economics issues of using that many drivers (a steep cost curve), and diffraction issues due to the use of multiple drivers, but at least you'll not lose on bandpass of the direct radiators if using lots of drivers in parallel. The farther you go down that path, the more it begins to look like a planar loudspeaker, and you know the issues with that--limited dynamic range (due to limited throw on the planar membrane) and poor off-axis polars, resulting in a poor sounding speaker in any reasonable-sized room. I suppose that monopolar planar speakers would be nice (like a Beveridge) but massive amounts of power are required, and usually the cost of high quality power (e.g., tubes) goes out of sight. (See Toole's discussion of the benefits of good loudspeaker polars and its correlation to positive listener reviews in his book.)

It seems to me that it's much easier just to use horns and be done with it, unless of course you're trying to make your speakers smaller and you really don't care about sound quality... [8-)]

Chris

[Edgar]

The efficiency or "sensitivity" would be the same, but not the modulation distortion ...

I'm not so sure about that, but I can't cite the math off the top of my head to say one way or the other. I suspect that two systems with the same radiating area and the same sensitivity will have the same modulation distortion properties.

Remember that PWK said that a 36" diameter woofer would be needed to equal the area of the Khorn bass bin placed in a good room corner.

I suspect that PWK was embellishing a bit. A 36" woofer would have an Sd of around 900-950 square inches. A KHorn has a mouth area of around 475 square inches. All other things being equal, the 36" woofer will win.

AM distortion would also be much higher for the direct radiator than the equivalent horn with like mouth size.

Can you cite a reference for that? I'm not challenging you; I just want to study the math for myself.

Greg

[Chris A]

I'm not so sure about that, but I can't cite the math off the top of my head to say one way or the other. I suspect that two systems with the same radiating area and the same sensitivity will have the same modulation distortion properties.

Horns always win on modulation distortion (on equal mouth area vs. cone area comparisons). Think of the relative moving masses: air + very small compression driver area loaded with higher pressure air vs. very large direct radiating cone/cones working on much lower pressure air (remembering the impedance mismatch issues, too, of working on low pressure air, and the nonlinear effects of diaphram motion at the root of AM distortion creation--the dominant source of modulation distortion at low frequencies). Modulation distortion is inherent in moving diaphragms - the more that they move, the more modulation distortion (AM and FM).

All other things being equal, the 36" woofer will win.

Win what? Not when a figure of merit that includes the absence of modulation distortion is used as a significantly weighted decision variable.

Can you cite a reference for that?

Yes: see the articles on modulation distortion referenced above written by PWK (Part 1 and Part 2), repeated and expanded by file above to include an additional article (see enclosed pdf).

modulation-distortion-in-loudspeakers-1-3.pdf

[Edgar]

Horns always win on modulation distortion (on equal mouth area vs. cone area comparisons). Think of the relative moving masses: air + very small compression driver area loaded with higher pressure air vs. very large direct radiating cone/cones working on much lower pressure air (remembering the impedance mismatch issues ...

Yes, I am aware of all the issues of pressure vs velocity, impedance mismatch, etc. However, the statements above are only assertions. It is a fact that increasing cone area decreases cone excursion, and some point is reached at which the direct radiator cone excursion equals the horn driver cone excursion. Even PWK says that distortion is closely proportional to cone excursion (Part III of the articles that you cited; Conclusion). The only questions are: At what point are direct radiator cone excursion and horn cone excursion equal (for the same output)? Is it when cone area equals horn mouth area?

Win what? Not when a figure of merit that includes the absence of modulation distortion is used as a significantly weighted decision variable.

Win exactly the comparison that PWK was trying to make by stating that the KHorn was equivalent to a 36" woofer. It's not.

... see the articles on modulation distortion referenced above written by PWK (Part 1 and Part 2), repeated and expanded by file above to include an additional article (see enclosed pdf).

In Part I of the articles that you cite, PWK said on page 145: "Among means to reduce distortion, one of the most obvious would appear to be to increase the diaphragm area. [...] Increasing the number of direct-radiator loudspeakers has also been used. This also improves efficiency, although not to the extent realizable with well-designed horns; however, the bulk and cost equal or exceed that of horns, and difficulties with polar response arise."

Apart from the polar response issue and the cost issue, which I addressed in an earlier post, it would appear that PWK agrees with me. Again, the only question is, how big does a direct radiator system have to be in order to have performance comparable to a horn with respect to everything that we have identified: transient response, dynamics, sensitivity, and distortion? I suspect that it's when cone area equals mouth area, or something pretty close to it, but I cannot say for sure because I just haven't studied the math in depth.

[CECAA850]

It is a fact that increasing cone area decreases cone excursion, and some point is reached at which the direct radiator cone excursion equals the horn driver cone excursion.

That's so true. I have over 1000 sq. in. of radiating area in my IB. It puts out unbelieveable ammounts of bass and the drivers barely move. As stated, the downside is expense, the upside is no large horn box taking up floor space. All subs are compromises of one type or another. There's no one sub to rule them all. Pick the best sub for your application and enjoy it.

Feel free to continue the debate, it's fascinating.

Carl

[Chris A]

Apart from the polar response issue and the cost issue, which I addressed in an earlier post, it would appear that PWK agrees with me. Again, the only question is, how big does a direct radiator system have to be in order to have performance comparable to a horn with respect to everything that we have identified: transient response, dynamics, sensitivity, and distortion? I suspect that it's when cone area equals mouth area, or something pretty close to it, but I cannot say for sure because I just haven't studied the math in depth.

Don't get me wrong, but I have to say that these sort of discussions are not energizing for me personally. Let me relate a story that might help you:

I once heard the Jubs in Hope: the KPT-KHJ-LF bass bin, and the K-402/K-69-A on top, crossed, EQed, etc. Then the bass bin was swapped with a KPT-904 (dual 15" woofers) and re-EQed. It sounded like a very good Cornwall with a huge soundstage.

Then the 904 was swapped with a KPT-415 (quad 15" woofers - 600+ square inches of woofer - double that of the 904), and re-EQed. It sounded big and muscular - like that which you hear at a disco. But it didn't sound like the Jubilee. There were others in the room that liked that "sound" of the 415 a lot. All I could hear was a LOT of authoritative bass...and bass distortion--of the sort that I hear with any direct radiator box. I personally didn't like the sound of that type of bass distortion, so I eventually left the room and watched some tests being run in the chamber.

I did really like the TAD TD-4002s on the K-402s, on top of KPT-KHJ-LF bass bins that preceded that particular A-B-C test, however. I now own a pair of those drivers.

End of story.

Chris

[DizRotus]

If a cone moves, it produces distortion. If you want to REDUCE cone movement, put a HORN on it. Power input will go down, power output will go UP, distortion will go DOWN more than the power input does (very cool). Horns will play louder and more like live music because the cones don't have to move as much, so they ACCELERATE faster for more lifelike TRANSIENT response.

The greatest movement in any loudspeaker regardless of size or brand is in the woofer section. The least amount of of movement is in the tweeter section. Where is the horn needed most? In the Woofer section, of course, yet this is the LAST place people put a horn. Why? SIZE and COST. But for those who don't wish to compromise, like me, I insist on ALL HORNS in the entire sytem, except for surrounds because they are down so far in output.

This is why I use Tapped Horn Subs (18 to 24 feet long), which will give the greatest bass output into the TEENS in the smallest cabinet possible working in concert with ROOM GAIN.

Since I was just a kid out of college, I have owned Klipschorns with a LaScala in the middle using PWK's resistor box. I know this for sure because I went to his house and saw/heard for myself. I had the exact same system that PWK had for 30 years.

I spent the last 5 years doing about 10 iterations of horns going from 5-way to 4-way (including the TH sub in this statement). My whole system is based on the K-402 with the Klipsch 1133 driver, found in over 50% of newer theaters around the world. There's a reason for this.

So if that setup produces the lowest distortion possible for a great big room like a theater with hundreds of sears, imagine what the same horns do in your home?

Talk about ridiculously LOW DISTORTION using only MILLIWATTS from little CHIP amps of about 6 Watts per channel (except for the subs), where one can measure between 0.010-0.025 W in the main channels and 0.25 (1/4 watt) in the woofers for normal listning with maybe 10X that for "cranked up demo mode."

Spoiled ROTTEN you say? You are darn RIGHT!

I've had the pleasure of experiencing Claude's system. It's ALL HORNS (except Cornwall center) and it's superb. Claude knows of what he speaks!

[Edgar]

Then the 904 was swapped with a KPT-415 (quad 15" woofers - 600+ square inches of woofer - double that of the 904), and re-EQed. It sounded big and muscular - like that which you hear at a disco. But it didn't sound like the Jubilee.

That is a very significant observation. But it raises more questions. The fact is that, given enough cone area, the modulation distortion from a direct radiator will be as small as that from a horn -- PWK himself said that modulation distortion is "closely proportional" to excursion. Under such circumstances the sonic difference is not a result of modulation distortion. So ... what is the cause?

One can hear differences in the "character" of the bass from the same woofer in different enclosures (particularly closed vs vented), even though distortion is comparable in all cases. What is the cause?

These questions are rhetorical. My point is that not all of the sonic character of a horn can be attributed to reduced modulation distortion.

And I like horn bass, too. But, as we all know, it is a real challenge to get that last octave with a horn.

[DizRotus]

" . . . But, as we all know, it is a real challenge to get that last octave with a horn."

Claude's system convinced me that a tapped horn is up to the challenge. His coffin sized unit functions as a coffee table.

Tapped horns can be integrated into a system in many ways. They can be in an attic, under the floor, in a closet, in a garage, as part of a bookcase, used as a coffee table, etc.. Despite their size, with a little creativity they can be hidden or camouflaged. They're easy to build and relatively inexpensive.

In other words, "We don't need no stinkin' direct radiators."

[MikeFord]

With some LONG links of reading it is taking me some time to follow along. (some notes below as I read).

Bass is funny, below 100 hz or so and certainly below 40hz, a simple spl measure is not adequate to adjust level, especially not with the purpose of SQ comparison. Dipoles, monopoles, and horns excite room modes differently and to different levels.

Bass has three types, Natural from musical instruments etc which doesn't often go below 30 hz, created sounds or eqed natural sounds usually not below 20hz, and natural environmental "room" sounds that don't appear to have a lower limit, at least not that know. How low? How loud? How clean? The range of good answers seems to be from 10hz or so maybe 80hz, and not much consensus on how loud, with a fair amount of distortion frequently tolerable.

I like the bass on my Forte's, it seems close to just right on the fat/thin scale, no clue how deep or how loud, but likely its below my plenty loud threshold. I like chest slap bass, but not sure my neighbors would.

[ClaudeJ1]

" . . . But, as we all know, it is a real challenge to get that last octave with a horn."

Claude's system convinced me that a tapped horn is up to the challenge. His coffin sized unit functions as a coffee table.

Tapped horns can be integrated into a system in many ways. They can be in an attic, under the floor, in a closet, in a garage, as part of a bookcase, used as a coffee table, etc.. Despite their size, with a little creativity they can be hidden or camouflaged. They're easy to build and relatively inexpensive.

In other words, "We don't need no stinkin' direct radiators."

It's all now in a new place, huge room, it dwarfs what you heard before. My "coffee table" is now standing up in a corner of an 18x24 loft inside of an 18x45 foot room. It goes to 15 Hz. easily. Anyone who has hear this latest setup, same speakers, new room, has commented on the definition, TRANSIENT response, depth, etc. of the bass and the total lack of boominess.