Corner-Horn Imaging FAQ

[Ski Bum]

Thanks, Ski Bum. Very interesting.

I did comparethe Pass to my SETs as well. As you would expect, they had a somewhat more technicolor presentation with all their peculiarities, less power, and less solid bass. Frankly, and in spite of that, I preferred the SETs. I admit it, I have weird taste, but it just makes sense to go for something really different, rather than extreme expense for maybe some slight difference. That's probably the reason why pp tubes never appealed that much to me, too expensive and not different enough from good old SS power.

[Chris A]

"What About The Importance of 'Controlled Directivity' Speakers: What's That All About?"

 

There are two good external sources on this subject:

The bottom line is that speakers that exhibit nearly constant and suitably wide-angle horizontal polars show up at the top of the list of preferred speaker types in testing with listening subjects.

Dr. Floyd Toole (former chief engineer of Harman International, which includes JBL, Crown, dbx, Harman Kardon, Infinity, Lexicon, Mark Levinson, Soundcraft, Studer, etc.). Toole wrote an excellent book, Sound Reproduction: The Acoustics and Psychoacoustics of Loudspeakers and Rooms (Focal Press, 2008) that discusses the preference that listeners in blind test have for speakers that have relatively constant polars (like certain horn-loaded speakers and some direct-radiator speakers) over those speakers that do not exhibit constant polars (e.g., dipolar speakers).

 

Dr. Earl Geddes (formerly of Ford Motor Company, et al.) who has written on "acoustic waveguides" in his book Audio Transducers, Gedlee, 2002. Geddes points out that the most important characteristic of horns (which he calls "waveguides") is controlled directionality, not acoustic efficiency. He independently acknowledges that controlled directivity of loudspeakers is most preferred among loudspeaker characteristics, including freedom from distortion (certain types). He introduces the concept of "higher-order modes" of waveguides (horns) that augments typical horn equation approaches (i.e., the Webster "one dimensional" horn equation).

 

"What is Controlled Directivity?"

 

For the moment, think of a theoretical pulsating sphere in free space away from all acoustic reflections and room boundaries: its output is not sensitive to the horizontal azimuth or elevation that you hear it--if you are looking straight at it. At the other end of this extreme, if you think about a "sound bar" transducer, its output is more like that of a laser or a collimated beam of acoustic energy.

In between these two extremes is the concept of "controlled directivity" - the speaker limits its acoustic output to an angular sector of space, just like a directional antenna does for radio frequencies.

The assumption is that the speaker's output is almost constant within that covered angular sector. We don't want its acoustic SPL output to droop toward the edges of the covered angular sector for all frequencies that it covers (i.e., flat, neutral-color sound coverage in its passband horizontally and vertically), and we want its SPL output to be almost zero just outside of that sector.

 

"Again...Why is Controlled Directivity Important?"

 

Because it keeps the direct room output frequency spectra at different azimuths within the coverage sectors from being different (i.e., different "color" of sound) at different output angles. This is the key to controlled directivity horn speakers being preferred among listeners. The exact reasons why this is are covered in Toole's book, which is highly recommended for those who want to know more.

 

"Can You Just Tell Me, In a Nutshell, Why Controlled Directivity is Important?"

 

For the key to understanding this effect, think back on the discussion of the Precedence effect: early reflections (from ~1 to ~30 ms delay relative to the initial direct sound) are perceived differently by listeners than those reflections that are less than ~1 ms delay and those that are greater than ~30 ms delay.

 

 

If you are putting a lot of reflected energy on the walls, ceiling and floor around your speakers, what this does is smear out the sound image into a larger area (which can be a good thing from a casual listener's perspective) AND introduces phase delays in the direct-to-reflected energy (which is always bad in terms of small-room imaging performance) if you are in the "Haas zone" of delayed reflections.

We want to present a sound image that is in-phase horizontally and vertically from ~10 ms up to ~30 ms (depending on frequency content and loudness), without introducing "diffraction" of that image to the listener. But this is true only for delays of greater than ~10 ms and less than ~30 ms. Strange...but true.

 

"Why do Some Speaker Manufacturers Design Their Speakers to Reflect Off the Walls, Ceiling, and Floor (such as Bose)?"

 

If you are not interested in presenting a focused stereo image to the listener, the effects of spreading the energy around the front of the room (like Bose does...) is actually perceived pleasantly by casual listeners because it increases the "apparent source width" (ASW) of the speakers (see Floyd Toole's book for more on this).

The typical "anti-Bose story" goes something like: "Why do Bose speakers sound like they do? Because they sound uniformly bad everywhere you listen, but they sound the same way at all listening azimuths".

 

"Why Do Planar Speakers Have a Focused Sound Image (Albeit in a Small Azimuth of Listener Position)?"

 

Because they reflect very little SPL off the side walls, floor and ceiling due to their dipole characteristics, but most of the non-direct energy comes from the front wall reflections, AND planar speakers should be placed out into the room more than the 10-40 ms delay (total path length to the listener) from the front wall.

However, planar speakers (such as MartinLogans) still exhibit non-constant polars in azimuth and elevation--which is not preferred in blind testing by many listeners, especially when playing mono or "hard-cut" stereo material (see Toole's book, once again). Also, planar speakers can't be toed-in effectively to increase their imaging performance unless the room is quite large with no near-field azimuth reflectors present, i.e., greater than 10-20 feet (3-6 meters) in all directions around the speakers.

[Chris A]

"What Can I Do About Ceiling Bounce of Klipsch Heritage Midrange Horns (I.E., K-400 Series Horns)?"

 

A quick fix is to place an absorbent tile such as this to the top of the speaker cabinet, cut to fit appropriately but allowed to extend forward of the front shear line of the speaker's cabinet a couple of inches. This will have the effect of absorbing some widely dispersed polar acoustic energy exiting the horn (400--1700 Hz) that would otherwise wind up undesirably on your ceiling. While this technique is certainly limited in effectiveness, it may partially mitigate having to place absorption tiles on your ceiling in low-ceiling environments.

 

It will have a secondary effect of precluding random placement of drinks on top of your speakers (i.e., the pad will discourage placement of drink glasses and cups there due to its sponginess and corresponding uncertainty in supporting drinks by those random individuals). It will also act as an acoustic coating over any plate glass that you might have placed on top of your cabinets --which reflects more acoustic energy towards your ceiling.

 

Note that there will be a slight timbre shift (i.e., on the Khorn, Belle, La Scala, Cornwall, and Heresy) when you do this, and slightly improved imaging.

 

[garyrc]

Somewhere on the forum there is a multi-colored energy plot showing the smoothness of bass energy distribution with the speakers placed in the corners v.s. elsewhere in the room. It seems to show that the distribution is the most even with corner placement, contrary to one of the articles of faith of many writers. I don't remember the poster's name. I've been searching and searching the fourm, and can't locate it. Can someone tell me where to find this plot?

[Chris A]

http://forums.klipsch.com/forums/p/120269/1212300.aspx#1212300

[garyrc]

http://forums.klipsch.com/forums/p/120269/1212300.aspx#1212300

Thanks!

[mikebse2a3]

Somewhere on the forum there is a multi-colored energy plot showing the smoothness of bass energy distribution with the speakers placed in the corners v.s. elsewhere in the room. It seems to show that the distribution is the most even with corner placement, contrary to one of the articles of faith of many writers. I don't remember the poster's name. I've been searching and searching the fourm, and can't locate it. Can someone tell me where to find this plot?

http://forums.klipsch.com/forums/p/120269/1212300.aspx#1212300

Caution should be used when looking at these plots and drawing conclusions from them.

Look at the frequencies involved in these plots and you will see that they are more in subwoofer territory and not very complete when it comes to using a corner horn like the Jub or Khorn.

Corners have real world benefits and drawbacks that need to be understood for any loudspeaker system/room and it's learning how to balance the options that will help to get the most out of any loudspeaker/room system.

Real world test.........Place any loudspeaker (even the Khorn LF and Jub LF) in the corner of any room and play selective frequency tones (100Hz to 500Hz) through them and walk around the room. You will experience SPL variations that are far from even..! It's a very rare room in my experience that when you place a loudspeaker in the corner that at least one or two room modes don't show up as problems that will need to be addressed if maximum performance in any given room is to be acheived.

mike tn

[mustang guy]

http://forums.klipsch.com/forums/p/120269/1212300.aspx#1212300

Caution should be used when looking at these plots and drawing conclusions from them.

Look at the frequencies involved in these plots and you will see that they are more in subwoofer territory and not very complete when it comes to using a corner horn like the Jub or Khorn.

Corners have real world benefits and drawbacks that need to be understood for any loudspeaker system/room and it's learning how to balance the options that will help to get the most out of any loudspeaker/room system.

Real world test.........Place any loudspeaker (even the Khorn LF and Jub LF) in the corner of any room and play selective frequency tones (100Hz to 500Hz) through them and walk around the room. You will experience SPL variations that are far from even..! It's a very rare room in my experience that when you place a loudspeaker in the corner that at least one or two room modes don't show up as problems that will need to be addressed if maximum performance in any given room is to be acheived.

mike tn

That is a very good point. Loading is definitely more effective at some frequencies than others. My THT's are losing some efficiency being in the center pointed at the seating position, but there are definitely fewer dead spots. The whole room seems to be filled evenly.

[Chris A]

Quote

Look at the frequencies involved in these plots and you will see that they are more in subwoofer territory and not very complete when it comes to using a corner horn like the Jub or Khorn.

 

Hi Mike!

 

One concern that I have with the plots is that they say 1.4-2.8 Hz...(i.e., one octave...!). That's probably a little low for most rooms to have room mode issues... Maybe I'm not reading the annotations on the plots correctly.

 

Chris

[garyrc]

Look at the frequencies involved in these plots and you will see that they are more in subwoofer territory and not very complete when it comes to using a corner horn like the Jub or Khorn.

Hi Mike!

One concern that I have with the plots is that they say 1.4-2.8 Hz...(i.e., one octave...!). That's probably a little low for most rooms to have room mode issues... Maybe I'm not reading the annotations on the plots correctly.

Chris

They are not the way I remembered them. While most of the plots are in subwoofer frequencies, the one with the speakers in the corners appears not to be. Chris, I think those are commas, not decimal points on the corner position plot. I think the center frequency is 2,000 Hz., and the octave is 1,416 to 2,832. I think society should eschew ultra-fine print, and get rid of pennies, too.

I'd like to see some plots done octave by octave, from about 30Hz to, maybe about 10K.

[mikebse2a3]

Hey Chris!

I believe Gary is right about the frequency range of the corner plot.

It's hard to know what all parameters were chosen to make these plots but they aren't sufficient nor intended I believe to define/predict a 2 dimensional let alone a 3 dimensional soundfield in a typical room that will contain modal and constructive/destructive interference.

mike tn

[Chris A]

I have updated the FAQ messages in this thread with pictures and increased explanations to make some discussion clearer. Please let me know if there are any concepts that are still unclear and that need to be clarified.

Regards,

Chris

[Delicious2]

Nice job! That Jub theater is classy. On the question of absorption on floor and ceiling. I have bare floor and bare 8ft ceilings now for my Khorns. How far out into the room should the absorption go? 2 ft in front of the Khorns? 4 ft? I assume the ceiling should match?

Considering using cut pieces of area rug both on floor and perhaps velcroed to the ceiling.

What shape should they be in? For example I can imagine the absorptive rug pieces on the floor forming a triangle in front of the speaker, or a square or even extending all the way out to the side walls.

[Chris A]

Quote

How far out into the room should the absorption go?

 

 

There is some additional information on this subject that says that the first 0.7 milliseconds (0.0007 seconds) should be covered with absorption, which translates to ~9.5 inches (~24 cm) of absorption on the sidewalls out from the edge of the midrange horn mouth where the Khorn midrange horn is in closest proximity, otherwise you get stereo image distortion. Try it on each side with a pinned-up quilt or a comforter on the side walls to see it it firms up the stereo image. Any more absorption from that point should help with focusing the image clarity. I use 2 x 2 foot squares stacked 3-high on each sidewall. Since I have a lot of issues between the speakers with hard brick reflectors, I use the 2 x 2 squares liberally on the front wall and on the sides and top of the center Belle Klipsch, as shown in the picture.

 

For Khorns, you will probably benefit greatly from placing a throw rug on the floor in front of each Khorn, or even one big rug across the front of the listening area. This is a high WAF exercise, but can get expensive depending on the tastes of your interior decorator.

 

The ceiling can be treated about halfway between to your listening position and each Khorn, Again, you can use fuzzy material pinned to the ceiling in order to hear if the stereo image moves downward toward the floor and coalesces. It will also help to hear the LEDR "up-and-over" demonstration track on the Chesky Records Jazz Sampler and Audiophile Test CD. If you try about a two-foot-square patch first, and then keep extending the coverage area it to see where the benefits reach diminishing returns. Then you can decide whether 1) you can afford it, and 2) whether you can afford it (...in the WAF sense...).

 

Also remember to put something fuzzy on top of the Khorns to let it stick out in front ~2+ inches to absorb the spilling 400-~3000 Hz acoustic energy coming out of your midange horns. You will hear a bit of a timbre shift - for the better, since you are absorbing mostly the energy coming out of the midrange, and the vertical off-axis energy spectra are what I'd call "heavily red-shifted".

 

The shape of the absorption areas is not really that critical - it's the %-coverage area on the nearby reflective surfaces that really counts.

 

The room will feel a little bit smaller across the front, but the resulting imaging and timbre improvements should go off the chart...

 

Chris

[Panacea Engineering]

The Fuzies on top of my K Horns, I think are Cob Webs....I can see them swaying in the acustical breeze. Also the half inch of dust acts as a great dampner. Ah.Such is the life of a bachelor....! I have stuff in here the CDC hasen't even classified yet. If it doesn't crawl out of the fridge on its own, it is still edible....!

[Chris A]

Sorry, I'm still married, and hope to stay that way...

 

Try the fuzzy material on top trick, letting it stick out a couple or inches in front of the speaker, then put some good stereo imaging music on and listen carefully for differences in timbre and vertical position of the stereo image. Try something on the floor (if you don't already have rug down) and ceiling.

 

The center of my Jubs' stereo imaging moves downward about half a foot when I place absorption material over the flat screen above the mantle.

 

Chris

[Don Richard]

A good, cheap way to control ceiling reflections is to hang 4 ft. X 6 ft. or so of felt on the ceiling so that is attached at the four corners of the felt and droops down so that there is about 18 inches to 2 feet of sag. You could use heavy cloth or drapery material at a higher cost. I have seen this type of thing done in trendy restaurants and it can look attractive if the colors are right. Seems to help bust standing waves also.

[Chris A]

Cool. I might try that out on the WAF...

[Panacea Engineering]

Sorry for "OFF Topicing" the thread....the Fuzzy thing sent me over the edge......[] Maid's been a bit tardy around here.....I tried the WAF thing four times...never worked...me and the Dogs are HAPPY.....

[garyrc]

There is some additional information on this subject that says that the first 0.7 milliseconds (0.0007 seconds) should be covered with absorption, which translates to ~9.5 inches (~24 cm) of absorption on the sidewalls out from the edge of the midrange horn mouth where the Khorn midrange horn is in closest proximity, otherwise you get stereo image distortion. Try it on each side with a pinned-up quilt or a comforter on the side walls to see it it firms up the stereo image. Any more absorption from that point should help with focusing the image clarity. I use 2 x 2 foot squares stacked 3-high on each sidewall. Since I have a lot of issues between the speakers with hard brick reflectors, I use the 2 x 2 squares liberally on the front wall and on the sides and top of the center Belle Klipsch, as shown in the picture.

What kind of absorption panel or pad would you use ... are there any that would be "invisible" to the Khorn bass, i.e., that would prevent the absorption from constituting an obstruction, since PWK warned that the wall should be unobstructed for 4 feet out from the corner. I think Craig Stark [?] preferred 5 feet. I'm not saying that this would be a problem, just wondering ...