The making of: The Two Towers (a 25 driver Full Range line array)

Hi percaval,

Think you maybe by mistake confuse Easy Audio Copy to Exact Audio Copy (EAC), well its most probably same German enthusiastic coders behind but first need license and have not one single settings to worry about and have totally simplest UI with magic engine running in auto, where free version as you say needs user skills and probably reading manuals to get the magic rips.
 
I'll go back to my "regular scheduled program" again.

A few posts ago I wrote about the mid/side chain and what it does. The reason for that was to create awareness of what the split signal looks like.

If we take my ambient information, which I have described as band passed, delayed (L-R) and (R-L) with a touch of phantom inserted, it should become clear that inserting back the phantom in this ambient information brings the signal closer to "being" stereo again.

I first ran only the (L-R) and (R-L). Though I noticed using the ambient part did create a difference in tonal balance. The center focal parts were sounding a lot dryer than the side information. I figured a real Haas Kicker would not be able to separate the center and side channels, so it should be obvious that even the best room would also affect the phantom material.

I have split the ambient part in two again, one being the (L-R) and (R-L) ambient signal and the other a (L+R) center part (this, in turn, also split in 2 channels to be able to run FIR correction and proper delay on each separate channel).

Both chains are followed by the reverb plugin, however each set of reverb can be dialed in separately this way. The adjustability is infinite.
This is exactly like I had done before, when BYRTT and xrk971 visited me.
However the EQ is a bit different in detail and I am still constantly playing with setting up the ambience to my liking. Dialing in the preferred room sound, so to speak. Keeping in mind to use the lowest levels I can get away with while keeping the benefits. As I said before: to hide the real room.
 
It looks like after years of 'more or less' rectangular shaped boxes, Magico has finally returned to smooth flowing shapes. First with their M-Project and later on the M3 and the new M6:

attachment.php


I only found out about Magico and their Magico Mini's long after I had planned my own design. In a search for using Aluminum baffles (after seeing koldby's sexy Aluminum columns, my true inspiration for that choice) showed me that their Mini design and my proposed arrays did have a lot in common. The stacked Birch ply, Alu as baffle and the overall shape.

attachment.php
attachment.php


They moved on from the Mini to the more square shaped enclosures. That really surprised me. To me, the Mini enclosure shape made way more sense...

attachment.php


Their cheaper (relatively speaking) "S" line featured more smooth, round shapes in newer generations. They finally seem to return to minimum diffraction designs again in the top of the line products. I wonder if they still employ their elliptical crossovers.

Just an observation, having never seen or heard a Magico speaker in real life I wouldn't know how they sound, I did like the look of that Mini though :).

I like the new feet on that M6 too :). Even that 'feels' a bit familiar.
 

Attachments

  • Magico M.jpg
    Magico M.jpg
    119 KB · Views: 358
  • Magico_mini.jpg
    Magico_mini.jpg
    128.1 KB · Views: 362
  • Magico-Q1-casse-monitor.jpg
    Magico-Q1-casse-monitor.jpg
    15.5 KB · Views: 377
  • BaffleDetail3.jpg
    BaffleDetail3.jpg
    38.4 KB · Views: 366
Last edited:
Well, in Magico's case they didn't exactly go to cheaper construction..

They went from this:
attachment.php


To an all metal construction (using Aluminum and Copper):
IMG_3201e.jpg


I do get that straight lines is the cheaper way (as everything is milled out of solid Alu) and they only kept a bulging baffle, milled out of a solid piece of aluminum (just like the Mini)...

The new M6 has carbon fiber sides which are easier to sculpt to round shapes..
Magico | M6

They've never been a company destined to save a few bucks in materials. Even the M6, with it's 'light' carbon fiber construction still weighs 177 kg each :eek:.
Their product prices do reflect that.
 

Attachments

  • Magico Mini Bracing.jpg
    Magico Mini Bracing.jpg
    22.8 KB · Views: 331
foot in mouth again... serves me right to post without looking things up again.

At first sight, I thought the new ones were pdf...

So... I went to their site to learn a bit more about those... yep, not exactly cutting corners, are they?

Haven't looked up prices for those carbon or alum monsters... not sure I want to! Ha ha!
 
Their subwoofer was an inspiration for my own sub design:

attachment.php


And my proposed "mini" version, only slightly different in appearance :D:
subassembly.jpg


However I doubt I can mimic their build materials:
attachment.php


I like the "all out" approach they have. Like I said though, I have never heard any of their speakers and it's not likely I will any time soon. In the video I linked I did like the "Form Follows Function" statement, which I believe is the approach I took and have mentioned before.
In my humble opinion that does not have to lead to ugly designs.
 

Attachments

  • image.jpg
    image.jpg
    31.3 KB · Views: 355
  • 3s21.jpg
    3s21.jpg
    500.6 KB · Views: 333
wesyaso,

Picture in your post of the sub-woofer that they are making (looks like the model with 18 inch drivers)- according to the Magico website that sub weighs 258 kg (570 lbs)! :eek:

I wonder of it comes with safety instructions on what sort of floor to place on...
 
Last edited:
Got the drivers, lacked the time ;). They have been on my mind all year, I just couldn't find the time to start building, but I am gathering tools, like router bits etc..

Had a bit of a strange year that went by way too fast for my liking... :)

They will happen though. I've got the build planned, have most materials. Still not sure about the amplifiers to get for them.
I'm hoping to get more spare time in the spring of next year. I'll take that time to figure out the details.
I might try some Jensen isolators on the Xonar Essence card I run to see if that cleans up the power hum on the signal.
The mains are completely isolated from hum with SP/DIF, my ambient channels are less clean in comparison.
A separate power supply for the soundcard might work too, not sure which direction to take there.
 
Last edited:
Mid/Side processing again...

I ran a few virtual checks on the mid/side processing. With the measured response as input I checked the graphs that result from each way the input signal was configured. So no swapping of channels, playing with input signals instead.

- Measured impulse as positive left and negative right channel.
- Input only one channel, either left or right
- Stereo input (as if it were phantom material)

Each of them will react different with the mid/side processing chain.

The left with inverse right will cancel out in the phantom part of the chain, so it will be a pure "side EQ" result.
When we input only a left or only a right signal we get an average between the mid and side EQ.
The stereo input will give the phantom EQ-ed result as output.

I made a graph of it, while comparing it to the JBL steady-state room curves:
SPLgraph.jpg


If you look closely you'll see I made a few changes to the EQ scheme as published on fluid's thread. Largely it's still the same, except I added a few changes in the "side EQ chain" that make the average "side" become a bit more flat.

Seeing it with the JBL graphs overlaid I guess I'll have to try a few tweaks and listen. :)
It's pretty close as is... I did mention the similarities before.
 
Last edited:
....A separate power supply for the soundcard might work too, not sure which direction to take there.


Know its not a separate PSU but will let you know if Seasonic's newest computer PSU flagship series will help here, its tested to mythical performance and released as 650/750/850w versions but a new 550w is shipping now and as soon it hits the shelves will test one in computer that have similar family ASUS soundcard, can say from past that when upgraded computer PSU to Corsair AX series it helped a lot for that family of ASUS soundcards so is optimistic about investment. With Corsair AX performance is good enough, one can hear some hum with ear into speaker drivers when volume is set above twelve o'clock but the little hum is totally killed when music plays so don't think its a big problem and probably because my power amps have separate mono PSU, most trouble seems to be reported when power amps L/R channels sits on same united psu.

Link to review: The Seasonic PRIME Titanium PSU (650W, 750W, 850W) Review: Mythical Performance


I ran a few virtual checks on the mid/side processing. With the measured response as input I checked the graphs that result from each way the input signal was configured. So no swapping of channels, playing with input signals instead.

- Measured impulse as positive left and negative right channel.
- Input only one channel, either left or right
- Stereo input (as if it were phantom material)

Each of them will react different with the mid/side processing chain.

The left with inverse right will cancel out in the phantom part of the chain, so it will be a pure "side EQ" result.
When we input only a left or only a right signal we get an average between the mid and side EQ.
The stereo input will give the phantom EQ-ed result as output.

I made a graph of it, while comparing it to the JBL steady-state room curves:
SPLgraph.jpg


If you look closely you'll see I made a few changes to the EQ scheme as published on fluid's thread. Largely it's still the same, except I added a few changes in the "side EQ chain" that make the average "side" become a bit more flat.

Seeing it with the JBL graphs overlaid I guess I'll have to try a few tweaks and listen. :)
It's pretty close as is... I did mention the similarities before.


Nice curves there you really good at this.

Admit its a bit over my head to figure out the MidSide mix but down the road when getting stereo will look into it and hope my brain can be trained up to figure it out ;), also because after trying their study for head phone curves and how good it sounds which by the way is based on sound from good speaker in a good room can imagine its serious good target curve.
 
Eager to hear more about that power supply. Right now I'm just running the original Lenovo supply but would be willing to upgrade if I can get a model that fits.

Originally, my most preferred curve, which I got by listening, adjusting, and repeat until I was happy was very close to this one. Lately I have been deviating a little in recent times. Mainly the top end was a little hotter from what I ran before. We can get used to a lot, however my preference still is with a little more decline on the top end, just like that JBL curve has. I keep coming back to it. The only thing I clearly remember is that I prefer the 200-300 Hz to be just that tiny bit up from the Harmon/JBL curve. A touch of warmth, which is easily overdone. Too much there and it starts to sound muddy.
Again the devil is in the details.

Remembering how close I was to the Harmon/JBL curves will make me experiment by ironing out the last few deviations and try it again.
I've done so before and I'm planning to repeat that test, after all the little details that have changed over time in my processing chain.

Disclaimer: This isn't an easy test, as the input (a measured result) will skew the results if there are large differences between the left and right channels.
The mid/side routine will react to everything that deviates between the two channels. So a little 'helper' EQ was needed (only for the phantom channel) to produce the graphs, but that should not be needed in regular music playback. That's why I advise anyone willing to try it, to place any mid/side chain before every other EQ tweak in the chain. Most EQ work can be in any order you'd like and still add up fine, not this one.


BYRTT, if you ever want to know more about how this all works (mid/side) I'll help you out. :) It can be a bit confusing at first.
 
Last edited:
Touching offer you will help me out how complicated (mid/side) works when time come thanks for that, regarding Seasonic PSU as 550w model it was announced at TechPowerUp some months ago but just checked Sesonic website and they don't list or mention any 550w model anywhere yet so will probably order their SSR-650TD model next week so we can get to know about its performance when also physical wired to sensible audio system, should it happen not work better than present Corsair AX series money is not totally wasted in top series Seasonic offer a whooping 12 years guarantee, and probably last 2-3 motherboard generations.