AVS Forum banner

1 - 20 of 32 Posts

·
Registered
Joined
·
72 Posts
Discussion Starter #1 (Edited)
I am posting information about a pair of speakers that I have designed and built by myself. I am doing this because there are innovative design elements that I feel should be applied to consumer speakers. I am happy to host listening sessions for interested parties. I live in the Northern Virginia area, and hopefully I'll have 15 posts soon and be able to PM.

The basic design philosophy for this speaker was that baffle diffraction, ceiling reflections, floor reflections, and front wall reflections should be reduced drastically, while lateral reflections should be maintained. As such, I refer to this design as the "Radical 1" or "RaDCL 1" since they are "Reflection and Diffraction Controlling Loudspeakers". They are also different from any other speaker I've ever seen, thus giving the name a double meaning. They differentiate themselves even from other line arrays, since the are a 4-way design utilizing a fully passive crossover without the need for large amounts of equalization, and the combination of the AMT tweeter and close spacing between the mids prevents comb filtering common to most line array speakers.

It's difficult for our brains to differentiate the secondary sounds we hear coming from near our speakers from the original sound. These diffraction and reflection sources rob us of the detail in the recordings we play, and probably contribute to the perception that the sound is originating from the speakers, rather than coming from an instrument in a performance space.

In contrast, the lateral reflections can be differentiated from the original source, and provide us with a sense of envelopment. These are a benefit rather than a detriment, enhancing our perception of the music, and giving it a more lifelike quality.

The speakers themselves are a 4-way passive design, standing 5'4" tall, 20.5" wide, but only 4.5" deep. This form factor was chosen for on-wall placement, without creating interference between the direct sound and reflection off the wall, because the speakers are so wide and shallow. The internal volume is limited, which in turn limits the bass response. However, it's become clear that in-room bass is better addressed by using multiple subwoofers, with their locations chosen to mitigate standing waves (not necessarily placed where the left and right speaker are). I supplement these speakers with 2 separate 12" high output subwoofers, which cover frequencies below 100 Hz.

The speakers are quite efficient (about 94.5 dB @2.83V) and offer substantial power handling, resulting in clean, dynamic sound even when listening at high levels. At 100 hz, they should reach 105 dB before the woofers move past their linear range. However, floor and ceiling reflections contribute substantially to the loudness we perceive from most speakers, so these must be played a little louder (I'd estimate 4 to 5 dB) to reach the same perceived loudness. Still, they have no problem playing way louder than I should ever actually listen to music, and hearing damage could easilly be achieved if one were not careful, since there is little evidence of how loud they are playing. Earl Geddes has shown that diffraction is a major contributor to what people perceive as distortion related to loudness.
 

Attachments

·
Registered
Joined
·
72 Posts
Discussion Starter #2 (Edited)
These speakers present a significant challenge to measure. A substantial distance is required for the drivers to fully integrate, with appropriate relative levels. As such, I decided NOT to try to eliminate reflections in the measurements as most others do. Instead, I measured them where I use them: up against the wall. My speaker measurements all include the front wall reflection, and the floor and ceiling reflections. Most speakers would suffer pretty substantial interference problems if their measurements included reflections off the front wall, floor, and ceiling. But since these were designed to reduce the impact of those reflections, a nice result is possible anyway. Still, the measurements must be considered in context (which is always the case).

All these measurements were taken at a distance of about 6 feet from the LEFT speaker, at tweeter height. For the first picture, the mic was placed directly between the speaker and my listening position. The response is gated, with no other smoothing. The impulse response shows the floor reflection at about 2.5-3 ms delay, and the ceiling reflection at a 4 ms delay. A lateral reflection is visible at 7.5 ms, but is excluded from the gate.

The second picture supplements the measurement described above with 2 more measurements. These measurements are pointed where someone would be sitting if they were 3 seats away from center to the left (blue) and right (green). For reference, my seating is about 11 feet back from the speakers, and these seats are about 11 feet apart. If my speakers pointed straight out from the wall (which they don't) these would be at about 0 (blue) and 45 degrees (green), with the main listening position near 22.5 degrees (red).

The 3rd picture shows the impact of changing the gating of the measurement pointed to the main listening position to exclude the direct sound, but include both the floor and ceiling reflections. Similarly, the 4th picture excludes the direct sound, but captures a single lateral reflection off the left wall.

The final picture utilizes the frequency-dependent gating feature in Holm impulse that aims to approximate our perception of the sound. Since lower frequencies use a longer gating, they capture more reflections, which contribute to increased variance, with higher peaks. The slanted frequency response is natural, and speakers that measure flat in an anechoic chamber typically show such a trend when measured in-room.
 

Attachments

·
Registered
Joined
·
972 Posts
Return of the wide baffle--interesting design. Nice build. Just wiring up the drivers alone would be very time consuming.

The wider baffle would lower the baffle step. Did you have any issues with the compensation? Also, wouldn't you want larger mids and woofers to better control directivity?

Most importantly, how do they sound?
 

·
Registered
Joined
·
72 Posts
Discussion Starter #4
Return of the wide baffle--interesting design. Nice build. Just wiring up the drivers alone would be very time consuming.

The wider baffle would lower the baffle step. Did you have any issues with the compensation? Also, wouldn't you want larger mids and woofers to better control directivity?

Most importantly, how do they sound?
Wiring up the drivers was fairly time consuming. By the time I was done I felt like I should have had a robot or something instead of a speaker.

I didn't have any problems with baffle step. The speaker always radiates into pi-space. Sometimes I think of it as a 180 degree horn. In order to achieve that, the depth has to be a tiny fraction of the width. I didn't want them to be ridiculously wide, so they had to be very shallow. The tiny depth is part of what led to the tiny woofers, because I needed something with very little Vas. Also, I wanted wide horizontal dispersion (besides the 180 degree horn aspect of integrating it into the wall), which requires smaller (or at least narrower) drivers. In order to create vertical directivity I used multiple small drivers, except for the tweeter. The tweeter is a single AMT with appropriate proportions, which doesn't suffer from the comb filtering so common among line arrays.

They sound great. I have had several musicians (who were also audio enthusiasts to varying degrees) tell me they have never heard a speaker system so accurately re-create the sound of instruments they play. The speakers are incredibly revealing of the source material, but in a completely non-fatiguing way. Also, they do a great job of re-creating the spatial cues on the recording. I don't intend to wax poetic about the sound. I'd rather have some others offer their impressions, and see if this approach is something that commercial speaker manufacturers ought to consider.
 

·
Registered
Joined
·
603 Posts
This thread deserves a lot more activity than it's received. I've heard the Radical on several occasions and it delivers on its design objectives. It has confirmed my belief, based on years of speaker design, that broad lateral dispersion is extremely important to realistic sound reproduction, and broad vertical dispersion is not necessary or even desirable, assuming it's not so narrow as to produce a "head in a vice" listening position. When I first saw the Radical, I was very skeptical that so many drivers could work together to produce a coherent, neutral sound. After listening to my own Philharmonic Audio test CD, I was convinced that the basic sound signature of the Radical was as neutral as any I had heard, and I also sensed a more immediate presentation of the recording venue, be it a studio or concert hall. The sound was simply cleaner, with superior differentiation of individual instruments and voices. To confirm that the Radical's response really was significantly truncated in the vertical plane, I knelt down to floor level and experienced a dramatic reduction in sound level. It was a little spooky. I was so fascinated with the Radical that I came back for a second listen, this time with a pair of my own Affordable Accuracy monitors. The AA's are perfectly conventional 2-way monitors, albeit with a complex crossover. They are as neutral as I could make them, and tonally there wasn't that much difference between them and the Radicals. But the AA's were simply "noisier" than the Radical. I had a much greater sensation that I was listening through the Radicals into the recording venue. Unlike my experience with so many controlled directivity speakers with wave guides or horns, I never sensed a constriction of the sound stage or any coloration. The Radical is obviously complex and wouldn't be practical in all listening environments, but in my opinion it would make a killer home theater speaker as well as a truly superior speaker for dedicated music listening. I was so enthusiastic about the Radical that I convinced a certain speaker manufacturer to fly in for a listen. He was as impressed as I was, and simply doesn't have the time or resources to make the Radical part of his lineup. I'm hoping that will change in the coming year. The Radical is an important step forward in speaker design and deserves wide exposure. Dennis Murphy
 

·
Registered
Joined
·
207 Posts
I've already had my brain broke once today. This is cool, and I wish I understood it better! Thank you Dennis for directing our attention this way!!! @spkr_diy : Thank you for sharing, and not being afraid to innovate!!!
 

·
Registered
Joined
·
1,864 Posts
Sometimes I think of it as a 180 degree horn. In order to achieve that, the depth has to be a tiny fraction of the width...
I'm not sure I follow-can you expand on that?
And I'd observe that in some way these are like an in-wall, getting rid of that first strong reflection. I had one house with in-walls and there was definitely a nice quality which I attributed to that. Good work!
 

·
Registered
Joined
·
72 Posts
Discussion Starter #8
I'm not sure I follow-can you expand on that?
And I'd observe that in some way these are like an in-wall, getting rid of that first strong reflection. I had one house with in-walls and there was definitely a nice quality which I attributed to that. Good work!
At high frequencies, the wall behind the speaker (I'm talking specifically about the thick side of the speaker here) is in an acoustic shadow. At lower frequencies, the sound pressure can bend around the baffle and head toward the front wall. However, since the wavelengths are long, and the speaker is shallow, the reflection is still in phase with the direct sound, and serves to reinforce that sound. That's how it effectively always radiates into pi space, and works like a 180 degree horn. If I pull the speaker even a few inches away from the wall, the path from the baffle edge to the front wall and back will no longer be in phase with the highest frequencies that bend around the baffle, and cancellation will occur. The thickest part of the speaker is only 4.5" deep (outside dimensions). It was a difficult criteria to accommodate, but a necessary one.
 

·
Registered
Joined
·
3,475 Posts
With the 20.5 inch wide baffle, the baffle essentially behaves like a 180 degree horn down to 228 hz, into the schroeder region of most rooms. Placed near the wall like shown is a good design choice.



228 hz has a wavelength of 59.47 inches, sound wrapping the baffle edges and reflecting off the wall will be in phase. (
 

·
Registered
Joined
·
173 Posts
At high frequencies, the wall behind the speaker (I'm talking specifically about the thick side of the speaker here) is in an acoustic shadow. At lower frequencies, the sound pressure can bend around the baffle and head toward the front wall. However, since the wavelengths are long, and the speaker is shallow, the reflection is still in phase with the direct sound, and serves to reinforce that sound. That's how it effectively always radiates into pi space, and works like a 180 degree horn. If I pull the speaker even a few inches away from the wall, the path from the baffle edge to the front wall and back will no longer be in phase with the highest frequencies that bend around the baffle, and cancellation will occur. The thickest part of the speaker is only 4.5" deep (outside dimensions). It was a difficult criteria to accommodate, but a necessary one.
Do you have a plan for these speakers going forward? It sounds like maybe you are trying to license them to a speaker manufacturer but I for one would be very interested if a DIY design were released as I imagine these would carry a hefty price tag from a commercial source.
 

·
Registered
Joined
·
72 Posts
Discussion Starter #11
Do you have a plan for these speakers going forward? It sounds like maybe you are trying to license them to a speaker manufacturer but I for one would be very interested if a DIY design were released as I imagine these would carry a hefty price tag from a commercial source.
I don't have a specific plan for these speakers at the moment. I am not posting here to find someone to license the design to, though I do want to keep that option open. I am trying to influence the consumer speaker industry a little. A couple of the effective technologies in this speaker are rare or non-existent in consumer speakers.

1) If speakers are wide enough and shallow enough, on-wall placement can be a benefit rather than a detriment.

2) Vertical dispersion can be limited with a line array design variation without the usual problems of comb filtering, nor the need for EQ or numerous amplifier channels.

3) Large round-overs are effective at reducing diffraction.

I am currently working on another project to drive home that last point. I will be modifying a commercial speaker design. I hope to post about that in the next several weeks.
 

·
Registered
Joined
·
54 Posts
The radicals look fascinating to me. I like that they are efficient, and though large, the slender profile compensates for that.

I would also be interested in a DIY option.

When you post about your current project maybe link to it on this thread so I will get an alert☺

Thanks for sharing.
 

·
Registered
Joined
·
2,864 Posts
Just a point for aesthetics - perhaps use black screws for the drivers rather than stainless steel ones.
 

·
Registered
Joined
·
72 Posts
Discussion Starter #14
Just a point for aesthetics - perhaps use black screws for the drivers rather than stainless steel ones.
I agree in theory. However, this project suffered a lot of delays due to other things having priority (job, wife, kids, house) and I refused to add to the delay to source the right color screws. Have you ever heard that "done is better than perfect"? It would be interesting to see how nice they could look. Mine look very industrial... but I'm perfectly happy with that.
 

·
Registered
Joined
·
2,430 Posts
@spkr_diy

Your DIY speakers look amazing! What a labor of love.

Don't mean to go OT, but what is the center speaker in your 1st pic?
 

·
Registered
Joined
·
72 Posts
Discussion Starter #17
Any thoughts on what you'd change if free of the depth constraint and mounting them flush in a baffle wall ?

Cheers,
If not for the depth constraint, I might use a different tweeter, though the AST2560 has proven to be amazing. Most ribbons have a large enough transformer that they would be too deep for my cabinet. I originally had a different tweeter in mind (I think it was a Fountek) but eventually figured out that it wouldn't fit.

Unless there was a volume constraint (requiring low VAS woofers), I'd probably build it with a bigger woofer, too. I can't really justify it, since these go deep enough to cross to my subs, and they play plenty loud and clean. Sometimes we go overboard for our passions.
 

·
Registered
Joined
·
72 Posts
Discussion Starter #19
@spkr_diy

Your DIY speakers look amazing! What a labor of love.

Don't mean to go OT, but what is the center speaker in your 1st pic?
Thanks for the kind words.

The center speaker is a very old DIY speaker I made specifically to fit on the top of a CRT TV that had a very shallow top (about 4 inches). The crossover is pretty low, so the dispersion isn't as troubling as it would appear. There's a port on the left side where there would normally be a second woofer in a center channel. I have plans and parts to build a truncated version of the Radical speakers that will fit under the screen, but that project never makes it to the top of my priorities. I do all my critical listening in stereo, and will continue to do so if/when I finish the matching center channel. The center channel only gets used for movies, and what I've found is that you don't need perfection for movies, which are mostly about dialogue and effects.
 

·
Registered
Joined
·
511 Posts
Do you have a plan for these speakers going forward? It sounds like maybe you are trying to license them to a speaker manufacturer but I for one would be very interested if a DIY design were released as I imagine these would carry a hefty price tag from a commercial source.
+1

Would love to attempt a DIY project with this design for stereo listening.
 
1 - 20 of 32 Posts
Top