Greetings from Poland!
I’m excited to share my rendition @perrymarshall's Live Edge Dipoles (minus the live edge). I am a 20-year-old industrial design student from Poland, currently studying in the Netherlands, and this was my journey with the build 🙂
The audiophile bug bit me thanks to my dad, who has always had a passion for HI-FI audio. Until now, I’d only ever experienced box speakers. About three years ago, the idea of building my own speakers began to take shape. For my 17th birthday, I received a beautiful piece of Sapele wood (a type of exotic African mahogany from the Meliaceae family) and a handful of tools. Naturally, my first naive thought was, "I'm going to build the ultimate best-sounding speaker from what i got." However, I quickly realized it wasn't going to be that simple 😀 For almost two years, that wood 'dried' in my room (which is a politically correct way of saying it blocked the entrance) while I dove into learning about speaker design, some basic acoustics, and physics.
Fast forward to about a year ago, just as I was about to cut the wood for a 100-liter box (the only thing I knew), I discovered some pretty amazing open baffle designs, including Perry Marshall’s Bitches Brews and Live Edge Dipoles. Perry turns out to be not only a terrific speaker designer but also an incredible salesman. So naturally, my design took a 180-degree turn to an open baffle. I decided to trust that it would be something different from what my dad and I had ever heard before. Plus, Perry's design was stunning in its elegance and simplicity and had a convincing track record—likely much better than any freestyle attempt from my inexperience.
Oh boy, am I grateful that I took that route. The speaker turned out to be everything I had hoped for. My main concern was the bass output since I'd never heard open baffle bass before. The fact that 99% of speakers I'd seen were box speakers made me sceptical. My dad loves deep bass, and his previous custom-built speakers delivered impressive low-end performance. But the prototype of my open baffle design alleviated all my worries with a single drum hit @90db.
To challenge myself, I redesigned the speaker in CAD and spent half a year building it in the university workshop despite a tight schedule. I had to excuse the big pile of wood as part of my “final bachelor project” exploration, and thankfully, I wasn’t kicked out. Near the finish line, though, I had to find an alternative location to cut out the big 18-inch holes and do the final long hours of sanding.
I have used the aluminium diagram version of the Radian 8'' Coaxials because of well... I'm working with a student budget. I have made a slight modification where the top part of the baffle is bent backwards about 4 degrees + the 5 degrees of the original baffle tilt. But this still resulted in a little too much tilt for the way the room in Poland is set up. So, I have corrected this by adding spikes to the speaker base. Additionally, I have modelled the crossover according to Perry’s design and experimented with my own tweaks to understand how each component influenced the build. However, I stayed true to the original design for the most part.
I brought the bad boys to Poland this summer and I’m fine-tuning the design with IIR and FIR filters to perfectly suit my dad’s man cave, however, the measurements are looking great already, and the wife appreciation factor remained high, even after ditching the live edge. They disappear when moved a bit further into the room than in the picture below, with the wide baffle facing slightly away from the listening position. I knew the sacrifice of having such a 'wall of a baffle' around the coaxial but that was mostly driven by aesthetics.
This project has been a deeply fulfilling endeavour, combining my love for design, craftsmanship, and high-fidelity audio. I’m proud of the final product and excited to continue refining it. My dad has no complaints, and I am inspired to build some more speakers. A huge thanks to @perrymarshall for his generous and professional outline of this open baffle design! Thank you for sharing the knowledge in such an open fashion. All the Best! Jakub
I’m excited to share my rendition @perrymarshall's Live Edge Dipoles (minus the live edge). I am a 20-year-old industrial design student from Poland, currently studying in the Netherlands, and this was my journey with the build 🙂
The audiophile bug bit me thanks to my dad, who has always had a passion for HI-FI audio. Until now, I’d only ever experienced box speakers. About three years ago, the idea of building my own speakers began to take shape. For my 17th birthday, I received a beautiful piece of Sapele wood (a type of exotic African mahogany from the Meliaceae family) and a handful of tools. Naturally, my first naive thought was, "I'm going to build the ultimate best-sounding speaker from what i got." However, I quickly realized it wasn't going to be that simple 😀 For almost two years, that wood 'dried' in my room (which is a politically correct way of saying it blocked the entrance) while I dove into learning about speaker design, some basic acoustics, and physics.
Fast forward to about a year ago, just as I was about to cut the wood for a 100-liter box (the only thing I knew), I discovered some pretty amazing open baffle designs, including Perry Marshall’s Bitches Brews and Live Edge Dipoles. Perry turns out to be not only a terrific speaker designer but also an incredible salesman. So naturally, my design took a 180-degree turn to an open baffle. I decided to trust that it would be something different from what my dad and I had ever heard before. Plus, Perry's design was stunning in its elegance and simplicity and had a convincing track record—likely much better than any freestyle attempt from my inexperience.
Oh boy, am I grateful that I took that route. The speaker turned out to be everything I had hoped for. My main concern was the bass output since I'd never heard open baffle bass before. The fact that 99% of speakers I'd seen were box speakers made me sceptical. My dad loves deep bass, and his previous custom-built speakers delivered impressive low-end performance. But the prototype of my open baffle design alleviated all my worries with a single drum hit @90db.
To challenge myself, I redesigned the speaker in CAD and spent half a year building it in the university workshop despite a tight schedule. I had to excuse the big pile of wood as part of my “final bachelor project” exploration, and thankfully, I wasn’t kicked out. Near the finish line, though, I had to find an alternative location to cut out the big 18-inch holes and do the final long hours of sanding.
I have used the aluminium diagram version of the Radian 8'' Coaxials because of well... I'm working with a student budget. I have made a slight modification where the top part of the baffle is bent backwards about 4 degrees + the 5 degrees of the original baffle tilt. But this still resulted in a little too much tilt for the way the room in Poland is set up. So, I have corrected this by adding spikes to the speaker base. Additionally, I have modelled the crossover according to Perry’s design and experimented with my own tweaks to understand how each component influenced the build. However, I stayed true to the original design for the most part.
I brought the bad boys to Poland this summer and I’m fine-tuning the design with IIR and FIR filters to perfectly suit my dad’s man cave, however, the measurements are looking great already, and the wife appreciation factor remained high, even after ditching the live edge. They disappear when moved a bit further into the room than in the picture below, with the wide baffle facing slightly away from the listening position. I knew the sacrifice of having such a 'wall of a baffle' around the coaxial but that was mostly driven by aesthetics.
This project has been a deeply fulfilling endeavour, combining my love for design, craftsmanship, and high-fidelity audio. I’m proud of the final product and excited to continue refining it. My dad has no complaints, and I am inspired to build some more speakers. A huge thanks to @perrymarshall for his generous and professional outline of this open baffle design! Thank you for sharing the knowledge in such an open fashion. All the Best! Jakub
Attachments
Last edited:
Very nice build! They look great. Have you considered putting some room treatments in that room? It looks like it could benefit tremendously. I'm guessing the sound you're getting isn't really what you were hoping for, and the room is the reason.
Thanks for the kind words! Yes, room treatment is next on my list. I’m aware of how much a well-treated room can impact the sound, and I’ve heard that open baffles can benefit from a bit of a reflective environment so i don't want to go overboard (though I'm not sure how true that statement is).
Currently, the room does have a variety of bookshelves and furniture on the sides and at the back, but I’m thinking about adding a thick curtain to the two-door openings and a rug right in front of the speakers. From then I would probably experiment with a couple of acoustic panels at the first points of reflection. If you have any comments about this setup or specific recommendations for panel placement or types of treatments that worked well for you, I’d love to hear them!
Currently, the room does have a variety of bookshelves and furniture on the sides and at the back, but I’m thinking about adding a thick curtain to the two-door openings and a rug right in front of the speakers. From then I would probably experiment with a couple of acoustic panels at the first points of reflection. If you have any comments about this setup or specific recommendations for panel placement or types of treatments that worked well for you, I’d love to hear them!
I think the room is arguably more important than the speakers themselves. I just moved from a few states over, just got my listening room set up in an empty room that is actually larger than my previous listening room - but I have no room treatments. It sounds like garbage. I can't even listen to it right now because I know what it should sound like. With all those hard surfaces in that room - it's going to sound like an entirely different setup once you get it treated. It's true that open baffles benefit from some reflection, but you have quite a lot of treating to do before you'd be overdoing it. I put absorptive panels on the first reflection points on the walls and the ceiling. You're also going to need a lot of bass trapping in that room - mostly in the corners. Then you should look at diffusion - some diffusion panels behind the speakers and a few on the walls will make a big difference.
Do some baseline testing beforehand - do some clap testing for echo/decay. Listen to some music and see how good the imaging is - can you pinpoint where instruments are? Do vocals sound centered and real? After you do some testing, try again. You'll be amazed.
Do some baseline testing beforehand - do some clap testing for echo/decay. Listen to some music and see how good the imaging is - can you pinpoint where instruments are? Do vocals sound centered and real? After you do some testing, try again. You'll be amazed.
Compared to a box speaker which is monopolar at low frequencies, a dipole radiator will output 4.8dB less acoustic power into the room, for the same on-axis spl. Lots of info here: https://www.linkwitzlab.com/AES'98/aes-98.htm
I think it’s the dramatic lighting accentuating the form of the drivers making them look bigger than they really are. However, the woofers are indeed the Eminence Kappa Pro-18LF-8, and they measure according to spec. I managed to grab them at a pretty good price compared to others on my Excel sheet (Precision Devices PD.1850 and B&C 18TBW100 were some alternatives I was considering).
As for the aesthetics:
The front baffle is a 4cm thick piece of sapele hardwood coated with a transparent water-based sealer to preserve the natural beauty of wood. It has a 4mm 45-degree chamfer to help mitigate some of the diffraction. The panels are arranged horizontally and joined in a way that allows some harmonious expansion and contraction, with the grain flowing seamlessly from one speaker to the other just because I could. The wings and base are cut from MDF and coated in black textured Duratex, which thins them out quite a bit in a darker room and provides a durable protective seal. I love how the transition turned out between the natural wood and the contrasting black surfaces.
I opted for adding a base, as it serves the dual purpose of housing an external crossover and solidly bracing the U-shape structure. The spikes help to virtually lighten up the speakers by lifting them off the ground, as well as preventing excessive vibrations from going into the floor and back to the baffle.
I tried to be resourceful with the wood and saved up enough for another pair of speakers (no idea what I’m going to build next, but I’m already rubbing my hands together 😛 Additionally, from the scraps, I have assembled a soon-to-be-fancy external crossover. It just needs a makeover and some plexiglass.
But I still need to decide on the final configuration. I’m still not sure whether noise will be a problem, as you pointed out in your write-up. Back in my student house, I had a more affordable multichannel NAD power amp, so I opted for padding the horn down with 10-ohm resistors and pretty much stuck to the old crossover before the upgrade. Now would be the time to determine whether I can ditch some of those resistors and redo the final crossover network accordingly. Sadly, I'm not in Poland for the time being, so I can only test when I'm back. (I used 100+100uf caps instead of 300uf like in the picture).
Now, the setup is a bit more comprehensive. I have a horizontal bi-amp with two Cyrus Power X amps. A MiniDSP 2x4 HD is sitting between them and the Cyrus preamp. It's probably a pseudoscientific and irrational observation, but the MiniDSP looks awfully small and wonky compared to everything else around it—which is probably not a bad thing, but at the same time, it makes me raise an eyebrow.😀 I'm sure DSP has come a long way, but truly, I'm not sure how this little magic box can do so much with the signal without any audible artefacts. Well, the resolution is there, so I'm not complaining, but maybe someone can explain it to me so I can sleep a little better at night?🙂
In the long run, is it better to upgrade to a better DSP unit? Could I notice a difference or go fully active? And @perrymarshall any advice on how I should experiment with the crossover in the meantime?
As for the aesthetics:
The front baffle is a 4cm thick piece of sapele hardwood coated with a transparent water-based sealer to preserve the natural beauty of wood. It has a 4mm 45-degree chamfer to help mitigate some of the diffraction. The panels are arranged horizontally and joined in a way that allows some harmonious expansion and contraction, with the grain flowing seamlessly from one speaker to the other just because I could. The wings and base are cut from MDF and coated in black textured Duratex, which thins them out quite a bit in a darker room and provides a durable protective seal. I love how the transition turned out between the natural wood and the contrasting black surfaces.
I opted for adding a base, as it serves the dual purpose of housing an external crossover and solidly bracing the U-shape structure. The spikes help to virtually lighten up the speakers by lifting them off the ground, as well as preventing excessive vibrations from going into the floor and back to the baffle.
I tried to be resourceful with the wood and saved up enough for another pair of speakers (no idea what I’m going to build next, but I’m already rubbing my hands together 😛 Additionally, from the scraps, I have assembled a soon-to-be-fancy external crossover. It just needs a makeover and some plexiglass.
But I still need to decide on the final configuration. I’m still not sure whether noise will be a problem, as you pointed out in your write-up. Back in my student house, I had a more affordable multichannel NAD power amp, so I opted for padding the horn down with 10-ohm resistors and pretty much stuck to the old crossover before the upgrade. Now would be the time to determine whether I can ditch some of those resistors and redo the final crossover network accordingly. Sadly, I'm not in Poland for the time being, so I can only test when I'm back. (I used 100+100uf caps instead of 300uf like in the picture).
Now, the setup is a bit more comprehensive. I have a horizontal bi-amp with two Cyrus Power X amps. A MiniDSP 2x4 HD is sitting between them and the Cyrus preamp. It's probably a pseudoscientific and irrational observation, but the MiniDSP looks awfully small and wonky compared to everything else around it—which is probably not a bad thing, but at the same time, it makes me raise an eyebrow.😀 I'm sure DSP has come a long way, but truly, I'm not sure how this little magic box can do so much with the signal without any audible artefacts. Well, the resolution is there, so I'm not complaining, but maybe someone can explain it to me so I can sleep a little better at night?🙂
In the long run, is it better to upgrade to a better DSP unit? Could I notice a difference or go fully active? And @perrymarshall any advice on how I should experiment with the crossover in the meantime?
Last edited:
You shouldn’t need a preamp at all unless you have a turntable.
The mini DSP Flex 8 is quite a bit better than the 2x4HD. it has at least 10dB less noise and better sound. The 2 x 4 HD makes enough noise that with a super high efficiency speaker like the live edge dipole, typical power amp will pass through a noticeable level of noise unless you use an attenuator or resistors like you mentioned.
I think the minimal passive crossover is a very simple attractive option, much simpler than triamping. I suggest you use measurements and IIR DSP to get as far as you can, then add some simple FIR all pass filters to correct the phase. @Carrywaterr
Your project looks great.
The mini DSP Flex 8 is quite a bit better than the 2x4HD. it has at least 10dB less noise and better sound. The 2 x 4 HD makes enough noise that with a super high efficiency speaker like the live edge dipole, typical power amp will pass through a noticeable level of noise unless you use an attenuator or resistors like you mentioned.
I think the minimal passive crossover is a very simple attractive option, much simpler than triamping. I suggest you use measurements and IIR DSP to get as far as you can, then add some simple FIR all pass filters to correct the phase. @Carrywaterr
Your project looks great.
Answering this question from @bhmartin who is new to the forum, from
https://www.diyaudio.com/community/...perrymarshall-live-edge-dipole-design.417072/
I would make the baffle just wide enough at the bottom to accommodate straight-sided wings. Probably 20-23" wide. The existing height is perfect.
You're welcome to make the baffle rectangular like @Carrywaterr does above, but you might also consider tapering the sides so the baffle is only 9-10" wide at the top. That will make the radiation pattern more even in the 1KHz region. I discuss that in this post:
https://www.diyaudio.com/community/...ompetition-updated-design.405104/post-7498546
This will also reduce the output of the 8" midrange below 400Hz so you may need to add 2-4dB of boost in that region in the DSP. That's no big deal, the Radian 5208 can handle that just fine.
https://www.diyaudio.com/community/...perrymarshall-live-edge-dipole-design.417072/
I would make the baffle just wide enough at the bottom to accommodate straight-sided wings. Probably 20-23" wide. The existing height is perfect.
You're welcome to make the baffle rectangular like @Carrywaterr does above, but you might also consider tapering the sides so the baffle is only 9-10" wide at the top. That will make the radiation pattern more even in the 1KHz region. I discuss that in this post:
https://www.diyaudio.com/community/...ompetition-updated-design.405104/post-7498546
This will also reduce the output of the 8" midrange below 400Hz so you may need to add 2-4dB of boost in that region in the DSP. That's no big deal, the Radian 5208 can handle that just fine.
Last edited:
Regarding tweaks @bhmartin :
The stock design exactly as described here is outstanding and will make most people very happy.
I recommend the MiniDSP Flex Eight over the 2x4HD. It's a very noticeable improvement.
If you're really ambitious, use an Acourate or Danville Signal digital crossover instead. I have not used those but from what others tell me, these are the best D to A / DSP combos on the market.
If you're even more ambitious, triamp instead of biamp and skip the passive lower crossover. I discuss that here https://www.diyaudio.com/community/...ompetition-updated-design.405104/post-7516202 and here https://www.diyaudio.com/community/...ompetition-updated-design.405104/post-7516229
I think the Radian 5208 is still an outstanding choice, and the beryllium diaphragm is better than the standard aluminum.
The Eminence woofer is a perfectly fine choice. If you're feeling extra ambitious, use the Faital 18XL1800, it has 20mm Xmax vs 8mm for the Eminence.
We would all love to see pics and measurements of your finished work, @bhmartin
The stock design exactly as described here is outstanding and will make most people very happy.
I recommend the MiniDSP Flex Eight over the 2x4HD. It's a very noticeable improvement.
If you're really ambitious, use an Acourate or Danville Signal digital crossover instead. I have not used those but from what others tell me, these are the best D to A / DSP combos on the market.
If you're even more ambitious, triamp instead of biamp and skip the passive lower crossover. I discuss that here https://www.diyaudio.com/community/...ompetition-updated-design.405104/post-7516202 and here https://www.diyaudio.com/community/...ompetition-updated-design.405104/post-7516229
I think the Radian 5208 is still an outstanding choice, and the beryllium diaphragm is better than the standard aluminum.
The Eminence woofer is a perfectly fine choice. If you're feeling extra ambitious, use the Faital 18XL1800, it has 20mm Xmax vs 8mm for the Eminence.
We would all love to see pics and measurements of your finished work, @bhmartin
Last edited:
“ I recommend the MiniDSP Flex Eight over the 2x4HD. It's a very noticeable improvement.”
Hi Perry, it’s been a while. Hope you are well. Regarding changing from the 2x4HD to the Flex Eight … do you by chance have a settings file for the Flex Eight that works with the original version of the live edge dipoles that I built?
Tom
Hi Perry, it’s been a while. Hope you are well. Regarding changing from the 2x4HD to the Flex Eight … do you by chance have a settings file for the Flex Eight that works with the original version of the live edge dipoles that I built?
Tom
IMO spending $100 on a measurement microphone, learning how to use, then measuring your speakers to see what is radiating into said room, should always be the first step. I am not aware of any controlled listening tests to support the claims made by treatment advocates, but would be happy to read any provided scientifically rigorous links. No question the visual aspect of those things can have a profound impact on those who believe in them, so YMMV. Again, just my opinion, but get the speakers right first. Measured in every direction, prior to measuring the subsequent "room" responses.
Tom,
Attached. Your Mileage May Vary - some of the levels may require tweaking (tweeter levels relative to mid/woofer channel in particular) but these configuration files should work very well for the MiniDSP Flex Eight for the Live Edge Dipoles.
Thanks, Perry! And I built a new listening room in my new house in Geneva National. Let me know if you are interested in coming up for some critical listening sometime. I’m sure I could scrounge up a good bottle of wine or bourbon….
Hope I´m right here...
I need some advice, or even two.
I´ve built a 2 way open baffle system with 2 Faital 15PR400 and a ESS AMT per side.
It´s an active system driven by a minidsp flex (2in 4out) and tube amps.
Sources are DELL notebook running Win11 and Qubuz/Roon (USB connection to minidsp),
a Debian-Linux based pc with Music Player Daemon MPD (connected via ethernet to home network),
and a turntable section (connected to minidsp analog in).
All in all, I am very happy with the sound except for a few small tears with the minidsp.
It sometimes has very short dropouts and artefacts that sound like ringing/rattling. Not pronounced, but noticeable. I don't use FIR filters, just EQ and crossover.
In addition, the USB connection between notebook and minidsp often breaks down while I am making changes to the settings. Then the changes I've made are lost, which is annoying. All components involved have the latest software updates installed.
A friend of mine runs an active 3-way system with Dynacord amplifiers that have a built-in dsp. The problems described do not occur there.
The Dynacord amps come from the professional sector. All changes made via dsp are implemented in real time. Switching between different presets is also done in real time - without the delay seconds of the minidsp...
Unfortunately, Dynacord does not offer a separate dsp unit.
Now my question: Who knows a good, affordable dsp that could replace the minidsp in my application? Gladly also from the professional sector.
I need some advice, or even two.
I´ve built a 2 way open baffle system with 2 Faital 15PR400 and a ESS AMT per side.
It´s an active system driven by a minidsp flex (2in 4out) and tube amps.
Sources are DELL notebook running Win11 and Qubuz/Roon (USB connection to minidsp),
a Debian-Linux based pc with Music Player Daemon MPD (connected via ethernet to home network),
and a turntable section (connected to minidsp analog in).
All in all, I am very happy with the sound except for a few small tears with the minidsp.
It sometimes has very short dropouts and artefacts that sound like ringing/rattling. Not pronounced, but noticeable. I don't use FIR filters, just EQ and crossover.
In addition, the USB connection between notebook and minidsp often breaks down while I am making changes to the settings. Then the changes I've made are lost, which is annoying. All components involved have the latest software updates installed.
A friend of mine runs an active 3-way system with Dynacord amplifiers that have a built-in dsp. The problems described do not occur there.
The Dynacord amps come from the professional sector. All changes made via dsp are implemented in real time. Switching between different presets is also done in real time - without the delay seconds of the minidsp...
Unfortunately, Dynacord does not offer a separate dsp unit.
Now my question: Who knows a good, affordable dsp that could replace the minidsp in my application? Gladly also from the professional sector.
before I threw away my miniDSP, I would look more closely at that USB connection. I used miniDSP for years without problems. One time a friend had an issue similar to yours but he was using a very long cable in his electrically noisy wood shop. It was solved by using a USB switch as a repeater between two shorter cables. It goes without saying that all the software/firmware involved should be brought up to date.
Iringing/rattling/dropouts/artefacts could be a result of a buffer running dry in the usb output stream. Somewhere in the software there should be a place to configure buffer depth
Iringing/rattling/dropouts/artefacts could be a result of a buffer running dry in the usb output stream. Somewhere in the software there should be a place to configure buffer depth
It´s a high quality 5m/15feet USB cable.
Software is up to date.
Hm, I couldn´t find infos about buffer adjustments under Win11...
Software is up to date.
Hm, I couldn´t find infos about buffer adjustments under Win11...
5m "should" work. IIRC, the spec limit is 20'. I would try a shorter cable just to rule it out.
Buffer size wouldn't be a windows thing but the audio player app or DAC driver. For example, there is a buffer size parameter in JRiver. So too, in the driver for my MOTU audio interface. You definitely get artefacts if the buffer isn't deep enough. MiniDSP software itself may have a buffer size parameter.
Buffer size wouldn't be a windows thing but the audio player app or DAC driver. For example, there is a buffer size parameter in JRiver. So too, in the driver for my MOTU audio interface. You definitely get artefacts if the buffer isn't deep enough. MiniDSP software itself may have a buffer size parameter.
@baggerbole your PC streaming system is complicated compared to the simple WiiM I’m used to. I would look there. I doubt the problem is the MiniDSP. I haven’t had any problems like the ones you describe.