Congratulations on necroposting a nearly 3 year old thread to try to tell Tom Danley of all people that he is wrong.
I thought it had been determined by tests, oh so long ago. Of course it's going to depend a lot on the Equal Loudness curves, tones, speech, music, noise, etc.
Agree that it's not really precise.
+1 bel is twice as loud. There are 10 decibels in 1 bel. Guess who it was named for?
Esmeralda
Yes, but the thread has a lot of good summary information on multiple entry horns usable in a home space--info that apparently hasn't changed in the 4 or so years since the thread's inception. It's a useful starting point for those that haven't been following or doing their own multiple entry horns.
For me, I'd add only one thing that seems to not been discussed here: the effect of wider coverage angles--at least horizontally--than 60 degrees. Pano, I think that you might have one dilemma at least partially addressed (i.e., why the Synergies sounded better in a much smaller room). I'd start with Toole's book as a place to look at this little-discussed subject, and the evidence that wider angles (up to about 90-100 degrees) reportedly are preferred by listeners in blind trials. Some rooms do better with narrower coverage due to room acoustics and early reflection issues. Other rooms (much larger--like commercial-sized rooms) tend to surface other issues that are just not experienced in a home-sized room environment.
I also believe that a lot of folks look at perhaps minutia rather than first setting the stage for the application: they're looking at the loudspeaker itself (devoid of application), or perhaps further detail in its horn profile, or crossover, or a certain driver within, or even get wrapped up in something as esoteric as efficiency or sensitivity, etc. But then everyone talks past each other and oftentimes miss the total (system-level) view of integrating everything into a home-sized space for humans to listen to (hi-fi). They're making assumptions that aren't valid for others.
This is what I'd call "relative ranking of requirements for the application domain in which they're to be used". This I find to be much more useful when talking about all the pieces of a design and its evaluation.
Chris
Last edited:
Cask05 +1
Silly seeing speakers that could only be decent in a large room with luxuriously extensive floorspace, but get stuffed into tiny ones. Or gigantic horns close enough to the only listening chairs that you could almost eat dinner off the bottom horn wall. I once visited a guy who had humungous Martin Logans stuffed right up against the back wall, sure didn't sound thousands of dollars good....
Silly seeing speakers that could only be decent in a large room with luxuriously extensive floorspace, but get stuffed into tiny ones. Or gigantic horns close enough to the only listening chairs that you could almost eat dinner off the bottom horn wall. I once visited a guy who had humungous Martin Logans stuffed right up against the back wall, sure didn't sound thousands of dollars good....
When we have a look at the the equal loudness curves one can reasonably assume that the doubling of SPL (i.e. +6dB) is eqivalent to the a doubling of the perceived loudness.
Well - that depends on the case. While large speakers can be very fun to listen to in smaller rooms due to their generous headroom, there can be indeed some sacrifices in other "departments" - but that is always depending on the speaker type. The ML placed close to a back wall is indeed a very bad example IMO.
I am currently developing the crossover for a large MTM that will be used in a not-so-big listening room. With the first informal tests I could see that it is not the size of the room that gives the most problems but furniture and parts of the building structure that are rattling.
Regards
Charles
I owned a pair of Magnepan MG-IIIa's in the 1980s in a fairly large room with sloping ceiling 8'-20' behind them and plenty of room around them--6 feet or more from any wall. That turned out to be insufficient in size. I couldn't toe-in the speakers to stabilize the stereo listening field. The rearward-directed energy would split in the middle and disturb the "ambience", i.e. artificially added soundstage depth by the rear wave that reflects strongly off the front wall to form a false "depth of field" effect. Needless to say, I'm not a fan of dipoles.
As far as horns are concerned, I find that larger horn mouths that can control their polars to lower frequencies actually perform better in smaller rooms than smaller-mouth horns that lose their polar control at higher frequencies. If you look at any Danley designed multiple entry horn (Unity or Synergy) you'll see large mouths as compared to traditional multiway horn-loaded loudspeakers with much smaller horn mouths.
One of the biggest advantages for small rooms is the point source effect of the multiple entry horn for use in small rooms, along with controlled polars down to much lower frequencies than typical. This is one of those application-related requirements that goes with small listening rooms that turn out to rank near the top of the list in terms of loudspeaker-room listening performance. Tom Danley has talked about the point source capability pointedly, but I see a less than spectacular response by all adherents of this type of loudspeaker design. Perhaps more discussion and demonstration of this emerging requirement for small rooms is warranted.
Bill, I believe that your comment about large horns that are too close to the listeners are those typical fully developed exponential/hyperbolic horn profiles that result in non-constant coverage vs. frequency: which is great for outdoor loudspeakers, but not useful indoors in small listening rooms like those we typically have in our living spaces (this includes big Le Cléac'h horn profiles too). Those instances look ridiculous indeed when showing the listener within 15-20 feet of the mouths of a stereo pair of very long horns that extend out of the room into the yards of their owners' properties.
Other top-ranked requirements seem to be:
- low phase growth/low group delay at bass frequencies, and
- low modulation distortion (FM and AM types) at bass frequencies
I'd like to see more discussion on requirements for the application domain. For instance, this includes those requirements that most strongly affect performance in small rooms--and the loudspeaker's design response to those requirements. This to me is preferred over the more typical case of tacit assumptions of what the requirements actually are--with no discussion at all on relative ranking of those requirements--the entire discussion exhausting salient design points only.
Chris
The Ijit responds
To reply to original question: understanding a Unity (Synergy) is conceptually simple. You squirt some highs into a conical horn, wait a fraction of a millisecond, squirt some mids (down the cone), wait a few ticks more and inject some woof. Discover original patent (Unity) expired, so make some crafty changes to ports and give them an obscure Latin term. Crossover is re-designed by wizards with EEs and computer black magic
I by no means want to diss Mr. Danley's fine invention(s). I can't afford his gear but have been quite content with what I believe is the only available non-DSL product: the Yorkville Unity U15. They are truly the "poor man's Danley", being about 1/4 the price of a pair of the coveted SH-50s. I lucked onto a trashed pair about a year ago. I soon made them active EQ and custom tuned with my limited test gear and (many here would add) even more limited common sense
I've not heard the real Synergy which is just as well because I'd be mortgaging the house to buy a pair probably.
They are the best pair of speakers I've ever owned. They look the horror but sound quite pleasant
To reply to original question: understanding a Unity (Synergy) is conceptually simple. You squirt some highs into a conical horn, wait a fraction of a millisecond, squirt some mids (down the cone), wait a few ticks more and inject some woof. Discover original patent (Unity) expired, so make some crafty changes to ports and give them an obscure Latin term. Crossover is re-designed by wizards with EEs and computer black magic
I by no means want to diss Mr. Danley's fine invention(s). I can't afford his gear but have been quite content with what I believe is the only available non-DSL product: the Yorkville Unity U15. They are truly the "poor man's Danley", being about 1/4 the price of a pair of the coveted SH-50s. I lucked onto a trashed pair about a year ago. I soon made them active EQ and custom tuned with my limited test gear and (many here would add) even more limited common sense
I've not heard the real Synergy which is just as well because I'd be mortgaging the house to buy a pair probably.
They are the best pair of speakers I've ever owned. They look the horror but sound quite pleasant
See following thread about halfway down the page for Tom Danley's reply to questions on the SPL TD-1 and TD-2 designs. According to him, the crossovers for both designs were never "completed" per his comment: SPL TD-1 and TD-2
Having a clean set of crossover points, especially the midrange-tweeter crossovers in those 3-way designs, without phase mismatch issues will be an audible improvement to passive crossovers that don't compensate for the circuit phase shifts. No magic in that realization.
The effect of those mismatches due to crossover issues will be present in FR and phase, but not the polar coverage angles (i.e., external loudspeaker lobing vs. frequency) since the single aperture horn integrates the resulting output to a much higher degree than your typical multiple-horn multiway design.
I believe that fact gets lost in translation...at least to novices on multiple entry horns. That's a really big deal in terms of the hi-fi performance in small listening rooms like we have at home--much more than it would otherwise appear to affect the overall in-room performance.
Chris
Having a clean set of crossover points, especially the midrange-tweeter crossovers in those 3-way designs, without phase mismatch issues will be an audible improvement to passive crossovers that don't compensate for the circuit phase shifts. No magic in that realization.
The effect of those mismatches due to crossover issues will be present in FR and phase, but not the polar coverage angles (i.e., external loudspeaker lobing vs. frequency) since the single aperture horn integrates the resulting output to a much higher degree than your typical multiple-horn multiway design.
I believe that fact gets lost in translation...at least to novices on multiple entry horns. That's a really big deal in terms of the hi-fi performance in small listening rooms like we have at home--much more than it would otherwise appear to affect the overall in-room performance.
Chris
Excuse for not measuring in detail
Patrick (#90) thank you for the tip. However, I have two good reasons that give me immunity to so testing. (1) My active EQ system is JRiver which seems to preclude reliable sweeps with REW; (2) I am loathe to mount my speaker twenty feet in the air outside so that I can get good IR data.
If I suddenly find myself with recourse to better test gear and an anechoic chamber, I may run some tests.
Patrick (#90) thank you for the tip. However, I have two good reasons that give me immunity to so testing. (1) My active EQ system is JRiver which seems to preclude reliable sweeps with REW; (2) I am loathe to mount my speaker twenty feet in the air outside so that I can get good IR data.
If I suddenly find myself with recourse to better test gear and an anechoic chamber, I may run some tests.
Patrick (#90) thank you for the tip. However, I have two good reasons that give me immunity to so testing. (1) My active EQ system is JRiver which seems to preclude reliable sweeps with REW; (2) I am loathe to mount my speaker twenty feet in the air outside so that I can get good IR data.
If I suddenly find myself with recourse to better test gear and an anechoic chamber, I may run some tests.
You can get clean measurements in a living room. There's a dude on this forum that builds and measures speakers in his hotel room!
The trick is to set the gate properly. For instance, 500hz is twenty seven inches long. That means that if you can eliminate all reflections within twenty seven inches of the loudspeaker, you can get a clean measurement down to 500hz.
Obviously, it would be ideal to capture MULTIPLE wavelengths, not just one wavelength. IE, if you could clear all reflections within 270", you could record TEN wavelengths at 500hz.
But there's a fix for that too! Arta can record the same thing multiple times and average them. And THAT is how to make a nice clean recording in your living room.
Ok thanks! My bottle neck is getting REW and my JRiver "crossover" to play well together. I've pretty much given up on it -- too many glitches that invalidate measurement sweeps (but amplitude seems credible.) I think I can test drivers individually, but I think that to tweak an active x-over it is kind of necessary that it and its drivers be in the circuit.
Not sure if it was asked and answered, but....
I wonder about one small thing. They put so many low frequency drivers in those horns (especially Jericho) yet there is not much surface area by which it comes out of the horn. Like, these speakers must move quite a bit, pump a lot of air, and then it comes out trough those small holes visible from the front. Am I missing something?
I wonder about one small thing. They put so many low frequency drivers in those horns (especially Jericho) yet there is not much surface area by which it comes out of the horn. Like, these speakers must move quite a bit, pump a lot of air, and then it comes out trough those small holes visible from the front. Am I missing something?
So you're apparently worried about off-axis asymmetrical loading of the midrange diaphragms.
I'd doubt that centering the midrange drivers over the midrange ports would have much effect, even at maximum SPL output (which is doubtful would be reached in a ME horn design). The mechanism of nonlinearity would likely be unequal pressure distribution under the midrange cones due to ports located far off-axis from the midrange driver centerline, leading to rocking of the driver diaphragm under heavy load conditions, and associated cone motion nonlinearities. But I'd guess that you'd need a Kippel R&D test rig to see and measure those non-axial modes, and probably wouldn't be able to hear the effects.
If you asked this question regarding the woofers, then depending on the peak percentage of x(max) that you use--based on sizing of the total woofer diaphragm area and efficiencies of the woofers on ME horn, which is a function of the woofer port sizes, ME horn dimensions, and position/flare rate along the ME horn at the woofer off-axis ports--then I'd say you probably would be better off placing the woofers closer to center over the woofer ports.
All of these discussions probably assume that you will be loading your ME horns at or near their highest SPL output, something that many/most home DIY enthusiasts may not use often, depending on listening room size and tastes in music and listening levels.
Chris
You're apparently speaking of the gradual loss of horn gain by the woofers at some beginning low frequency due to finite (small) size and depth of the ME horn. In general, full-range Synergy horns done by Danley Sound Labs will experience a loss of horn gain starting at about 100-300 Hz. Below that frequency band, they increasingly rely on directivity gain only, which is the reason why Synergies are rated lower in sensitivity than traditional horn-loaded configurations. Arrayed Synergies will tend to offset this effect due to mutual self-loading.
You can have ports that are too small, just like you can have reflex ports that are too small in a vented box, leading to port air velocities that are too high and the accompanying loss of acoustic efficiency (ref: Olson's Acoustical Engineering). But I assume that you understand some of the physics involved with compression driver-horn efficiencies, and the resulting (mid-band) ~15 dB increase in output SPL on-axis due to compression loaded driver-horn loading over the same driver used in direct radiating mode.
Chris
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.