Yes. Because there is a distinct way to achieve different goals.
For example, to easily achieve accuracy to voices you may want to avoid notch filter or other complexities, which is required to achieve accuracy in cymbals and may be electric guitars at believable spl.
Complex music require accurate "pitch" which also has different way to achieve.
I think Lynn has set/stated the objective and limitation very clearly on his designs.
Very interesting response. The impulse seems to suggest that the higher frequency instruments will sound a bit in the back and soft, especially instruments like cymbals. The dip is very similar to what the Exaltation has. I wish to have tried flat response, but it was not possible with the simple circuit.
The CSD of the Aerial seems nicely controlled in the lower frequencies. It sould have a pretty fast bass impression.
Would be nice to see some phase data.
Last edited:
I hear what you're talking about ... the ability of a truly high-end system to "slice through" a dense, congested, distorted recording and pull out the singer, and reveal a human being in the dense wall of distortion.
This is a dangerous capability that is independent of natural sound. This may sound like a bizarre statement, but there are systems that have astonishing amounts of "pull out" but also sound brutal and harsh on acoustic instruments. That's a direction I don't want to go in.
The Karna amplifiers have the highest resolution of any amplifier I've ever heard. I don't think this is necessarily all that wonderful, since I'm not going for resolution as a sonic goal. It just sort of fell out that way. The Ariels are also extremely revealing at low, medium, and moderately high levels. So the system as a whole is one of the most revealing I know.
But that's never been a goal; it makes balancing more tedious, since everything sits on a knife-edge, and I never want the system to tip over into a cold, clinical "hifi" kind of sound, which to me is extremely unnatural and nothing like the sound of acoustic instruments.
From what I've heard of the first prototype, it is considerably more revealing than the Ariels, and much more efficient. In a way, there is almost an excess of resolution; there's certainly no point in chasing more from other components. The main task is adjusting balance and removing resonant colorations so the sound remains natural and consistent with the sound of human beings singing and the sound of acoustic instruments playing in real acoustic spaces.
In other words, I don't want more resolution; I want more natural sound, and will adjust the design criteria to get there.
DeonC: What bothers me about the slots is that they are inherently resonant, and rely on brickwall lowpass filtering for acceptable performance. I try to avoid high-Q resonances whenever possible, since there are very definite limits on how much any crossover can reject; some of the bad stuff always sneaks through. My crude test for any driver or enclosure is to audition it with a minimum of filtering, and listen at length to pin down the inherent acoustical and mechanical colorations. If there are hollow or resonant colorations, it's not acceptable, and another approach has to be chosen. For example, the 5.5" Vifa drivers in the Ariel sound quite good with the tweeter disconnected and no crossover at all; the crossover's just there to tidy things up.
Last edited:
Hi Lynn
Thanks for the answer. How then about just mounting the drivers straightforward on the wings. IOW the system in pic but with normal woofers in stead of slots. If you're going for a front horn, that creates a lot of cabinet space behind it. Very nice for two or three woofers on each side. And the sand-loading behind the horn's 'wings' will help reduce cabinet colorations.
The RA servo drivers all have 24dB/octave XOs built into the amps. Is it enough? I really would not know. That I leave up to the guys with more experience than me.
Thanks again for the answer. It is truly appreciated. It makes the effort worth while.
Deon
I think resolution needs to reveal the nature of the original sound and no more. It is true that sometimes one can be too obsessed with the delight of perceived resolution and loss sight of the music itself. It is generally more acceptable to sacrifice some resolution to avoid problems with match various equipment, but if the best is to me acheived, each equipment needs to be carefully evaluated.
Lynn, for those of us who haven't followed this thread from the beginning (and don't have the time to read the whole thing!), could you post a link to the description of your new prototype design? I saw your links in post #1, but I do not know if they are the latest "version" that you are prototyping...
Last edited:
My question in this now semantic discussion is what is your definition of resolution? It almost sounds like not really resolution but actually distortion products creating a false sound. What I mean by that is something like the phony high frequency response of say an aluminum Focal inverted dome tweeter, that high frequency glare that some seem to think is extended high frequency response but is in reality noise that is not related to any original signal.
As Lynn would call it any high Q resonance that is just unnatural but some find revealing of some sound. Is this what we are calling too revealing in this sense, I would not normally use this term this way myself.
As Lynn would call it any high Q resonance that is just unnatural but some find revealing of some sound. Is this what we are calling too revealing in this sense, I would not normally use this term this way myself.
DeonC,
A couple of comments on your bass enclosure design. First I would say that there are probably some who would like this type of design, it does address the extreme low bass that you would not produce from such a short horn. At the same time I would say that this design really doesn't seem related to the original Onken design as that from what I have seen is a folded corner horn design at least originally.
If I was going to build a design like you posted I would surely make sure that the slot loaded speakers are separated internally from the horn loaded driver, this will only cause problems with both sets of drivers due to the interaction from the rear waves modulating each other.
A couple of comments on your bass enclosure design. First I would say that there are probably some who would like this type of design, it does address the extreme low bass that you would not produce from such a short horn. At the same time I would say that this design really doesn't seem related to the original Onken design as that from what I have seen is a folded corner horn design at least originally.
If I was going to build a design like you posted I would surely make sure that the slot loaded speakers are separated internally from the horn loaded driver, this will only cause problems with both sets of drivers due to the interaction from the rear waves modulating each other.
Hi Kindhornman
Sorry if that was not clear, but I thought that would go without saying. The mid-bass will have a totally seperate enclosure. The reference to an Onken enclosure was just in terms of looks, not overall functioning. The full enclosure will have three (or four if you count the front-horn section) sections, all totally seperate from each other. It will be the mid-bass section, and either side of it the two low-bass sections. This will be so if you use the slot-loading or the regular bass loading. Sorry if this caused any confusion.
Deon
Last edited:
MF/HF: Radian 745Neo with either aluminum or beryllium diaphragm, or Altec/GPA 288.
MF/HF horn: AH425 Azurahorn, or larger JMLC models from Poland that have 1.4" entrances and entrance/ext angles that match the Radian or Altec/GPA driver (around 7~8 degrees). If larger horns are selected, crossovers in the 500~650 Hz range can be used, depending on horn cutoff.
Attenuator: Precision wirewound discrete-resistor L-Pad, or custom autoformer or transformer with 1 dB taps in the 12~15 dB range. The autoformer uses a 16-ohm shunt resistor in parallel with the primary, and the discrete L-pad presents a 16-ohm load to the passive highpass filter.
HP Crossover: Passive 4th-order 700 Hz highpass, intermediate between Bessel and LR4, for a 16-ohm load. High-quality caps (metal-foil preferred) and air-core inductors recommended.
LP Crossover: Passive 3rd-order 700 Hz lowpass, Bessel type, set for Z minimum of 15" woofer, which also has Zobel inductance corrector.
Woofer: If the system is a simple 2-way, an Altec/GPA 416 Alnico, 16-ohm version, in a 5~8 cubic foot vented box, with vent on the floor. Gary Dahl has measured the T/S parameters of the current-production GPA drivers and will be posting them. (Hint: Qts is 0.25~0.26 after break-in.) Others woofers are fine if you have a personal favorite.
More elaborate 3-way version: Altec/GPA 515 Alnico for the 160 Hz to 700 Hz range, in a 2.5~3 cubic foot closed box, or similar-size open-back box mostly filled with recycled cotton UltraTouch. If the latter approach is chosen, then a passive 2nd-order 160~200 Hz highpass filter is recommended to minimize excursion. Something I'm looking into is adjusting the frequency of the baffle peak of the open-back box so it becomes part of the electrical highpass filter.
The 30~160 Hz region is covered by a pair of 15" woofers in a separate closed box, on the left and right sides of the box, with reaction-cancelling rods going between the mounting holes of each driver. Bi-amplified with its own amplifier, active crossover and room equalization. Since the box is on the small side, Linkwitz transform equalization is used to synthesize the desired bass response.
The simple 2-way version has been built and auditioned, and performance is satisfactory. FR is +/- 1.5 dB (with no peaking), and impulse response is comparable to the Ariel.
Last edited:
For multiple listening positions, generally yes.
For absolute performance for a single listener, no. That goes to near-field stereo (..and can be good into the lower mid-range). However, even near-field mono can be better, and practically speaking: near-field mono could be placed (with multiple mono-sub.s) at each listening position (which necessarily is a distributed solution, but the principle goal is the near-field effect).
In fact, for out-right performance and *value* for a single listener it's tough to beat a good sub with fully variable phase adjustment placed right behind the listening position (..where the driver/ports are within 3 feet of the listener's head).
Lynn,
I follow what your design and appreciate your trading information as you do. I only have one question that I keep thinking about when I read your posts. When you specify a 16ohm driver is that only because you plan on using tube amplification, or is there another reason that you are using this old standard?
I follow what your design and appreciate your trading information as you do. I only have one question that I keep thinking about when I read your posts. When you specify a 16ohm driver is that only because you plan on using tube amplification, or is there another reason that you are using this old standard?
Tube amps with 4, 8, and 16 ohm taps don't care about impedance, so that's not a factor.
The choice of 16 ohms is a little embarrassing: it's to save money on the crossover capacitors, which have to be twice as big for a 8-ohm system. Air-core inductors aren't that expensive, and I don't recommend that people use the expensive and inferior-sounding flat-wire inductors.
The Ariel is a 4 ohm system, and works just fine with tube amps ... just pick the 4 ohm tap on the output transformer.
Yes, it's a knife-edge, but I believe strongly this is the final goal. Why systems "tip over" is because there is insufficient understanding of how the elements need to be refined and stabilised - and the latter is always about system troubleshooting, not at the individual component optimising level. Always, IME.
For me, a system that can't go straight from, say, Foo Fighters at room punishing volumes, to the classic, plaintive girl with guitar thing, and not get both right hasn't been fully sorted out. This is the capability I'm always refining a system to be capable of, to handle the extreme energy material and with no change do the pure, simple tonal soundscapes beautifully. In both cases it's reduction of low level distortion to inaudible levels that makes it possible, and currently the techniques for achieving this are not well understood, and normally not properly implemented. Hence, the belief that one needs one system to do the houserockin' stuff, and another for the sweet, innocent, quiet material. IME, this is not the case ...
Thanks for bringing this up, IslandPink.
Some of us tread a fine line in the diyAudio forums. A number of posters are long-time industry professionals ... Dr. Geddes and others come to mind. Others are on the periphery of the industry; that's where I am now, ordering custom-build devices from key vendors, and maybe at some date in the future offering products for sale. JMLC is a contributor to the industry, but makes nothing from his horn designs, offering them free to all.
All of us on the periphery, or directly involved professionally, have special relationships with key vendors. If a vendor offers unique products and I feel they are ethical business people, yes, I'll mention them. Tribute transformers; O-Netics transformers from Bud Purvine; Dave Slagle; Alexander of RAAL; custom design and construction services from Gary Pimm; Martin of Azurahorn; Bill of Great Plains Audio (which is the real Altec of today, since they are the same staff and production and repair facility); Radian; Materion (Brush Wellman); and several others I can't think right now.
The moment I cross the line and start offering products for sale to this forum or the public, the rules change. I'd expect to start a manufacturer's sub-forum at that point, and steer the conversation in that direction.
Some of us tread a fine line in the diyAudio forums. A number of posters are long-time industry professionals ... Dr. Geddes and others come to mind. Others are on the periphery of the industry; that's where I am now, ordering custom-build devices from key vendors, and maybe at some date in the future offering products for sale. JMLC is a contributor to the industry, but makes nothing from his horn designs, offering them free to all.
All of us on the periphery, or directly involved professionally, have special relationships with key vendors. If a vendor offers unique products and I feel they are ethical business people, yes, I'll mention them. Tribute transformers; O-Netics transformers from Bud Purvine; Dave Slagle; Alexander of RAAL; custom design and construction services from Gary Pimm; Martin of Azurahorn; Bill of Great Plains Audio (which is the real Altec of today, since they are the same staff and production and repair facility); Radian; Materion (Brush Wellman); and several others I can't think right now.
The moment I cross the line and start offering products for sale to this forum or the public, the rules change. I'd expect to start a manufacturer's sub-forum at that point, and steer the conversation in that direction.
Last edited:
Lynn,
I appreciate the position that we are all in here on this thread and others on this forum. I can't say how many times I have written a comment and then self deleted that as I decided not to post something negative about what I have read or about a product. Sometimes I do it just to make a point about a certain phenomena that most of us can identify with. But it is just to make a point, not to disparage a company. We do need to walk a careful line here, otherwise this would just be a shouting match like I do see in other threads at times.
I appreciate the position that we are all in here on this thread and others on this forum. I can't say how many times I have written a comment and then self deleted that as I decided not to post something negative about what I have read or about a product. Sometimes I do it just to make a point about a certain phenomena that most of us can identify with. But it is just to make a point, not to disparage a company. We do need to walk a careful line here, otherwise this would just be a shouting match like I do see in other threads at times.
High Q resonance is something you will see in CSD plots. But it is not the only measure associated with perceived resolution. Generally it is necessary to first identify the issue through the experienced listeners, then identify the problem through measurement or engineering assessment.
Another thing to consider is to what kind of listening loudness you are designing for. If you listen at same loudness settings used in studios, then the response curve should be flat. But we generally listen at levels much lower, so your design needs to take ear sensitivity into consideration. Working on a new amplifier, the idea I have in mind is to add some form of tone control to adjust the balance of speaker frequency response using simple controls, this gives some flexibility to fine tune to the listening levels one normally listens at. The idea came from a local store owner because he felt that sometimes tone controls are useful, I just needed a better technical reason for it, and did find one. The key is maintaining a list of criteria and specs for the first prototype, then figure out what is right or wrong.
The approach Lynn seems to be taking is more loosely controlled which is a good approach when developing new technology. For example, if you want to find out how to best match one particular compression driver and a horn, then it is more like doing a matrix of matches to determine the best way to identify it from analysis through design, and to measurement. In the end, you get your specific criteria for driver measurement and horn properties to select the best match so that you can reduce the trial and error process in future designs. Bear in mind that almost all horn designs assume a plane wave front with uniform velocity profile at the throat which few compression drivers can generate, so you need to figure out how to make sense out of this besides the other issues involved.
Looking at Lynn's data on the Aerials, it seems to me that just implementation without having that large inverse peak in the impulse should improve resolution. Would be interesting to see what the impulse of the current design looks like.
Last edited: