After owning a pair of Aerius for a few months, I felt there was something missing that I had with other ESLs. I'm very familiar with old JansZen designed panels and RTR made for infinity. Made an awesome DIY using 28 panels that had it all. Played reference level from 24Hz to 24KHz with only 2 correction points.
They melted in a fire.
Realizing that I would never go back to box speakers and the fact that I have had no income for the last year and had to buy a new everything since my insurance didn't cover my stuff, I started looking for a set of ESLs and was looking for a decent set of Old Rockford Fosgate Spectrums and see if I could get them sounding close to what I had.
I got lucky and spotted the Aerius for cheap and had some worries about how I read that ML panels just don't hold up like others. I really never considered it since my last panels were made in 1959, 1969 and 1971 and none had any issues other than rebuilding or fixing the power supplies.
The only issue with the older hybrid speakers that Infinity and JansZen made was the woofer. I had no problem using modern woofers and I overlapped the panel crossover point to have a much less detectible transition.
So, I figured I'd take a chance and got the MLs very quickly and they seemed to work fine just sitting randomly in my front hallway. After using them and trying to capture that same magic I had before, I couldn't get the results I wanted.
Lack of any real bass was obvious and I figured could easily be addressed, but the lack of perfect imaging was not there no matter where I placed them. I could run a sweep and see smooth results but there was something missing. I felt the curve may be the problem. I tried curving my panels when I made my JansZen DIYs and found that it took away from the perfect headphone like bubble that encompassed me.
Since I had 14 panels (10 - 5"x5" and 4 - 2.5"x5") per side, I could arrange them in any way I wanted. I thought line array might be best. I figured 2 columns or 7 with woofers on the side but tried curving them since the JansZens were removed from curved cabinets. I figured out why the 30 and 130 were lackluster in the 1st place. The attempt to disperse was not ideal for their operation and sealing the back was the 2nd problem that held them back from stardom.
Infinity had a different idea that only angled 1 panel about 12 degrees. It's too bad that they failed to make mirrored sets and the entire design was super flawed from the get go. Plus their woofer was just awful, they sounded horrible and I had no problem ripping all panels out and making a better design.
My year long experiments yielded some valuable info into how the panels layout and crossovers needed to be picked properly to get that perfect image.
The fact that ML has chosen the curved path may have been out of creative thinking, patent availability, gimmick or research, I can't say, but I do understand what they were hoping to achieve. The fact that surround sound had a very important inclusion factor was probably a big point that stopped people from investing in the ESL game. Price was probably a big factor as well as size and WAF.
What I see happening with the curve is the rear waves cross the path instantly as the are produced. This may offer a better off axis test result but it losses the effect that make ESLs so awesome. My Aerius had a good sound but it was just like owning a good set of box speakers that could be found used for cheap (Snell Es come mind, owned them and loved them till my DIY panels blew them away). I paid $60 for them.
I had $1300 in parts for my DIY panels, so the comparison is just based solely on sound.
I decided to try and get more control over the MLs by bypassing the crossover and got out my surviving active (2 year old Behringer 3-way), my old amps made it through the fire as well. I tested the panels full range and got this result.
The scale is represented by a percentage not a db so they're not nearly as rough as this plot may look. In REW this would be a flat line once the panel reaches 500Hz.
I found I get better, faster, much easier to do readings that are as good or better than REW using an app called SPECTRUM RTA made by a very nice Russian coder who will take them time and chat if you have any questions. What I found was Moto phones come with an exceptionally well made mic and work very well with the app.
I've tried every all out there and none of them are even worth the time it takes to download them. The accuracy of this app is outstanding and it has every feature needed to complete the task.
Ok, back to the topic, now I know where the panels operate and can use that to try and get the bass to a point that it's usable. I swapped out the stock woofers for some other 8" subs that came from some failed active subs, I have quite a collection of them and tried 8 different sets, since they aren't marked I'd have to guess that they are either Polk or Infinity since those amps failed quite a bit.
It appears that I deleted my first sweeps before changing the woofers but the plot had bass starting at around 110Hz and was dip and peak the entire way. So I am show the new single woofer plot running 1 speaker with close mic (2 feet away) just as reference to see how it performed, then compared it to the new dual woofer idea I had in my head.
This is a single replacement woofers with half the stuffing removed from the jam packed enclosure.
I went through a series of sweeps that start with the 2nd driver mounted facing to the rear and changed polarity, then turned it aiming at the back of the front driver with the magnet outside the cabinet. To get this to work, the cabinets have to be truly sealed and not just saying sealed as ML has done. The cabinets are far from being sealed and hot glue helped me achieve a much closer seal.
These next pics will show response from 1 speaker tested about 4 feet away as I went through the different mounting and wiring (refer to the image names).
The entire sweep was done to show how they interacted with panels, you'll notice the panels have changed plots with each change.
Look at the percentage as I changed them, no only was there deeper but also a large gain. Woofers are 8Ω wired in parallel.
The next plot shows after EQ (dual 31 band analog running into a BBE maximizer set to flat to correct phase) correction points are at 120Hz and 60Hz.
This is the final test on the one completed speaker, I just finished the other one before I started this post i have yet to run a sweep because I've been amazed at what has happened to the overall sound.
The improvement is vast, the imaging issues a gone, I can hear both speakers no matter where I'm sitting and the sound is now full and has a definition that sounds live. There's no missing weird suckouts or obvious peaks. The vocals have a completeness that's remarkable. The crossover is being used as a 2-way with 450Hz dialed in.
I can't even believe the difference in every aspect of these speakers. I just read a live track and it sounded outstanding. Typically live recordings lack everything good. I didn't realize it was live until it ended and the crowd started cheering.
There is no longer a drastic change in sound when standing up. The rear woofers reinforcement refecting it's out of phase signal with the panels out of phase signal fills the room with perfect sound now. There's only a bit of reflection from the crossing waves. The basd is accurate and detailed. It's not like my last set up but these speakers only have ⅓ the footprint and look far better.
I'll have to post the rest of the shots that show the inside and completed outside in another post. I can only do 10 each time.
Thanks for reading, this project can be done in a day, I ended up using 2 Mirage 8" subs from my BP150i active sub that survived the fire. That sub has always sounded excellent and was going to be the woofer part of my DIY build but couldn't find another to do it.
Amps are old 1980s SS that put out 400W per channel and are 2 and 1 ohm stable. The second LED doesn't even light up and there's plenty of sound to fill a room that has a 14' vaulted ceiling in an opening about 25' by 20'.
They melted in a fire.
Realizing that I would never go back to box speakers and the fact that I have had no income for the last year and had to buy a new everything since my insurance didn't cover my stuff, I started looking for a set of ESLs and was looking for a decent set of Old Rockford Fosgate Spectrums and see if I could get them sounding close to what I had.
I got lucky and spotted the Aerius for cheap and had some worries about how I read that ML panels just don't hold up like others. I really never considered it since my last panels were made in 1959, 1969 and 1971 and none had any issues other than rebuilding or fixing the power supplies.
The only issue with the older hybrid speakers that Infinity and JansZen made was the woofer. I had no problem using modern woofers and I overlapped the panel crossover point to have a much less detectible transition.
So, I figured I'd take a chance and got the MLs very quickly and they seemed to work fine just sitting randomly in my front hallway. After using them and trying to capture that same magic I had before, I couldn't get the results I wanted.
Lack of any real bass was obvious and I figured could easily be addressed, but the lack of perfect imaging was not there no matter where I placed them. I could run a sweep and see smooth results but there was something missing. I felt the curve may be the problem. I tried curving my panels when I made my JansZen DIYs and found that it took away from the perfect headphone like bubble that encompassed me.
Since I had 14 panels (10 - 5"x5" and 4 - 2.5"x5") per side, I could arrange them in any way I wanted. I thought line array might be best. I figured 2 columns or 7 with woofers on the side but tried curving them since the JansZens were removed from curved cabinets. I figured out why the 30 and 130 were lackluster in the 1st place. The attempt to disperse was not ideal for their operation and sealing the back was the 2nd problem that held them back from stardom.
Infinity had a different idea that only angled 1 panel about 12 degrees. It's too bad that they failed to make mirrored sets and the entire design was super flawed from the get go. Plus their woofer was just awful, they sounded horrible and I had no problem ripping all panels out and making a better design.
My year long experiments yielded some valuable info into how the panels layout and crossovers needed to be picked properly to get that perfect image.
The fact that ML has chosen the curved path may have been out of creative thinking, patent availability, gimmick or research, I can't say, but I do understand what they were hoping to achieve. The fact that surround sound had a very important inclusion factor was probably a big point that stopped people from investing in the ESL game. Price was probably a big factor as well as size and WAF.
What I see happening with the curve is the rear waves cross the path instantly as the are produced. This may offer a better off axis test result but it losses the effect that make ESLs so awesome. My Aerius had a good sound but it was just like owning a good set of box speakers that could be found used for cheap (Snell Es come mind, owned them and loved them till my DIY panels blew them away). I paid $60 for them.
I had $1300 in parts for my DIY panels, so the comparison is just based solely on sound.
I decided to try and get more control over the MLs by bypassing the crossover and got out my surviving active (2 year old Behringer 3-way), my old amps made it through the fire as well. I tested the panels full range and got this result.
The scale is represented by a percentage not a db so they're not nearly as rough as this plot may look. In REW this would be a flat line once the panel reaches 500Hz.
I found I get better, faster, much easier to do readings that are as good or better than REW using an app called SPECTRUM RTA made by a very nice Russian coder who will take them time and chat if you have any questions. What I found was Moto phones come with an exceptionally well made mic and work very well with the app.
I've tried every all out there and none of them are even worth the time it takes to download them. The accuracy of this app is outstanding and it has every feature needed to complete the task.
Ok, back to the topic, now I know where the panels operate and can use that to try and get the bass to a point that it's usable. I swapped out the stock woofers for some other 8" subs that came from some failed active subs, I have quite a collection of them and tried 8 different sets, since they aren't marked I'd have to guess that they are either Polk or Infinity since those amps failed quite a bit.
It appears that I deleted my first sweeps before changing the woofers but the plot had bass starting at around 110Hz and was dip and peak the entire way. So I am show the new single woofer plot running 1 speaker with close mic (2 feet away) just as reference to see how it performed, then compared it to the new dual woofer idea I had in my head.
This is a single replacement woofers with half the stuffing removed from the jam packed enclosure.
I went through a series of sweeps that start with the 2nd driver mounted facing to the rear and changed polarity, then turned it aiming at the back of the front driver with the magnet outside the cabinet. To get this to work, the cabinets have to be truly sealed and not just saying sealed as ML has done. The cabinets are far from being sealed and hot glue helped me achieve a much closer seal.
These next pics will show response from 1 speaker tested about 4 feet away as I went through the different mounting and wiring (refer to the image names).
The entire sweep was done to show how they interacted with panels, you'll notice the panels have changed plots with each change.
Look at the percentage as I changed them, no only was there deeper but also a large gain. Woofers are 8Ω wired in parallel.
The next plot shows after EQ (dual 31 band analog running into a BBE maximizer set to flat to correct phase) correction points are at 120Hz and 60Hz.
This is the final test on the one completed speaker, I just finished the other one before I started this post i have yet to run a sweep because I've been amazed at what has happened to the overall sound.
The improvement is vast, the imaging issues a gone, I can hear both speakers no matter where I'm sitting and the sound is now full and has a definition that sounds live. There's no missing weird suckouts or obvious peaks. The vocals have a completeness that's remarkable. The crossover is being used as a 2-way with 450Hz dialed in.
I can't even believe the difference in every aspect of these speakers. I just read a live track and it sounded outstanding. Typically live recordings lack everything good. I didn't realize it was live until it ended and the crowd started cheering.
There is no longer a drastic change in sound when standing up. The rear woofers reinforcement refecting it's out of phase signal with the panels out of phase signal fills the room with perfect sound now. There's only a bit of reflection from the crossing waves. The basd is accurate and detailed. It's not like my last set up but these speakers only have ⅓ the footprint and look far better.
I'll have to post the rest of the shots that show the inside and completed outside in another post. I can only do 10 each time.
Thanks for reading, this project can be done in a day, I ended up using 2 Mirage 8" subs from my BP150i active sub that survived the fire. That sub has always sounded excellent and was going to be the woofer part of my DIY build but couldn't find another to do it.
Amps are old 1980s SS that put out 400W per channel and are 2 and 1 ohm stable. The second LED doesn't even light up and there's plenty of sound to fill a room that has a 14' vaulted ceiling in an opening about 25' by 20'.
