[macrojack]

Right off the top I should state that I am not one of you. I can't solder, read a schematic or drive a slide rule. I did, however, create a horn based loudspeaker that provides a very dynamic, coherent, faithful and intriguing musical reproduction. This was accomplished by getting good advice, being a little creative about assembly and relying on my tried and true dumb luck.

Because I am under the influence of audio fantasy and devoid of the ability to leave well enough alone, I am always thinking about how to make it more better. That, I imagined, makes this the place where I should raise this topic.

JoshK has expressed a longing for a horn speaker project and maybe a few more of you might think to pursue such a thing as well. If enough of you are interested enough, I'm bound to be able to bleed off some info I can use if I can find a translator.

So, who wants to build horns? My story has already been told somewhere in the archives but I'm willing to provide info about my experiences if any one wants tolerate my non-technical bio.

[JoshK]

Currently I am finishing up on my current speaker project which uses a "waveguide" loaded compression driver for 1Khz up.  Its a hybrid of sorts.   I, however, have a growing fascination with horns.  I really want to tinker and experiment with horns.  Like I mentioned in the thread that inspired this thread, it will likely never amount to more than a basement project, but that is fine.

For the purposes of this thread, I think we can agree that we are interested in discussing front-loaded horns (FLH), not back-loaded horns, which is a different animal altogether.  Given how fractionalized the horn community is, that might be all that we agree upon, but that is ok.   I can respect other's decisions or rationale even if it doesn't agree with mine.

Everyone has a different take and spin on FLHs.  There is the conical crowd (you), the OS crowd (me, thus far), the tractrix, the Le'cleach, spherical, etc, etc.  But I think we should see past our differences and realize that we are drawn by a similar passion.  Different solutions based on differences in order of priorities. 

FLHs have a limited bandwidth and anyone looking to do horn loading for a predominance of the audio range is going to accept that the system will be huge.  Most draw a line in the sand where to desire to horn-load the system is met with practical limitations on size allowance.

I think the amount of solutions to the highest frequencies, be it somewhere above 600hz on up, are out there.  There still isn't much in the way of good solutions for the lower midrange, say 100hz-1khz, apart from serious DIY endeavors.  Horns for this range are as scarce as hens teeth, at least outside of the pro audio world.  Its exactly this range I have been most interested in.

I'd like to see more ideas drawn up on how to cover the lower mids to mate with a HF horn that can be accessible to the novice builder (I am assuming we would have to build any solution).  Maybe another larger conical  would be best.  Your thoughts?

[TRADERXFAN]

Is this about crafting your own horn, or just using a premade horn to incorporate into a loudspeaker project?

[sts9fan]

I am interested in a FLH project also.  I have a small area that is going to be a "reading room" that will need a system.  I am thinking one ST,FLH,Woofer,Sub. 

Do good mid FLH have to be 3' long?

[macrojack]

I am interested in a FLH project also.  I have a small area that is going to be a "reading room" that will need a system.  I am thinking one ST,FLH,Woofer,Sub. 

Do good mid FLH have to be 3' long?

Bigger horns can cover a wider bandwidth allowing for fewer drivers and fewer crossover points. Dispersion and driver integration issues can require a larger space as you need to have a certain listening distance for these to be effective.
There's a fellow named John in Philly (Nullspace) who created a very sophisticated OB horn system that you might like to investigate.

http://www.acoustichorn.com/news/index.php?id=3689716252287709688

As an aside to a fellow Zu aficionado, Sean Casey cut his teeth on horns and still thinks they are the best approach. The Zu sound and character arise from his attempts to approximate horn performance with dynamic drivers. Obviously Zu is trying to keep their hardware compact.

[JoshK]

Is this about crafting your own horn, or just using a premade horn to incorporate into a loudspeaker project?

Either.  This is meant as a topic to discuss all things horn speakers.  There wasn't a natural place for this thread and given that it typically requires some discussion of technical topics, we felt it best belonged in the lab. 

[JoshK]

If WAF wasn't an issue I would have jumped on the midbass horns JLH was selling (see his full sized 120hz horn ad on diyAudio).  The typical horn-lover simple doesn't have access to this type of solution, apart from hiring someone like JLH to build them for you.  Maybe the lack of products exemplifies the lack of demand.  Still there are many horn lovers out there.

I don't consider myself yet a horny.  I just have an interest to explore this niche of audio.  I haven't heard eny such setup, apart from the hybrid Avantgarde Duo in a difficult acoustic environment. 

So why horns?  I know some lurkers are probably asking that question.  A lot has been written on the subject, but I'll add my naive understandings of what makes horns interesting. 
1) controlled dispersion.  Horns naturally focus dispersion in a much narrower angle than dynamic drivers in a flat baffle.  This is beneficial in lowering particularly the early reflections from the reverberant field.  This can also be done like Geddes has done it with a large format midbass driver and a waveguide for the HFs.  So all horn solutions aren't necessary to acheive this end.  Its theorized that early reflections are the most problematic acoustic anomalies in reproducing audio in a small room.
2) lower distortion.  This might seem counter-intuitive to many traditional audiophiles who are used to thinking that pro audio drivers have roller coaster frequency responses.  Indeed many do.  But that isn't all that is important, and in many ways is often much easier to deal with then other issues. 
A good example might be looking at typical THD charts of a compression driver.  It is true that most CD have higher than a dome tweeter when comparing straight THDs, but that obscurs the real story.  CDs have a high constant second harmonic distortion (the one your ears don't hear well). 
Where CDs kick the dome's butt is level dependency of distortion and harmonic profile.  Dome's distortion sets in quickly at higher and higher levels.  It also starts compressing quickly with any real power applied.  For a domestic environment CDs have none of these issues.  The increase in distortion with level is practically negligible.  So tonal balance doesn't change from soft to loud.  Standard audiophile two-ways get very aggressive sounding to my ears at louder levels (higher harmonics coming up).  A hi-efficiency speaker like Geddes's (my frame of reference) does not.  It sounds the same just louder.
3) sensitivity.  This is one of the most popular notions of horns.  Everyone thinks horns are all about finding speakers for the flea amp crowd.  There is much more than that.  Truth is, it is exponentially easier to design and live with LINEAR amps that are lower powered than high powered equivalents.  Class A/B, class D, and other classes were all invented to get around the need for more power without creating huge monstrousities of amps that require their own power plant and generate huge amounts of heat.  However, if we only need a skosh of power, then class A amps are tolerable.  Class A amps don't have crossover distortion and typically have more tolerable harmonic distortion profiles.  In short the demands placed on the amps goes down tremendously.  I seriously believe that there is something to simple amp topologies that is lost in the complexity of most higher powered amps.  Notice I said nothing about tubes vs solid state.  I am open minded on that topic, although I enjoy building tube gear.

[JoshK]

Bigger horns can cover a wider bandwidth allowing for fewer drivers and fewer crossover points. Dispersion and driver integration issues can require a larger space as you need to have a certain listening distance for these to be effective.

No offense intended, but I am not sure I totally agree.  Horns are band pass devices of a limited bandwidth.  Some horns have better bandwidth than others but all are limited.  Trying to get too much bandwidth of a single horn is said to be a large part of why horns have such a bad name in popular thinking.

The larger the horn the lower the range it can cover all else being equal.  But this usually comes with the top end rolling off as well.  Its a balancing act.  I think this is what requires that the end user knows more than required in other areas of audio, unless plunking down a large sum of money for a turn-key solution. 

I say first have a very clear idea of what range you wish to cover with the horn.  Then size the horn to that range if doable and take care of how the dispersion of the two adjacent drivers matches with the horn at the crossover region.

[macrojack]

Josh - Don't worry about contradicting me. Anything I say on here is a result of supposition. I do not pretend to speak with authority on scientific matters.

 Before I chose to own these horns I had very little experience with any kind of horns and had only heard Cerwin Vega, Klipsch and JBL in home settings and was too closed minded about the subject to actually discover their strengths. I guess the AvantGarde, Jadis, Oswald Mill, etc. phenomenon opened my eyes. I started thinking that those crafty Japanese audiophiles that bought up all of our Western Electric and RCA stuff might just have known what they were doing. And well-healed audiophiles of my acquaintance who could afford Wilson and other such were beginning to buy, enjoy and endorse them. This must be a good thing. How can I afford to get in this game?

Are there others among us who have not thought about horns because they are too expensive to buy and too difficult to make? Or does everybody out there think they honk?

[Niteshade]

I use Klipsch RF-83's in one system and believe they're a very,very good speaker. 

In what sense are you talking about horn speakers?:

1. The tweeter
2. Tweeter & midrange
3. Folded horn woofer
4. Full range driver

[JoshK]

Waveguide is often used in connection with ideas popularized by Dr. Geddes.  He proposed mathematics for modeling of waveguides, waveguide theory if you will, instead of the the traditional webster equations.

For the layman, use of the word waveguide is more in reference to what your goal is than the device itself.   WGs are generally constant directivity (horns are often not).  Dumbed down even further, often WGs are more about controlling the dispersion characteristics rather than acoustic loading of the driver. 

[sts9fan]

I built a pair of Gedlee speakers and love them.  With nearfield listening what differences in sound would one expect going from the OS waveguide to a conical horn using the same driver?

[JoshK]

That is a very contentious issue.  One which I'd rather not touch.  I know what the theory says, but I've never heard the comparison myself. 

[sts9fan]

Ok so what are the camps?  OS is .... Conical is ....

I chose the Gedlee speaks because I liked the science behind them.  It made sense to me.  I did not really research other horn types. 

[macrojack]

Here's what Bill Woods says about Conical designs:

http://www.acoustichorn.com/tech/conical/index.html

I don't even know what OS stands for so I can't help you there. It isn't Operating System, is it?

[sts9fan]

Oblate spheroid

[macrojack]

Oblate spheroid

WOW!!! I was just thinking that "Waveguide" sounded like a proprietary  euphemism. Oblate spheroid just totally eclipses it though. I'm headed for the dictionary. Stay tuned.

[sts9fan]

The shape is what it is.  If I remember correctly Earl states in his white papers that waveguide is just another name for horn. 

[konut]

See JoshKs 2nd post on some types of horns. OS is 1 type. What the OS horn does, that is different from the others, is to limit diffraction artifacts. Is this important? Depends on who you ask. There are monster threads at DIYAudio on this very subject by leading proponents in the field. There are no easy answers. In my experience horns are like any other component in audio. Most are crap or mediocre, but those who know what they're doing can make some extraordinarily great sounding gems. And, like any other subject in audio, there are compromises in any design.

[nullspace]

I chose the Gedlee speaks because I liked the science behind them.  It made sense to me.  I did not really research other horn types.

I chose conical horns because I met Bill Woods at a show, liked both him and his horns. I did not really research other horn types.

Which is my way of suggesting that there are most likely few people that have heard all or even many of the different types of horns; it makes it tough to have an educated opinion -- mine certainly isn't.

Of the basic types (exponential, hyperbolic, tractrix, spherical, conical, oblate spheroid) conical & OS are actually pretty similar from a theoretical perspective, the difference being the throat transition. Specific examples tend to be different, as conicals you see out in the wild tend to be fairly narrow, 50-60deg, whereas the OS waveguides are usually 90deg and feature a large transitional roundover at the mouth.

Generally, though, I think that if you keep in mind what you're looking for from your horn & compression driver, any of the major types can work if you make informed decisions. For instance, I have a smallish room and a decent power response is something I value. I don't think a big 300hz horn would work in my smallish room, so I'm using a horn with a 1000hz cutoff and crossing around 1.2khz. Since I like a decent power response, tractix is definitely out. I think a narrow-directivity like the 50deg conical I have does better in a narrow room than a wide-directivity horn like Dr. Geddes' 90deg OS waveguide. So that's how I ended up where I ended.

Regards,
John