I rented a pair of SH50s recently, and they image really well. Intrigued by how they work, I decided to mothball my 'Monster Massive*' project and start a new one I'm calling 'Black Mirror.'
In order to understand why the SH50s image well, we need to understand how our brain creates a soundstage. Above 1khz our perception of location is mostly dictated by intensity. Below 1khz our perception of location is mostly dictated by interaural time differences. Around 1khz it's both.
IMHO, it's fairly simple to get a good image above 1khz. All you need to do is get a couple of well matched tweeters and sit equidistant between them. Waveguides can help here, because they control early reflections.
Below 1khz, things get trickier.
I've generally found that dipoles image better than monopoles. And that's one of the reasons that Monster Massive was dipole. But I found that the SH50s imaged as well as MM, and in many respects, imaged better. So, again, that got me thinking about what generates a good soundstage.
I think that one of the reasons for the 'focus' of the SH50s is because reflections in the octave between 500hz and 1khz are minimized. (Because of the waveguide.) The same holds true of dipoles, but if you review my measurements from the end of the MM thread, you'll see that there's no doubt that waveguides control directivity better than dipoles. IE, they both get the job done, but one works better than the other.
I think that reflections, or lack thereof, are a big part of that pinpoint imaging with the SH50s.
Here's some illustrations to show what I mean.
With the big SH50s, we get some reflections off of the floor, the ceiling, and the wall behind the listener. From 500hz to 20khz there's virtually no energy radiated towards the wall behind the loudspeakers, or the side walls. And I think that's a big part of the secret sauce. I've heard speakers that imaged like the SH50s, but those speakers were in very large rooms, where the nearest wall was as much as 2-4 meters away. I can't do that in my room; I live in San Diego and my listening room is under 150'.
In the second illustration, we see what happens with something like a Lambda Unity. The tweeter sits about 3' off the floor. Because the waveguide doesn't control directivity below 1khz, we're going to get reflections off of the floor, the wall behind the loudspeaker, and the ceiling. The worst case scenario with something like the Lambda would be if you pulled it 4' away from the back wall, and then raised it up a foot. If you did that, you'd get simulataneous reflections off of the floor, ceiling and back wall, all at the same time.
Time plays a big factor here. Ideally, we'd eliminate all reflections. But since we can't do that, we'd like to minimize them. If we can't do that, we'd like to spread them out in time. That's one of the things that's a bit nefarious about having a tweeter 4' off of the floor; it will tend to make the reflections happen simultaneously, and that's a recipe for crappy imaging.
The bottom illlustration is my 'Black Mirror' project. Basically move the waveguide down low, so that the floor and the side walls extend it. In the illustration, you can see that pushing the loudspeaker all the way into the corner basically eliminates the early reflections in the lower midrange. Basically the wall and the waveguide function as one.
Here's some pics to illustrate the size of the cabinet, how it fits in the corner, and the size when compared to my Summas.
In order to understand why the SH50s image well, we need to understand how our brain creates a soundstage. Above 1khz our perception of location is mostly dictated by intensity. Below 1khz our perception of location is mostly dictated by interaural time differences. Around 1khz it's both.
IMHO, it's fairly simple to get a good image above 1khz. All you need to do is get a couple of well matched tweeters and sit equidistant between them. Waveguides can help here, because they control early reflections.
Below 1khz, things get trickier.
I've generally found that dipoles image better than monopoles. And that's one of the reasons that Monster Massive was dipole. But I found that the SH50s imaged as well as MM, and in many respects, imaged better. So, again, that got me thinking about what generates a good soundstage.
I think that one of the reasons for the 'focus' of the SH50s is because reflections in the octave between 500hz and 1khz are minimized. (Because of the waveguide.) The same holds true of dipoles, but if you review my measurements from the end of the MM thread, you'll see that there's no doubt that waveguides control directivity better than dipoles. IE, they both get the job done, but one works better than the other.
I think that reflections, or lack thereof, are a big part of that pinpoint imaging with the SH50s.
Here's some illustrations to show what I mean.
With the big SH50s, we get some reflections off of the floor, the ceiling, and the wall behind the listener. From 500hz to 20khz there's virtually no energy radiated towards the wall behind the loudspeakers, or the side walls. And I think that's a big part of the secret sauce. I've heard speakers that imaged like the SH50s, but those speakers were in very large rooms, where the nearest wall was as much as 2-4 meters away. I can't do that in my room; I live in San Diego and my listening room is under 150'.
In the second illustration, we see what happens with something like a Lambda Unity. The tweeter sits about 3' off the floor. Because the waveguide doesn't control directivity below 1khz, we're going to get reflections off of the floor, the wall behind the loudspeaker, and the ceiling. The worst case scenario with something like the Lambda would be if you pulled it 4' away from the back wall, and then raised it up a foot. If you did that, you'd get simulataneous reflections off of the floor, ceiling and back wall, all at the same time.
Time plays a big factor here. Ideally, we'd eliminate all reflections. But since we can't do that, we'd like to minimize them. If we can't do that, we'd like to spread them out in time. That's one of the things that's a bit nefarious about having a tweeter 4' off of the floor; it will tend to make the reflections happen simultaneously, and that's a recipe for crappy imaging.
The bottom illlustration is my 'Black Mirror' project. Basically move the waveguide down low, so that the floor and the side walls extend it. In the illustration, you can see that pushing the loudspeaker all the way into the corner basically eliminates the early reflections in the lower midrange. Basically the wall and the waveguide function as one.
Here's some pics to illustrate the size of the cabinet, how it fits in the corner, and the size when compared to my Summas.
I noticed a similar thing when listening to my Trynergy speakers - the narrower directivity reduced the sound splashing off walls and floor and ceiling to give a very clean sharp soundstage. Very neat effect that I never had with another speaker. It may be true for horn mid/hi speakers in general.
I've posted this facetiously elsewhere, but I really think the logical end point of your current path involves tilting the listening room backward 45 degrees and building a 3-sided synergy horn at each apex of front wall/side wall/floor (drivers inside the walls). Hang a hammock at the MLP, add optional rope ladder and some sort of pulley system for beverage delivery, done.
More seriously, has anyone ever tried or thought about a 3 sided (tetrahedral) horn? Seems like it might be helpful here.
More seriously, has anyone ever tried or thought about a 3 sided (tetrahedral) horn? Seems like it might be helpful here.
One thing that I would worry with the tweeters in the bottom corners is the image would be shifted down, cymbals on the floor boards and such. Maybe just a midrange horn there in the corner and two smaller horns in traditional placement handling the most directional sounds.
Reminds me of the big diy JBL setup I heard a while back, I wrote a little about it here:
No Audiophile Field Trip #1
It was still too much like really big headphones with the imaging, I missed the reflections.
I've spent a few years chasing the same sort of things that you have, and tried a number of different configurations. One thing that I have not tried yet is the to maximize cancellation between the left and right speaker at the listening position. This is something that Legacy talks about in the marketing materials for their Whisper speakers.
The general idea is to take measurements with a mono signal at the listening position from both speakers at the same time trying to create a uniform null in the response over as much of the frequency domain as possible. This may be why symmetry in placement and room layout makes such a large difference in imaging performance. The good phase performance and uniform off axis frequency response of the SH-50 would probably do very well in this test from just about any position. Of course it's very possible that it's just marketing mumbo jumbo.
Going back to one of your other projects, the LX mini make over, the baffle width was a big deal. Maybe the CBT array is a something to consider: narrow baffle with reduced/controlled reflections and constant directivity. On paper it seems to check all of the boxes, but I've yet to hear one in person.
Reminds me of the big diy JBL setup I heard a while back, I wrote a little about it here:
No Audiophile Field Trip #1
It was still too much like really big headphones with the imaging, I missed the reflections.
I've spent a few years chasing the same sort of things that you have, and tried a number of different configurations. One thing that I have not tried yet is the to maximize cancellation between the left and right speaker at the listening position. This is something that Legacy talks about in the marketing materials for their Whisper speakers.
The general idea is to take measurements with a mono signal at the listening position from both speakers at the same time trying to create a uniform null in the response over as much of the frequency domain as possible. This may be why symmetry in placement and room layout makes such a large difference in imaging performance. The good phase performance and uniform off axis frequency response of the SH-50 would probably do very well in this test from just about any position. Of course it's very possible that it's just marketing mumbo jumbo.
Going back to one of your other projects, the LX mini make over, the baffle width was a big deal. Maybe the CBT array is a something to consider: narrow baffle with reduced/controlled reflections and constant directivity. On paper it seems to check all of the boxes, but I've yet to hear one in person.
Not necessarily, take Snell Type One as an example. Those things were thoroughly discussed in the FCUFS thread.
@The Black Mirror corner speaker - Gamper Martino corner speaker comes to my mind - all this has been discussed in the FCUFS thread
graaf,
I'm afraid not many know (or care about) this. And "FCUFS" might be confused as some kind of UFO, haha....
Here you are: http://www.diyaudio.com/forums/multi-way/121385-advantages-floor-coupled-up-firing-speakers-81.html
🙂
I'm afraid not many know (or care about) this. And "FCUFS" might be confused as some kind of UFO, haha....
Here you are: http://www.diyaudio.com/forums/multi-way/121385-advantages-floor-coupled-up-firing-speakers-81.html
🙂
It looks like I have some reading to do; I'm astounded that I missed such a long thread on this forum.
I feel confident that this project will work, so this time around, I'm building both speakers simultaneously. (In 80% of my projects I'm not so confident, and I generally only build a single speaker, then build a second one if the first one works as planned.)
Due to the fact I'm building both at the same time, things aren't moving as rapidly as they normally do. Right now I have the waveguides mounted on a baffle for both speakers, and I have the mounting plates attached for the midranges. The compression driver won't require a mounting plate, because this waveguide was designed for the compression driver I am using. (I'm using the big QSC waveguide, and the waveguide was designed for the Celestion CDX1-1425.)
Hahaha, Oh man,
I Really , really like your percistence!! Go John!
(And I like the fact that you followed my advise and checked out the sh50's in real-life)
I Really , really like your percistence!! Go John!
(And I like the fact that you followed my advise and checked out the sh50's in real-life)
Melo Theory is building a 3-sided horn...
http://www.diyaudio.com/forums/multi-way/263353-portal-controlled-dispersion-tripole.html
I'm not sure it's worth reading through but You can find a lot of information there using the search function.
Saw a three sided corner cube horn over here, looks like it's dying to have a compression driver:
http://www.diyaudio.com/forums/subwoofers/258706-study-dipole-cardioid-bass-horn-4.html#post3983145
http://www.diyaudio.com/forums/subwoofers/258706-study-dipole-cardioid-bass-horn-4.html#post3983145
Synergy corner horns make a lot of sense! Walls acting as secondary flare.
I have a dream of having such in my HT cave as mains. I am pretty sure that a center speaker is not needed. Actually I made a quickie proto last summer. But this setup would mean pulling down 4 m2 of book/record shelf. And finding a new place for it. I don't have that much energy now.
I have a dream of having such in my HT cave as mains. I am pretty sure that a center speaker is not needed. Actually I made a quickie proto last summer. But this setup would mean pulling down 4 m2 of book/record shelf. And finding a new place for it. I don't have that much energy now.
Last edited:
Hmmm... In my understanding, the center channel is mainly for correct pinna cues in the head-related transfer function.
Well, I'm not pretending I know those well, but I'm hooked after I tried a central speaker (with linear matrix 3-ch arrangement). It plays the major portion of the music at the correct location most of the time. It makes perfect sense.
So I really don't understand why 2-ch stereo is the 'standard' for all these years.
Well, I'm not pretending I know those well, but I'm hooked after I tried a central speaker (with linear matrix 3-ch arrangement). It plays the major portion of the music at the correct location most of the time. It makes perfect sense.
So I really don't understand why 2-ch stereo is the 'standard' for all these years.
If the soundstage is shifted too low from placement you could get really silly and add a pair on the ceiling corners. This might provide a phantom vertical center that you could tweak like a balance knob to raise or lower the image.
CLS, I agree that with well done multichannel recordings, the center channel is for good. Also for movies.
The physical realities make it impossible for me to install 3 cormer-loaded horns as L/C/R 😉 The C is a compromise and because my room is only 3m wide it is not that necessary. Now I have identical coaxial L/C/R speakers, but I listen to 2ch music with a pair (plus sub s 2.2), movies/tv with 5.2
The physical realities make it impossible for me to install 3 cormer-loaded horns as L/C/R 😉 The C is a compromise and because my room is only 3m wide it is not that necessary. Now I have identical coaxial L/C/R speakers, but I listen to 2ch music with a pair (plus sub s 2.2), movies/tv with 5.2
- Status
- Not open for further replies.