HLCDs, or "horn-loaded compression drivers", are a type of high-end car audio speaker design stemming from pro-audio experimentation in the late '80s. They have evolved into a science, and over the years, have been used with good results by many SQ competitors. They offer several advantages over conventional speakers, but there are some disadvantages as well.

So what is a "horn"? HLCD horns are meant to play as single point sources for all upper midrange and treble frequencies in a typical system. A horn has two parts: a compression driver with magnet structure, and a horn body. They are designed to mount under the dash of most cars, firing parallel to the floor from the far left and far right under-dash locations. The compression driver uses a moving "diaphragm" to produce sound, and by horn-loading this driver, the air space inside the horn couples with the diaphragm of the driver and allows the small element to move a larger volume of air than normal. This increases the sensitivity of the horn speaker in comparison to conventional cone drivers. It's the same principle used by cheerleaders when they scream their cheers at the game thru a megaphone (you know, that cone-shaped thingy). Typically, an HLCD speaker has a sensitivity of over 100dB/1m/1W, whereas most conventional cone drivers in the same frequency range are about 10dB lower. Currently, the manufacturers of mass-produced HLCDs include Illusion Audio, USD Audio, Veritas, Image Dynamics, Crossfire, and JRSpeaker (Crystal). All of these different horns are built on the same principles, but differ somewhat in driver selection, horn material, and horn geometry.

Horns for in-car use are built on the premise that near-equal path-lengths are optimal, and by design, the compression drivers of these horns are placed well under the dash in an attempt to equalize their path-lengths. Furthermore, the actual horn itself not only increases the driver sensitivity, but also serves to control the dispersion pattern of the speaker's sound wave. Car horns are actually "exponential-dispersion" horns, meaning at different axis positions they have different loudness levels. While I am not well versed in "horn-guru" lingo, I will attempt to explain in laymen's terms.

Car horns have several parts: the mounting flange, the "mouth", the throat, and the compression driver orifice at the far back. I won't waste time here trying to explain what they look like, but realize the horn mouth is the actual speaker opening and fires toward the listener. Also realize the throat of most standard-size horn bodies is several inches long, placing the actual compression driver well behind the front surface of the speaker, increasing the path-length. The mouths of most horns are approximately 2.5" high and 12" wide. When installed, this mouth is about 2-3" from the far side of the dash, and spans across longitudinally toward the opposite side of the car. The opposite side horn should be located in a "mirror-image" fashion as well.

A horn's interior shape is designed to direct the bulk of its sound off-axis toward the opposite listener. Recall what we discussed in the "path length /intensity trading" topic. Horns are built to do that for you. If you look at, say an Image Dynamics full-size horn body directly in front of you, you will notice the interior of the horn widens as the "mouth" goes toward the opposite side. You'll also notice the flange inside is chamfered and the opening very small toward the nearside of the mouth. By designing the horn in this fashion, ID is able to control the loudness levels across the mouth of the horn. I am not sure the exact distribution of the sound wave, but I believe it is 30% on axis and 70% off axis. Regardless of the actual numbers, what we need to realize is that horns are loudest toward the center of the car, focusing the bulk of their energy off-axis and creating good imaging and center focus. This is what most of them are designed to do. The proper terminology for this design is "controlled-dispersion", where the speaker itself is designed to have a specific dispersion pattern and axial response characteristics.

A typical system that employs horn drivers is as follows--- HLCD horns for right and left mid's and highs, a pair of midbass/midrange speakers either in the kicks or doors, and subwoofers somewhere. Many guys choose to do a true 4-way set-up with dedicated midbasses and dedicated midrange drivers to accompany the horns. Either way can net good results.

The frequency response of horns basically depends on two things; the size and build material of the compression driver diaphragm and the size and shape of the horn body. At the top end, horn freq response usually hits a brick wall at just under 20KHz due to sound wave interactions and reflections inside the horn body, but this often isn't audibly noticeable if the system is tuned correctly. On the other side of the spectrum, horn body size greatly affects low-end frequency response. The bigger the horn, the lower it goes. The ID mini-horns are the smallest available on the market. And being the smallest, their freq response goes down to only about 900 Hz at best. Step up to a full size horn body and they are able to play down to around 600 Hz. Some Veritas models have spec'd freq response of lower than 600 Hz, and there ARE guys out there with custom-built horns that play down to 400 Hz (Richard Clark, Mark Eldridge, and others from the "old school"). Recall that the "optimal" front stage speaker is one that can play all frequencies from a single point source. Well, HLCDs are designed to do ALMOST that. By playing all of the upper midrange AND highs from a single driver, they negate all the ill-effects of having a crossover point in the mid-to-high transition as well as having likely different path lengths between the midrange and high freq drivers. Couple this to the "controlled dispersion" pattern of the horn body, and you have the ingredients for superb imaging from the get-go.

So what are the benefits of using HLCDs instead of conventional drivers? First of all, you are assured of good image placement across the sound stage from the start, at least from the upper midrange and high frequencies. Second, the increased loudness of horns allows us to power them with a smaller amplifier compared to the other speakers in the system. Third, they can likely be installed in most automobiles and do not take up much space in the kick panel areas. Fourth, when mounted properly (firing parallel from under the dash), they couple sonically to the bottom of most dashes, and this *can* help project the sound upward, giving the sound stage decent height. In terms or dynamic output, no conventional speaker can match the dynamics of good HLCDs, but there are ill effects, as we shall discuss in a sec. In comparison to installing conventional drivers in kick panels, it is easier to install horns and get great staging characteristics simply b/c horns are specifically designed to give you near-exact stage results in a "drop-in" fashion.

As mentioned, there are drawbacks to using HLCDs in an SQ system. We discussed some of the design goals of under-dash horns and what we can expect in terms of performance. We can also expect tonality problems, which stem from the "megaphone effect" in the upper midrange/low treble region, as the internal shape of the horn body, which naturally boosts frequencies in the 3KHz region, affects the sound wave. This lends to the "nasal" sound character of virtually all horn drivers. The top end of the frequency spectrum is also affected in horns and results in a sharp frequency roll-off above about 17KHz, but only well-trained ears can sense this deviation b/c the perceived high frequency brightness of most horns masks this effect. On an RTA this roll-off is very pronounced. This is one reason many competitors using horns supplement the high end with an additional set of tweeters, especially for RTA testing.

Since horns are shaped to emit the bulk of their sound wave toward the opposite listener (great for imaging), the outer portions of the stage suffer from lack of width. I quick comparison of different mfgr's horn designs shows how this problem materializes. Veritas horns are the lesser offender in stage width as their design funnels a higher percentage of sound to the left and right as compared to the direct on-axis position. This is very tough to explain in words, but if you look at the internal shape of a Veritas horn mouth, it resembles an hourglass, with a very small cross-sectional opening firing right at the listener, but to the left and right, the opening enlarges, moreso toward the center of the car. USD horns are simply curvilinear in shape, firing smoothly toward the center of the car. They have no decrease in opening height to speak of, and thus the direct wave tends to be a little too loud. A technique that has been used with these horns is placing a block of fiberglass insulation or other absorptive material partially across the horn mouth to mask the on-axis sound wave just enough to allow for proper dispersion

Image Dynamics uses a decreased cross-sectional opening to attenuate the on-axis response, widening it toward the center. This directs the bulk of the output toward the opposite listener, but there is no "flared opening" directing sound outward to speak of. Illusion Audio uses both a decreased opening height on-axis AND small "vanes" inside the horn body in it's throat to direct the sound. While not as small an opening as the ID, the Illusion exhibits similar characteristics, but does make an effort to widen the dispersion toward the outer border of the vehicle. The Crystal and Crossfire horns are very similar to the USD design, using a curvilinear shape to direct the sound wave off-axis.

So, what does all of this mean in the real world? It means horns are great for imaging and getting the elusive "center image focus", but are not great for stage width. This is a byproduct of design factors and controlled linear dispersion. Conventional drivers have uniform dispersion that is not linear, and as such, they can greatly improve stage width in comparison. In the SQ judging lanes, stage width is determined by sonically localizing the far left and far right imaging cues and determining where this sonic boundary is located in the vehicle. The A pillars are most often used as a reference here, and pillar to pillar width is a good goal to have when designing your system.

The best systems can portray stage width beyond the pillars, and if the system is determined to have a width INSIDE the pillars, score deductions result. Seldom does a horn car have a stage that spans pillar to pillar w/o adding additional speakers to supplement them. Most horn car soundstages only span from about 2" inside the a pillars at best. Many guys using horns in the lanes will add a set of tweeters, either on the A pillars or sail panels, to help them with width and RTA. They tend to cross them over at 16KHz and up so these tweeters can pick up where the horns begin to roll off. While this does assist the width to a degree, it only does so in the higher frequencies, and seldom does it affect midrange imaging and width. A superb SQ system will have a stage that is wide no matter what frequency is played. This is one instance where conventional drivers have an advantage over horns.

To take this a little further, most high-end horns only play down to approximately 650Hz. The remainder of the front stage frequencies must be picked up with a dedicated midrange and/or midbass speaker of conventional design. Most horn cars use midranges placed underneath the horns in the kick panels and are angled for proper imaging characteristics. Sometimes, the midrange or midbass is placed in the door locations and consequently will have a large difference in path lengths. While HLCD systems take care of the upper octave x/o point that most conventional component sets have, they introduce a NEW x/o point in the low midrange, and the resultant phase shift that occurs at this point can and will be troublesome to some degree requiring "tweaking" to compensate.

Because the horn will project it's frequencies upward and can couple with the dash's front side in doing so and because the mids in the kicks do NOT have a controlled dispersion, many HLCD systems suffer from what a few audiophile SQ judges call a "layered" sound stage. What this means is you get a frequency-dependant stage height where the notes in the sound stage appear on top of each other. You get high frequencies at the highest level of the stage, the upper midrange just below that, lower midrange even lower, and so on. Sometimes it can sound like the separate instruments are playing above or below one another. This is most prominent in the low midrange/midbass region, where you can tell the highs are nice and high (near eye level) but the preponderance of the low notes seem to come from the floor.

Very careful tuning and mid/midbass speaker placement is vital to overcome this phenomenon with HLCDs. We must also keep in mind that mounting a full size horn under a dash will decrease the dispersion area available to the midrange drivers mounted in kick panels. This can also affect how "high" the mids can project their sound in the stage.

Another problem with horns is their frequency response curve. We've all undoubtedly heard the so-called "myth" that horns need massive EQ to sound good. Well, this is a misunderstanding, b/c they can sound "good" with minimal EQ. What we should say is that "horns need several bands of EQ to smooth out there freq response to a proper curve". Due to their design, the sound waves emitted from the compression driver encounter many different interactions inside the horn body as the sound travels outward. There are reflections and resonations occurring at several different frequencies, and even the horn body itself can resonate (this is another problem, and some guys actually damp the horn bodies with dynamat to combat this effect).

The resulting freq response curve of HLCD drivers consists of a series of several peaks and dips, some of which are several dB deep. It should be common knowledge that in order to achieve excellent SQ we must have a response that has smooth transitions from freq to freq or the system just will not sound realistic, warm, or natural. Equalization is vital to smoothing out HLCD response curves, and it is for this reason that most horn competitors utilize 1/3 octave EQs in their systems. When you couple this to the fact that the response also will depend on the car's acoustic properties, the need for an EQ heightens.

I am sure we will touch on some more areas of sonic problems with both designs during our discussions, but for now it has been suggested I do a comparison of HLCDs to conventional drivers, just a quick glance at the pros and cons of each design. We will break it down into categories, which are important in a competition system, so here goes:

Tonality Out of the box, conventionals are better. With proper tuning, equal tonal characteristics can be had with both designs, although the top end of the frequency spectrum is most realistic using conventional drivers.

Dynamic Output Horns rule, hands down. Remember, they are 10dB or so MORE efficient compared to cones. Only a creatively mounted on-axis conventional system can rival HLCDs in dynamics.

Position to sound stage Good results can be obtained with either. Depends on tuning and speaker location. Stage Depth Again, good results with either depending on above, however, many listeners will feel like HLCDs provide an "in your face" kind of sound with little delineation of stage depth cues. Again, it depends on tuning.

Imaging Very tight focus of upper midrange and high freqs goes to the HLCDs. Cones can also image perfectly, but can often need much more effort to do so. In either case, "proper" imaging (that is, ALL frequencies) is a byproduct of careful driver placement mainly, in addition to all of the topics covered in the previous SQ articles.

Stage height This is a toss-up and depends on system design and driver placement (and negating that pesky layering effect I mentioned). The best horn systems can portray an eye-level stage, but so can systems with kick panels and especially dash speakers.

Stage width Conventionals Ambience Most guys feel conventionals, having a more natural sound and sound wave dispersion pattern, can produce more "lifelike" ambience.

These are a few of the main categories; we can discuss these and others as the time arises. Pictures of a typical horn installation will follow. The important thing to realize is we are comparing two totally different speaker designs, each with it's own strengths. So whether to use HLCDs or not is a question we all must figure out for ourselves. Try to listen to properly tuned cars using both designs and find the sound you like, and then run with it!

An Alternative View
author- Lee Cao

Also known as compression horns and wave guides, HLCDs represent what could be the biggest waste of money in car audio. The only benefits of HLCD are good dynamics, high efficiency and better imaging.

However, the disadvantages are numerous. HLCD have frequency response from approximately 1kHz to 20kHz. Thus they are essentially large tweeters with very good low frequency extension.

However, they cost anywhere from $300 to over $1000 for a pair. So by themselves, the HLCDs already cost as much or significantly more than most competition quality component speaker systems. You will also have to purchase a separate set of mid bass drivers to cover the frequencies below 1kHz. A pair of quality mid bass drivers start at about $150, adding more to the cost. Installation of HLCDs are very costly and time consuming. Labor quotes can exceed $1000 just for the installation and setup of a pair of HLCDs.

The installation location of HLCD is usually under the dash. Even though this position allows for a more focused image, the overall sound stage is much narrower than conventional tweeters mounted on the doors. The tradeoff for high efficiency is poor transient response, tonal accuracy and overall frequency response. There are mid-bass drivers that have better transient response than HLCDs.

The frequency response and tonal accuracy of HLCD are so bad that it's generally accepted that in order to make a pair of HLCD listenable, one would need 30 band 1/3 octave equalizers, which cost no less than $300. And even then, the sound quality of the HLCD are not even comparable to a pair of good conventional tweeters.

In summary, HLCDs are expensive and they exhibit all the characteristics that we do NOT want in a tweeter, especially one that would require the owner to spend thousands just to get mediocre results.