What I described is the path that allows reliable analysis. You will be hearing with no reliability in judgement of why you are hearing what you are hearing in the first place.
The jmlc does have a proper roll off in all directions. You should create a list of expectations, that will help people, help you.
You are getting a little overly worried. The vertical off axis of the bi-radials are not optimal according to technicality but obviously its not a deal breaker considering some of the best horn designers on this forum have favor towards them. Matter of fact, just save yourself the time and go with the trends and opinion of the much more experienced and get the Yuichi/4001 horn. At least you can be confident in the choice of the people with much more experience.
There isn't a flawless option btw. Every horn you think about will present another set of pros n cons. My advice was to choose the biggest, quality, one, you can afford and to use a healthy dose of room treatment. This will get you your healthy dose of direct sound that you claim to desire.
https://audiohorn.net/next-gen-bi-radial-horn/
Certainly beyond dispersion issues, probably the biggest fear which most have coming from cone speakers is how colored or honky some horns may sound. Troy Crowe claims that his horns are free of this. But what is it about the geometry and other factors which causes those nasty problems? And how bad might it be with the A290 and TH4001?
I wouldn't say that. DIYers routinely design horns with low reflection.
However issues such as reflection are not simple to apply, not all reflections are equal. Also, they tend to work on a sliding scale. It makes no sense to say they don't exist, but it may be so that they are below audibility.. or perhaps below audibility at low volume settings at least 😉
Not only that.. it also has to be practical, achieve certain goals and meet certain constraints.
Understood.
Even with my stunted grasp of those CD plots, those for Audiohorn biradial looks way better with that JBL driver. But would it remain that good with the Radian 745Be-and when equalized for extended HF response?
Furthermore, contrary to what you've been suggesting all along, this is hardly a wide horn. That page says that the biggest version is 36cm; less than 15 inches, or about 9 inches smaller than the other two horns. And numerically, what do you suppose they mean by "mid-distance usage"? Sitting 11 ft from the speakers with them placed 2.5 ft from the rear wall?
As for crossing with the Altec 416, what's meant by "Effective crossing area: 650hz"? How would that relate to the Altec's ideal 500Hz Fc?
Also, owing to the horn's design, would the 4" Radian 951Be still be a better choice than the 3" Radian 745Be, especially for crossing with the Altecs?
However, measurements are one thing and the horn looks to be too new to the market and/or too costly for users to have done actual listening sessions even though I could handle those prices.
https://www.diyaudio.com/community/search/1848228/?q=audiohorn&o=relevance
Last edited:
No, its not coming from that POV at all. The issues created by a very wide dispersion angle is not being able to fully control the reflected sound, specifically the midrange sound which is much harder to control being longer WL. Treble is fairly easy to tame and in some cases can help create a little "spaciousness" which isn't always a bad thing. When you have more of the midrange coming along for the ride, it tends to become unbalanced due to how different this range acts on various materials in the room. This is when you get issues with difficult to treat small cavities and hollow spaces. These all create resonances of their own and can be hard to pinpoint, often feeding off each other. The worst culprits here are sheet rock resonances, window frames, hidden return springs and heating ducts. The surface is easy to treat, but these other sources can sometimes have you come close to pulling out what's left of your own hair. If you're like me, a string instrument player, you'll have your prized possessions hanging proudly in your room. Those can be a whole different nightmare to deal with acoustically than anything else, as they're designed to resonate by design, mainly around 220 hz (A) but rather wide banded.
When I had my electrostats, this was no issue whatsoever because the sound coming from them was highly directional. Mind you, these were very large Soundlab speakers which could throw close to 110 dB at you within reasonable listening distance. The backwards sound was absorbed by large traps and panels of baffled wool, so practically all of the sound came out of the front, focused right at the listener. These ESLs weren't your typical low output design, being able to convey a big drum kit right into your chest without the help of a separate subwoofer. They were flat to 30 hz without any room gain. The beauty of this type of directional speaker is being able to play music loudly at night without bothering anyone. You can't do that with wide radiating speakers, not to mention needing highly directional (cardioid) bass.
The opposite of the above is a minimal baffle speaker with very wide radiation, depending on the limitations of the drivers employed. You could go through lots of effort getting a speaker like this to produce a very even FR throughout the entire range, but the temporal issues it would likely render highly percussive styles of music difficult to enjoy.
In the right setting, a wide radiating system implementing effective and strategic acoustic treatment, could in fact be a suitable system for listening to specific types of music ie. large orchestras, string ensembles, opera, large choirs pipe organ music or other music highly reverberant in nature.
For highly analytical listening, wide dispersing speakers are only suitable if the room is carefully and thoroughly treated. To most people, this sort of room would be too dead. At minimum and in most cases, it would be expensive and visually intrusive to live with (WAF aside). You'd have to be heavily invested in your system, being willing and able to provide the treatments needed to make it all sound accurate, both temporally and FR balance wise, especially in spectrum decay. Transient rich music would definitely be challenging to reproduce. A directional speaker would solve or avoid most of those issues. Just using DSP alone won't fix it to the point of not requiring extensive room treatments, not to the extent of it being suitable for critical listening.
These are the 2" PVR D3220Ph drivers which have been outfitted with the hybrid Ti/Mylar diaphragms. They have a shallow throat exit angle and sound very extended for an off the shelf reasonably priced driver. With the diaphragms they were only $160 each. They badly needed a layer of wool felt in the chamber to stop some of the hash above 8k. Otherwise they're exceptionally smooth considering the price tag. I'll measure them again once they're back together with properly aligned VCs.
And BTW, I align most VCs in compression drivers using THD measurements, not just relying on manufacturers machining tolerances being accurate enough to just blindly bolt them on. There have been times I needed to sand a spot or two on the pole plates from them being out of center.
And BTW, I align most VCs in compression drivers using THD measurements, not just relying on manufacturers machining tolerances being accurate enough to just blindly bolt them on. There have been times I needed to sand a spot or two on the pole plates from them being out of center.
Attachments
The best place to ask question is likely here.
https://www.diyaudio.com/community/search/1848578/?q=4%22&t=post&c[thread]=382609&o=relevance
Bur having the Radian 950 and the TAD 4001, did you usually find that 4” drivers are generally preferable to 3” for a two-way speaker? Understandably, it’s because the 4” can cross lower towards the woofer’s ideal crossover point. But are there other reasons, like reduced likelihood of lower midrange beaming? But downsides, such as crossing that low would preclude a two-way system? Other trade-offs either way?
I didn't look at the dimensions, I would think that they could make a bigger horn upon request, most horns are scalable. Profile wise, you are looking at probably one of the best horns available according to modern philosophy. With that being said, I personally am not convinced that constant directivity horns are a step forward, but rather a sep to the side. A smooth off axis transition without peaks or nulls is a higher priority. Horns with a smooth rising DI (Tractrix, Jmlc, Exponential, etc) are acceptable in my book. They pose limits on the listening window, but using the beam width described by the polar one could predict whether the resulting window size meets their needs. These horns tend to image better at the top of the register due to less indirect sound creation in the room, also, very large rooms and well dampened rooms tend to have a tapering power response so a horn with a rising DI sounds more like those acoustical environments when used in a small room even before acoustical treatment. Large rooms and well dampened rooms tend to attenuate HF more so, but lose the affect as we get lower in the spectrum. There are situations where that generalization falls apart.
I think that elliptical horns are a middle ground between circle horns and the rectangle biradial horns. Their polars are something below
Not to detour you, it seems that you are set on by biradials and I think it's a good choice. I would look into the modern biradial that I posted above and see if they can make it in a larger size and if not, just go with the Yuichi/4001 horn. You also have to think about availability/cost.
I think you're in the same place I am with much of your system, recognizing the weaknesses and trying to address them using a logical path of modifications which don't call for an excessive amount of money and effort. Unfortunately, much of what needs to be done to make this work out affects something else which is co-dependant on other major changes. Its basically all tied together in a very specific way.
I've also sort of hit a wall trying to plan the supporting midbass to the A290. I discovered that trying to mate a direct radiator to a horn further down in.frequency range is much more difficult than just tacking on a super tweeter to a smaller horn. Theres so much more to consider and go wrong working in the critical midrange around the A290s preferred crossover frequency. Anything between 300 and 1000 hz is very delicate, especially between two different types of drivers like the A290 and another direct radiating cone with abrupt directivity changes.
The problem with driving a highly reactive load like a large compression driver and horn combo is the hard time the amp has absorbing the EMF, even with a highly inductance linearized driver. The very thing which gives the midrange the harmonically rich sound is what fights you trying to make it sound tighter in nature. This is why I decided to go solid state fet based amplification with the fets in charge of voltage gain and bipolars doing the hard high current drive. I went with the Parasound JC5 and never looked back. It fulfills all these needs and never leaves me wanting more, being a high bias AB design driven with fets and bipolar outputs. Not cheap, but when you factor in how much I've spent on other stuff trying to get close, its actually a bargain. Only drawback aside from price is the power consumption. It sits there, idling at 400 watts of power. The huge plus is being able to drive virtually any speaker without needing to ever buy a different amp, unless of course you want to biamp...
The issues with your onken enclosures are exactly the ones I've run into. Most boundary placed port walls tend to emit more inner cab pollution already sharing a reflective boundary. Placing ports away from enclosure walls helps alot. Lowering the LP point also helps, but then you need a separate midbass driver to cover the missing bandwidth. This is however a blessing in disguise not having the midbass / midrange potentially leak out of the ports. It also tightens up the midbass having a lower compliance enclosure which is easier to dampen. The dampening can then be relaxed on the lower bass enclosure which IMO can potentially improve reproduction of fine details in the bass. Dampening the bass enclosure is a delicate balance of getting it just right at the desired frequencies. That can take a long time to optimize depending on your level of OCD. Some people practically spend a life time trying to get it right and their tastes can also change over time. It doesn't help that even the smallest changes like air temperature and humidity can affect it.
Based on experience I would say relatively high source impedance seems to have an effect on response flatness (generally addressable with DSP) what is less clear is the high level of THD and it's audibility and masking effects. When I originally built these amps they had 0.45% thd at 1W, and a little under 2% at 6.5W just below clip. These amps have input transformers which might be contributing significantly to the distortion spectra. I have been working on super low distortion headphone amplifier designs, and the latest one which is in the final stages of debug and listening sounds really good - it has the highest feedback margins of anything I have designed and my CFP diamond buffer output. I am planning on something vaguely similar with Exicon lateral fets in the output stage (also a diamond buffer)
The woofer is a 515-8G and the QTS is 0.19, I have been warned that this driver will not work in most ported boxes, the Onken being a notable exception. It was really intended for horn loading. This one is very much up in the air for the moment.
The woofer is a 515-8G and the QTS is 0.19, I have been warned that this driver will not work in most ported boxes, the Onken being a notable exception. It was really intended for horn loading. This one is very much up in the air for the moment.
Hi Profiguy,
We are completely on the same page! I have been going through all of what you described and even briefly considered a dedicated mid-bass driver, but there was no where to put it!
I have combing issues with the bass and harmonic and fundamental shoot through the ports. Crossing at 675Hz although I found crossing at 600Hz mitigated those issues by a few additional dB, things sound better at the higher XO point.
We are completely on the same page! I have been going through all of what you described and even briefly considered a dedicated mid-bass driver, but there was no where to put it!
I have combing issues with the bass and harmonic and fundamental shoot through the ports. Crossing at 675Hz although I found crossing at 600Hz mitigated those issues by a few additional dB, things sound better at the higher XO point.
Last edited:
I also came to the conclusion that extending the low bass at least over 100 hz avoids much of the detached sound you can get from too much group delay down low. The phase lag many ported boxes have lower down can really ruin the transient response around the second impedance peak. I've seen it wrap around over 200 degrees in some cases with a softer LP filter slope. That creates the weird, separate location sub woofer sound that's hard to fix without a very steep filter on the HP.
The pipe dream of that perfect wide banded 2 way horn system in many cases turns into a nightmare trying to realize. There are only a hand full of drivers which are capable of getting close to this idea. You already have those unicorn TADs which IMO are likely the pinnacle of large CDs capable of the extended top end on the A290s while going low enough to mate with a large cone or midbass horn.
The pipe dream of that perfect wide banded 2 way horn system in many cases turns into a nightmare trying to realize. There are only a hand full of drivers which are capable of getting close to this idea. You already have those unicorn TADs which IMO are likely the pinnacle of large CDs capable of the extended top end on the A290s while going low enough to mate with a large cone or midbass horn.
Last edited:
I made a modification to the power amplifiers last night that seems to have largely addressed an unmeasurable but very audible glare in the mid-range. I suspect if I spent enough time I could probably find something to do with spectrum of distortion that the 4001s are all too obligingly reproducing. I am going to try moving the cross back down to 600Hz and see what that gets me. (Measured great in the last iteration, maybe now it will sound good too.)
This is all too true, and in my neck of the woods there was absolutely no alternative than to plunge blindly ahead. It's been a rough ride, but even auditioning something in someone's home before buying can still lead to unexpectedly poor results in your system/space. Regardless of choice you are going to have to work for it. I am going to go out on a limb here and say between the two it probably matters most in a smaller room - the Yuichi is reputed to perform marginally better in a small space than the TH-4001, if you are closer than 5 meters listening distance I would say that pushes you more towards the Yuichi otherwise both are reasonable.
Loose the idea that these are not wide horns otherwise you will end up in a heap of trouble in an untreated room. They are much wider than the 90° horns I used in the past. The key to good performance with these is attenuating the early reflections and to find the best compromise just keep adding treatments until either the room is over damped (remove some) or you are happy with the result. Consider hanging diffuser/absorbers from the ceiling in addition to diffusion and absorption on wall surfaces. Bare floor isn't great unless the room is pretty well damped already, carpet is recommended, but it is not very effective below a few kHz (and then not great) If you can splash on a professional acoustical consult that may not be a waste of money.
In my 3.5m wide room things were bad until I treated adjoining walls and surroundings. (I also treated behind me, and behind the speakers. It immediately made things better. The impressive thing here is that tonal balance remains very similar over a wide area. The room is still too small, and some of the best sound I have heard down here is in the bathroom behind the listening space. It's mono, but sounds right. 😢
Last edited:
Valuable descriptions of room acoustical problems; thanks, though lots of bad news, some which is very confusing given Pierre's system, who I don't believe has done much if any room treatment; see posts 15266, 15276. https://www.diyaudio.com/community/threads/beyond-the-ariel.100392/page-764 Obviously, Pierre has wide dispersion speakers, which he continues to rave about. The only flaw they have he said are "midrange aberrations" from woofer beaming due to the 3" Radian745Be and 15" TAD 1601B diaphragm differences.
But unless you had heavily damped your room why on earth did you swap out those amazing SoundLab ESLs for the A290, TH4001 or some other wide dispersion horn? Or are you using a narrow directional horn like JMLC or even something like Troy's? https://josephcrowe.com/products/3d-cad-files-horn-no-1994-es450-biradial-for-jbl-2446-2-throat Indeed, as I think Pierre mentioned in those two posts, he was a long time JMLC Azurahorn 425 user in that same room. He loved that "ESL + dynamics" sound which Lynn Olson raved about from the 425s-as Gary Dahl still does (Yamaha JA6681B), whose Altec midwoofers I cloned. https://galibierdesign.com/wa-trip-01/
However, Pierre eventually wound up preferring the wide sound stage of the 4001 horns. But speaking of low level listening, when asked if he finds himself wanting a center channel speaker for movie/TV sound, Pierre said he's mostly quite satified, though said that a recent session with a friend's high end home theater system shows that a center speaker definitely gives better dialogue coverage. Speaking of which, between Gary's Azurahorn 425s and Troy's ES450s, which do you believe would give better center channel dialogue coverage?? And if not a tie, please explain why one's the winner. Movie sound reproduction would be as critical as music to me if so many of vintage movies and TV shows I have on DVD and BD weren't of med-fi quality at best. Uusally over compressed and what often almost sounds like a bandpassed or midrange boosted FR. So, as I may have room for two pairs of speakers + subs in my room, I was hoping for one pair of horn speakers (JMLC 425 or ES450?) for movies-and good enough dialogue without need of a center speaker-and one pair (A290 ir TH4011?) for music. But I will ask Pierre how much he treated his room and how badly reflections impact the sound.
Wow, you have awesome restorations skills. If you, Radian and Rewind could find a perfect replacement for those beryllium"fingers" a lot of 6681B owners would be forever grateful. https://www.diyaudio.com/community/...aha-ja-6681-compression-driver.163469/page-27
The SB Audience 65CDNT would likely be my only choice over the B&C DCM50 in that ES450 horn. https://www.diyaudio.com/community/threads/sb-audience-rosso-65cdn-t-test-review.383917/
@oltos I know its hard to understand why I went from something like the big Soundlab ESLs to a horn setup with the A290s (athos clones) and now searching for the complementary FLH bass. I used to live in a smaller place with the ESLs and didn't have the real estate to have larger speakers which could fill the room to listen from a consumer POV and not always an engineer's perspective. I used to edit and master material for smaller acoustic acts, solo string and wind instruments, percussion, etc. I needed the ESLs to hear all the details which a dynamic speaker would swallow. I was very critical about using very little if any processing unless absolutely necessary. The Soundlabs were the only ESLs which could keep up with a larger dynamic multi way speaker while sounding neutral and accurate. They were basically just a big pair of open back headphones. I ended up selling them due to lack of work and wanting something more "fun" to listen to. Combine this with limited time left on this planet (terminal illness), I wanted something a little more substantial to enjoy, plus giving my son something meaningful to inherit and remember me by.
The driver optimization I do on the higher end side is due to having dealt with very picky clientele demanding a higher level of performance. Some companies like TAD, Faital, Fostex and Altec require hand centering of diaphragm assemblies using THD measurments. Even many mass produced drivers require tweaking by ear and measurments. Running drivers off of SETs and other minimal feedback amplification make the driver more prone to unwanted movements (lateral at some resonances) in the VC gap. The driver has to be able to cope with rocking diaphragm modes in these situations without impedance compensation. Left unchecked, it can damage the driver in some instances.
The A290 horn is IMO one of the most capable horn designs for a large 2 way system that sounds more like a live performance taking place in your living room. It has a very natural sounding midrange which isn't as "horn" sounding in character as some other large format wide range horns. That is of course assuming the use of a driver designed to work with it ie. the big TAD. I originally was using the B&C DCX50 which sounded very good, but the top end was wonky using the original adapter plate. It requires a different plate to get the fullest potential from the DCX50s. I may end up using a separate super tweeter if I have to unless the big PRV with the hybrid diaphragm can handle the treble better on its own. This driver is a very good performer above 10k, unlike many other large diaphragm 2" drivers which can't manage over 10k, even on axis. The TAD is of course the preferred driver for the A290, but as @kevinkr has already mentioned dealing with this very same issue, the Athos A290 clone adapter plate isn't fully optimized to achieve the best possible performance from this combo. I'm experimenting with other drivers due to lack of funds, hoping I'll come across something which will work decently without much butchering or hacking.
The driver optimization I do on the higher end side is due to having dealt with very picky clientele demanding a higher level of performance. Some companies like TAD, Faital, Fostex and Altec require hand centering of diaphragm assemblies using THD measurments. Even many mass produced drivers require tweaking by ear and measurments. Running drivers off of SETs and other minimal feedback amplification make the driver more prone to unwanted movements (lateral at some resonances) in the VC gap. The driver has to be able to cope with rocking diaphragm modes in these situations without impedance compensation. Left unchecked, it can damage the driver in some instances.
The A290 horn is IMO one of the most capable horn designs for a large 2 way system that sounds more like a live performance taking place in your living room. It has a very natural sounding midrange which isn't as "horn" sounding in character as some other large format wide range horns. That is of course assuming the use of a driver designed to work with it ie. the big TAD. I originally was using the B&C DCX50 which sounded very good, but the top end was wonky using the original adapter plate. It requires a different plate to get the fullest potential from the DCX50s. I may end up using a separate super tweeter if I have to unless the big PRV with the hybrid diaphragm can handle the treble better on its own. This driver is a very good performer above 10k, unlike many other large diaphragm 2" drivers which can't manage over 10k, even on axis. The TAD is of course the preferred driver for the A290, but as @kevinkr has already mentioned dealing with this very same issue, the Athos A290 clone adapter plate isn't fully optimized to achieve the best possible performance from this combo. I'm experimenting with other drivers due to lack of funds, hoping I'll come across something which will work decently without much butchering or hacking.