OK
The MEH will never match the smoothness of performance that I get because it has too many internal possibilities for resonances etc. The MEH approach is ideal when very high SPLs are required, which is Tom's target market. But for smooth objective performance in a smaller space where high SPL is not a requirement, I have yet to see one that matches the NS-15 in performance and I don't expect to.
MEH is also quite appealing from a construction standpoint since its waveguide is very easy to assemble. The NS-15 is a real PITA to put together, so for DIY I'd have to say that MEH is very appealing. It's just not the ideal that I was after.
I tried to get Tom to supply me with data on his designs so that we could compare them in my app showing performance across many designs, but it never came to pass.
The MEH will never match the smoothness of performance that I get because it has too many internal possibilities for resonances etc. The MEH approach is ideal when very high SPLs are required, which is Tom's target market. But for smooth objective performance in a smaller space where high SPL is not a requirement, I have yet to see one that matches the NS-15 in performance and I don't expect to.
MEH is also quite appealing from a construction standpoint since its waveguide is very easy to assemble. The NS-15 is a real PITA to put together, so for DIY I'd have to say that MEH is very appealing. It's just not the ideal that I was after.
I tried to get Tom to supply me with data on his designs so that we could compare them in my app showing performance across many designs, but it never came to pass.
Last edited:
These issues would exist regardless of if the MEH is OS?
Don't mean to get too off topic. It's just been said by some that an MEH design is champ of phantom center...
I think that the holes in the sides would always be detrimental to performance even with OS. And the lack of a substantial mouth radius will always be a negative as well.
As to "champ of phantom center" this is not based on any valid subjective testing that I know of.
As to "champ of phantom center" this is not based on any valid subjective testing that I know of.
As usual, I agree with Earl.
Though I've published a zillion Unity horn projects, a lot of it is academic: I'm curious to see how far I can push the boundaries. For instance, the MCM 55-1870 shouldn't work on a Unity horn, but somehow it does if you make the midrange taps a particular shape.
Here's the polars from my latest projects. Though I've bent over backwards to reduce the impact of the midrange taps, it's still there, to an extent. And right now I'm listening to the Kali LP6, which is a bit like a pint size Summa (waveguide tweeter, two-way speaker) and I gotta admit the treble on these things is REALLY nice.
The phantom center on the Cosynes is *ridiculous*, it feels like you can reach out and touch it. On my 2nd set of Unity horns, way back in 2006, one of the reasons I discarded the project after a week was that it made everything sound mono... then it turned out my music is mostly mono, and the Unity horns were the first to demonstrate this solidly.
Whether the phantom center is due to the waveguide, or the crossover, or the MEH concept... That's open to debate, and I honestly don't know which it is.
My 'hunch' is that it's mostly the waveguide and the xover. But this raises an interesting question, because certain crossovers *demand* a MEH configuration.
For instance, look at the Klipsch Jubilee. It's basically impossible to get acceptable vertical polars unless you stick the midranges IN the horn. That's a HUMONGOUS waveguide.
The waveguide on the Kali LP6 is four inches tall; it's not difficult to cross over to a midrange. But once the height of the waveguide is exceeding six inches, it gets really tricky to get vertical polars that are acceptable. There are some tricks you can do with delay and offset to 'thrown' the off axis lobes somewhere besides the floor and ceiling, but everything gets more complicated with the crossover when the midrange and tweeter are approaching or exceeding a wavelength apart. I think with the Jubilees, Klipsch just said "screw the vertical polars."
Of course, this entire topic is one of those things we can go in circles forever. Right now, listening to the LP6, there's a TON of things it does that the Cosyne doesn't, in particular the stage is HYOOOGE. But the phantom center isn't even remotely close. A lot of this boils down to personal preference, many will prefer a big stage over a solid phantom center. (My wife prefers the LP6, but I think she was just relieved to have those giant monkey coffins out of the living room.)
That still sounds like a phase or polarity problem. IME weird phase can make for a very large/wide soundfield but at the expense of a center image. Tricks like that are often used in the "surround sound from 2 speakers!" gimmicks. It does do some wild things.
Also found the crossover specs. The attached is not hard to figure out except that you would need the passive crossover layout. Both active and passive crossovers are in the listing, but the active is turned off.
I'll try and locate the schematic, but as of yet I haven't found it.
Could not upload. The file is in a XML format and has a unique .lss tag that I cannot seem to get rid of. It is all text, but with a .lss ending it will not upload. Sorry. Anyone know how to strip this off? In Win 8.1 it does not seem possible. If I rename it it still retains the .lss. Even if I use "save as:" it retains the .lss - can't figure this out.
I'll try and locate the schematic, but as of yet I haven't found it.
Could not upload. The file is in a XML format and has a unique .lss tag that I cannot seem to get rid of. It is all text, but with a .lss ending it will not upload. Sorry. Anyone know how to strip this off? In Win 8.1 it does not seem possible. If I rename it it still retains the .lss. Even if I use "save as:" it retains the .lss - can't figure this out.
Last edited:
I don't disagree, but the wiring is correct.
Years ago I was measuring some speakers in my apartment, and I set my gates wrong. I discovered something really interesting, which was that the reflection off of a hard surface is a near perfect replica of the original sound.
This was something that just blew my mind; reflections aren't twenty or thirty decibels down, they're nearly as loud as the original signal. That just blew my mind.
Obviously it wasn't a *perfect* replica, it was attenuated by a few decibels and the treble was rolled off. But the reflection off of the wall wasn't a whole lot different than the sound generated by the actual loudspeaker!
Here's my living room, I'm listening to the LP6 right now.
Here's the response of the LP6 that I measured.
For comparison's sake, the Cosyne radiates a fraction of the energy to the sides, due to that big ol' waveguide.
So the LP6 is going to produce a strong 'bounce' off of that backwall, delayed by about one millisecond. 1350Hz is one millisecond long.
I'm doing this off the top of my head, but if I'm not mistaken, the 'bounce' should produce a notch at 2700hz.
Obviously, I should drag a laptop and a mic in here to confirm. But I think that some of the additional ambience, along with a reduction in dynamics, is being caused by reflections off of that big ol' back wall.
I've been kinda tempted to put the speaker stands on cinder blocks like Greg Timbers recommends, but I don't think my wife will tolerate that 😉
For instance, look at the Klipsch Jubilee. It's basically impossible to get acceptable vertical polars unless you stick the midranges IN the horn. That's a HUMONGOUS waveguide. I think with the Jubilees, Klipsch just said "screw the vertical polars."
For home use, the suggested active crossover frequency was 375 hz between the LF bin and HF device.
Here is the raw beamwidth of the HF device.
Only at wavelengths that effectively 'see' the holes, frequencies below that, no ill effect.
Only at frequencies that will 'see' the radius, at the low end of the passband of the device.
All of this can be drawn out in a ripple tank simulator.
Does this help? In windows explorer, 'Tools' menu - Options - 'View' tab - uncheck "hide extensions for known file types". That option causes most problems with trying to modify file extensions. For the very first step, if the file/edit/view/tools menus are hidden, alt+T should bring it up.
Yes, I've had a similar experience when measuring. Those reflections were looking pretty good. 🙂 Don Keele talks about his CBT having a mirror image on the floor that doubles its length. That's easy to grasp when you've seen reflection in the measurements.
I suppose we'd all like to see comparison phase plots of your two different speakers in the room, since they give such different results.
When I play dual mono, my large very ambience-sounding ESL panels which are a few feet from the front wall, produce a razor sharp center virtual image.
Golden-ear theories aside, wouldn't it be true that if you can produce a razor sharp and unvarying virtual dual mono image, you have all the ducks in a row to extract all the stereo information on a recording when you play in stereo? And conversely, if you can't produce a sharp centre image in dual mono, your system just isn't any good?
I've been pondering this matter of stereo recording depth for some months. Obviously, mikes don't capture depth, esp in modern recording practice. Just for this discussion, I think it is fair to distinguish perceptual cues to depth from what might be called practical cues. Recording engineers try to get some practical cues into the tracks. For example, they might have a mic in the back of a hall just to gather echoes to mix in to an otherwise (and almost universally) disgusting dry pick up. Between the skill of their cooking and the active imagination of the listener, you get depth. Listeners with really good imagination can place the oboe in one place and it never moves (even if it really is on the other side of the orchestra in reality).
Historical note: egocentric conductors prevailed on the board to re-engineer Orchestra Hall in Chicago, a few decades ago. What had been a venue for great recordings became useless and recording engineers had to add reverb.
B.
Golden-ear theories aside, wouldn't it be true that if you can produce a razor sharp and unvarying virtual dual mono image, you have all the ducks in a row to extract all the stereo information on a recording when you play in stereo? And conversely, if you can't produce a sharp centre image in dual mono, your system just isn't any good?
I've been pondering this matter of stereo recording depth for some months. Obviously, mikes don't capture depth, esp in modern recording practice. Just for this discussion, I think it is fair to distinguish perceptual cues to depth from what might be called practical cues. Recording engineers try to get some practical cues into the tracks. For example, they might have a mic in the back of a hall just to gather echoes to mix in to an otherwise (and almost universally) disgusting dry pick up. Between the skill of their cooking and the active imagination of the listener, you get depth. Listeners with really good imagination can place the oboe in one place and it never moves (even if it really is on the other side of the orchestra in reality).
Historical note: egocentric conductors prevailed on the board to re-engineer Orchestra Hall in Chicago, a few decades ago. What had been a venue for great recordings became useless and recording engineers had to add reverb.
B.
Last edited:
I believe this absolutely to be the case and yet when I suggest such a thing to an audiophile audience who adjust speakers to achieve some sort of, of necessity, arbitrary notion of the correct "soundstage", usually involving pointing the speakers straight down the room or timidly toed in, as if to appease the toe-in Gods, I am met usually with resounding silence as a response. As I said in an earlier post, achieving a coherent phantom image is all that needs to be done—all the rest is guesswork and wishful thinking, presuming the mixing and mastering engineers have their ducks in a row in their monitoring system—but unfortunately I feel sure that is often not the case.
As for your system being no good if you can't produce a sharp center image, if the speakers are good, especially in terms of pair matching of frequency response, the problem may well be the set up. Is the acoustic environment of one speaker as close to a mirror image of that of the other as feasible, for example? Sometimes across the corner is the best option, as opposed to square on to the long or short wall.
Last edited:
This is completely true. And a comparison with listening to mono via a single loudspeaker will demonstrate the shortcomings in central phantom images that no loudspeaker will likely help. The exception is where wide dispersion adds more room reflections and these limitations are smeared over. But such smearing is not the target of "hi-fi" one would assume?
This bit is not true, however. The perception of depth arises from predominantly early reflections. It is a temporal 'signature' that coincident mic techniques capture well. Where it is not evident (if it was there in the recording to start with!) is also because of (amongst other things) the same smearing in the reproduction chain. Reverberation is somewhat different to depth perception.
Despite all the efforts to afford scientific rigour to the recording process, it will continue to be the subject of considerable artistry. I have been fortunate to sit in the prime seats at several recorded performances, but it never ceases to amaze me that listening to the same recording in the recording engineer's hideaway afterwards affords a better listening experience.