Pete,
I'm not quite certain, but I think you'd do well to read through a few of the longer running threads concerning various horn types, including Dr. Geddes, the JMLC, and maybe the Synergy/Danley designs.
What I'm trying to get across is that this is not a trivially simple subject, despite what Wayne "pi" wants people to believe. So, checking out some stuff is not a bad idea, before getting too far into the deep end of the pool. (Beware of fresh water sharks... piranha too!)
That would get you started with an overview, and what some of the considerations might be.
You're not going to slap some random driver, compression or "dome" (whatever) on some random horn and get results that will be much more than random.
Btw, an aside, but what "truth(s)" did Aczel discover? Very curious about that.
Did you say that he was a guiding influence for you?
...sorry that I couldn't follow up right away. But I guess the discussion has been perked up... btw, what is/was the dirty little secret??
I'm not quite certain, but I think you'd do well to read through a few of the longer running threads concerning various horn types, including Dr. Geddes, the JMLC, and maybe the Synergy/Danley designs.
What I'm trying to get across is that this is not a trivially simple subject, despite what Wayne "pi" wants people to believe. So, checking out some stuff is not a bad idea, before getting too far into the deep end of the pool. (Beware of fresh water sharks... piranha too!)
That would get you started with an overview, and what some of the considerations might be.
You're not going to slap some random driver, compression or "dome" (whatever) on some random horn and get results that will be much more than random.
Btw, an aside, but what "truth(s)" did Aczel discover? Very curious about that.
Did you say that he was a guiding influence for you?
...sorry that I couldn't follow up right away. But I guess the discussion has been perked up... btw, what is/was the dirty little secret??
Well let's see Aczel is my elder, but far behind me in discovering the hifi industry is full of BS. What he us is one hell of an intellect, who should have caught on sooner and Gods gift to audio journalism once he did, taking no prisoners. Not an influence because if I would have met him before he wised up it may have been the other way around had I had an outlet at the time. The secret well its not dirty is it now, its being answed is it not, and you are one of the ones who can. (-:
OK, so what was the question?
What!!! 😱 😱 You can't be serious!? I am devastated!
Yeah, it "works"... the math works fine.
The implementations are just now becoming, how can I put it, "listenable".
Which is fine by me.
Pete, my man, don't drink the Kool-Aid?
Sorry if I don't share the same admiration for Mr. Aczel as you do.
To me he came off a bit like Consumer Reports... the way they tested hi-fi gear and speakers was laughable... but otoh I never took the comic book "equipment-of-the-month-club" publications seriously either. So, what does that mean?
You do know that Columbus "discovered" that the New World was already inhabited and civilized with operational societies? Lewis & Clarke reported back that the Northwest was already inhabited and chock full of people all the way out the the coastal regions of the Pacific ocean. It's good that you before Aczel made your discoveries... sorry for the sarcasm. It's not polite.
Back to the horns... what was *your* "dirty little secret"?
Was it that you say they need "EQ"?
Or?
I lost the thrust of this.
_-_-
The implementations are just now becoming, how can I put it, "listenable".
Which is fine by me.
Pete, my man, don't drink the Kool-Aid?
Sorry if I don't share the same admiration for Mr. Aczel as you do.
To me he came off a bit like Consumer Reports... the way they tested hi-fi gear and speakers was laughable... but otoh I never took the comic book "equipment-of-the-month-club" publications seriously either. So, what does that mean?
You do know that Columbus "discovered" that the New World was already inhabited and civilized with operational societies? Lewis & Clarke reported back that the Northwest was already inhabited and chock full of people all the way out the the coastal regions of the Pacific ocean. It's good that you before Aczel made your discoveries... sorry for the sarcasm. It's not polite.
Back to the horns... what was *your* "dirty little secret"?
Was it that you say they need "EQ"?
Or?
I lost the thrust of this.
_-_-
gedlee;3[LIST=1 said:
Ok then I'm willing to go there, fine. Yes it works this well:
You know those old round antique Keele CD horns from the 70s (and many other horn designs that actually load the driver low enough to bring the crossover down below the place were the ear is most critical) those that put that critical area, out of the reach of destructive lobing, allow the driver to excurse less and therefore have less distortion? (supposedly inaudible to most people)
Well it may turn out that with a little more processing filter power than a few coils caps and resistors (Yeah that DSP) those old horn designs just not be obsolete at all and in fact coupled with DSP instead of simple analog filter, might just be superior. So the secret is that one and the question perhaps posed to the wrong person is "Is it possible that diffraction is not so bad after all? Is it possible this trade off of not loading the driver to at least 500 Hz is not a good one? I'm wondering if diffraction is a straw man in a way, and is further studies on the audibility of harmonic distortion (renamed I suppose) are not warranted. I'm not so so sure about this. I have a K12 here and this old SP1 with DSP. Hands down the SP1 over B&W and the K12, it is cleaner sounding. I think among other things I'm heating far less dion from the od antique Peavey. Hopefully I've edited this into clarity from this phone.
Last edited:
Being polite is tough when you have tough questions. To summarize I'm not sure diffraction is so bad, and I'm not sure harmonic distortion is all that inaudible, neither of which means that waveguide speakers are not great speakers. I do not think the door should be closed on other horn speakers a possibly being superior when combined with DSP. I'm betting on the point source thing being the most important issue based on what I'm seeing from my old Peavey horns behaving nicely 500 up with DSP the vultures circle but I know there gas been some clarity brought to others here in the middle of this small pile.
Laugh, but imagine if all of the people who don't know that suddenly were made aware. You and a few others would have to square your production. I really do have a problem with some of this crap, therefore I ask the kind of things I do. Some of the crap I see makes the stupidity of some of this little puddle of butt pale, and at least people can read about things that ate worthy of disagreement on as opposed to grown men debating ohm's fn law.
Last edited:
1. It's the converters that have acheived maturity, the math has been squared away for 15 years.
2. DA conversion and than another ADDA is still questionable at best. You must still enter your DSP in the digital dim
That was not my point, not at all.
I have studied horns and loudspeakers that use them for several decades, and if anything, I am prone to perfectionism and a tendency to overanalyze things.
However, I would suggest that they are pretty simple systems, usually no more than a few moving parts. Certainly not something so complex they cannot be discussed in detail, even in a public forum. My words were intended to encourage discussion, not stifle it. Basically, my comment was intended more to say, "bring it on."
I was replying to the assertion that a compression driver does not create a plane wave. The wavefront generated approximates one, but not exactly. And even if a compression driver created a perfect plane wave, that would not be the right shape for all horns.
So to carry on the discussion, one might ponder exactly how much the wavefront distortion at the source matters. I mean, we all (myself included) get pretty wrapped around the axle when talking about horn flare shapes, but then again, we all use the same drivers, which present a wavefront that we know is not ideal.
It has always seemed to me to be something of a circular argument. We all know the source wavefront isn't perfect for the interface, yet we talk about how important the interface is.
We can't have it both ways. If we're going to count the number of angels on the needle, we cannot overlook their size. Or if we're willing to let that part go, then it isn't realisic to talk about how "pristine" our flares are. Each is a compromise, and I think the best we can hope for is to compare measurements. Because if we fall back on hypothetical angel counting, they're all imperfect. There is no 100% solution, just a handful of 90%-ers. And of course, plenty of 50%-ers and 70%-ers too.
Pete
These are all very good questions. I asked them myself many years back. How will you go about getting the answers? Not around here since few here do much valid psychoacoustic studies. So unless you do them yourself, these questions will just have to go unanswered in your mind.
In my mind the answers are clear, but you've already rejected those.
1) Diffraction is bad in as much as it may be benign at low levels, but sounds worse at high levels. Having compared the same drivers equalized flat on conical and exponential horns with smooth throats compared to "pinched" diffraction throats, I have a strong preference for the former.
Distortion is bad if you desire playback to sound like the recording or live source. Odd order harmonic distortion is worse than even order.
Compression drivers have both, and both increase with throat SPL.
Larger diaphragm compression drivers with larger throats do not require as much throat SPL for a given SPL at the listener position.
2) Proper implementation of DSP can correct many problems, but can't correct poor horn designs.
3) A single point source is desirable to avoid vertical response changes from a seated to standing position.
A large center to center distance between the LF source and HF source as in your Peavey SP1 means you must be several meters away before you can have a smooth transition from standing to seated, at a distance of one meter or so, the vertical response is too ragged for mastering or critical listening unless your ears stay in the same vertical plane.
That said, at a few meters in the usual listening position the difference between a single point source and a closely spaced 15" and horn probably would not be a factor.
Having compared a DSL SH-100 (8" co-ax Synergy) to a Mackie HD1521 (two-way speaker with built in amps and DSP) found that with the SH-100 you could literally move your head around within a few inches of the horn in any direction and not hear tonal changes, while the HD1521 required a few meters for as much vertical movement to not be noticed.
However, within the HD1521 horizontal pattern as long as your ears were on axis between the horn and 15, polar control was excellent.
Now, sound and cost- at $1000 street price, the HD1521 is about 1/2 what you would pay for the SH-100 and an amp.
The SH-100 is not nearly as smooth, and does not go as low as the HD1521.
I have been told that the SH-50 is in a different class compared to the SH-100, it does measure smoother, and is about three times the cost.
It does get louder than the 125 dB the HD1521 can do in the real world, but more than 125 dB above 60 Hz is not generally needed in a home or control room environment.
That said, the $1000 HD1521 measures way smoother than the SH-50 and damn near anything I have ever seen, regardless of cost.
The HD1521 can be driven in to hard limit and still sounds good.
Why, with decent, but not exotic components can the HD1521 sound so good?
Because the DSP implementation stems from Dave Gunness.
Over his career at EV (he was responsible for the EV MT system design) and EAW, where he learned how to make some decent horn designs sound much better through "Gunness Focusing", patented acoustic correction DSP algorithms specific for the horns and transducers used.
EAW was bought by Loud, who are the owners of Mackie, Dave's secrets have trickled down, he has figured out how to correct most of horn and loudspeaker's "dirty little secrets".
Dave now runs Fulcrum acoustics.
After A/B/C tests with the HD1521, SH-100 and some of my designs, I realize that other than designing for specific niche applications, there is no reason for me to build speakers any more, I can't compete in terms of cost/performance.
That's my "dirty little secret".
P.S. The Smaart measurement of the HD1521 shows a ground bounce dip at 200 Hz, and the "High" control was turned up a few dB from the standard setting, which looks more like the spec sheet curve above 10 kHz.
There is no smoothing on either the SH-100 or HD1521, with smoothing the HD1521 looks like the spec sheet.
Art
Attachments
Last edited:
Pete,
My opinion is that diffraction generally speaking is bad, and that EQ is not always required in a horn. At least not the sort that you are thinking of, the DSP method of attempting to provide an inverse correction to make everything minutely "perfect".
It's probably nearly impossible to take something (a speaker) with inherent flaws and apply "pre-distortion" (read: DSP correction) and make it work so well that the inherent flaws melt away. I can see a time when there might possibly be some sort of incredibly sophisticated and dynamically variable means of applying digital corrections that might possibly ameliorate certain aspects effectively, but that ain't just yet.
Wayne Pi, I think we're on the same page more or less...
...if it matters at all my sense of this is that Pete is looking for a problem upon which to apply a (DSP) solution, more or less.
Oh and fwiw, my own horn system runs with a single compression driver down to the rolloff of the horn (250-300Hz range) now, without any EQ required at all. It's rather flat in fact. So, running compression drivers down even below this range is not new. Just gotta select the proper driver. How high you can run also depends on the driver selection as does the max SPL and distortion vs. SPL. But done well generally speaking the distortion is pretty low and the max SPL higher than almost all "normal" hi-fi speakers. Fyi.
Dr. Geddes and I may differ in some regards as far as practical approaches to a hi-fi system, but I certainly can't argue with him on theory.
My opinion is that diffraction generally speaking is bad, and that EQ is not always required in a horn. At least not the sort that you are thinking of, the DSP method of attempting to provide an inverse correction to make everything minutely "perfect".
It's probably nearly impossible to take something (a speaker) with inherent flaws and apply "pre-distortion" (read: DSP correction) and make it work so well that the inherent flaws melt away. I can see a time when there might possibly be some sort of incredibly sophisticated and dynamically variable means of applying digital corrections that might possibly ameliorate certain aspects effectively, but that ain't just yet.
Wayne Pi, I think we're on the same page more or less...
...if it matters at all my sense of this is that Pete is looking for a problem upon which to apply a (DSP) solution, more or less.
Oh and fwiw, my own horn system runs with a single compression driver down to the rolloff of the horn (250-300Hz range) now, without any EQ required at all. It's rather flat in fact. So, running compression drivers down even below this range is not new. Just gotta select the proper driver. How high you can run also depends on the driver selection as does the max SPL and distortion vs. SPL. But done well generally speaking the distortion is pretty low and the max SPL higher than almost all "normal" hi-fi speakers. Fyi.
Dr. Geddes and I may differ in some regards as far as practical approaches to a hi-fi system, but I certainly can't argue with him on theory.
Isn't that what Kahneman is talking about? Although the answers seem to be quite clear, there's no objective data to support it.
Not exactly. I'm not entirely bewildered (-:
There was a hilariously funny little blues songs from the 70's by Tommy Collins. The gist of which is he is asked to be more specific, he would really rather not but..... "Opal, you asked me"
Opal You Asked Me - YouTube
"I reject your answers" Not exactly, not my words either, but since you ask...
What I reject is generalization and obfuscation via generalization, whether intended, not intended, or by accident, which I consider highly unlikely among this brain trust.
1. We all know what harmonic distortion is, IM distortion, all sorts of distortions, hell some of even know what real transient distortion is. Not enough of know what dynamic distortion is and that one is RAMPANT, yet here are generalizing distortion by lumping it all into the "non-linear" distortion pile. That way it is easier to simply say "people are insensitive to non-linear distortion" In this manner more people are likely to nod and go umm hmmm "The Emperors clothes look great" instead of saying WTF?
Well I'm the guy who goes WTF? Non-linear distortion is when the damned output differs from the damned input, this now makes it pretty obvious that we have to draw a line somewhere. The industry has drawn that line decades ago and we all know that around 500 to 2-3k we are extremely sensitive to all sorts of distortion, including harmonic especially those those above 2nd.
We used to use big efficient horn speakers that kept the crossover point low and loaded the horn appropriately well into the lower end. These are damned expensive to manufacture and the ruling female class has more or less banned good speakers from whatever house she is allowing the convenient man of the moment to tacitly pay rent on, but I digress.
So most of us really can't (aren't allowed?)to have those gorgeous big speakers that actually do present harmonic distortion levels in the single digits (and even lower in some cases). I saw where one famous DIYer moved your flagship unobtanium units that pretty much does meet all my criteria to his bedroom for gods sake!
So we need to compromise, and here is where the gelding occurs. We move the crossover point higher, we make the horn shallow. We are now well aware that even with computers to do all the tremendous slide rule work that PWK JBL et al had to do (god bless them I'm assuming there is one of those) we are still presented with vertical flat spots and drivers that are excursing just a bit more than we would like them to. Ergo much higher ummm "non linear distortion" (please refer back to the line where I casually mentioned that non-linear distortion is "when the damned output differs from the damned input", and the inevitability of a null somewhere right in the sacred passband.
There is a line in the sand where output not equaling input DOES becomes significant and audible, the industry (all of the industry not just the idiots) has determined this line over a period of decades and decades, Electronics are no longer an issue DSP is no longer an issue as these are at the point of "straight wire with gain that you may bend to any shape you desire"
Speakers are and will continue to be THE issue. Rooms are and will be the issue, but assuming our goal here really *is* constant directivity over the entire bandwidth, we can leave that aside for now I suppose.
OK so here we are. You want to provide us, the endangered species who purchase said units, with a great speaker. W,e the endangered species who wish to buy these great speakers also have to make compromises.
You E.G., have decided that that compression drivers are one of the bestest compromises ever. I concur. Apparently however, the dimensions allowed by the ruling class (we call that WAF here right?) and the economy they have created do not allow that luxury. So what the hell move the metric? Not needed, the speakers sound great and speak. let them do so.
2. Diffraction. Thank goodness, I think this takes less time to think through. Straw man. We are stuck with diffraction, since every time the wave splashes on something you in effect get another speaker. Here is what I think counts. The diffraction, which *will* occur needs to occur early and not repetitively in the horn. Bumps in the road make things bounce.
3. Studies. No, *I* do not need to do a study on how sensitive the ear is to xdistortion, Neither do I your reject your clarity on these issues, but I certainly do think that broader studies than the one you conducted could be done, but IMO probably not needed. The entire world has been doing that study since at least the 50's, we know well we are dealing with single digits and percentages in most cases at least from Shroeder up. However *if* a new study is done, It needs to be similar to the one Toole did Huge, did I mention huge? Tightly controlled beyond question and most of all, regarding "non linear distortion" include *lots* of people that do what I do, Industry pros, musicians and hard core audiophiles (well those that are not completely nuts or worse) need to be included in large proportion in such a study if actually done. Hint here: Sane audiophiles can not hear the difference in cables that measure the same. Yeah this costs paper, but Beyond that all bets are off, because people live with 128k Mp3s, that ain't us is it? I really do not think that a small pointed albeit nice study trumps that much historical and empirical world knowledge, but what the hell, I do not have academy stripes, I might have a couple of other creds sitting around somewhere though.
Opal, you asked me.
Last edited:
Opposed to that, there are the well-known benefits of direct radiation drivers and the lesser manipulated sound path 😛😉😀
Again the neo directs are giving much higher sensitivity and justify a look with their other technical improvements.
- Status
- Not open for further replies.