"Shadow of The Colossus" build thread

Here's the Hungarian Yuichi Arai horns prior sanding and wood oil-waxing.








I should buy a equally wonderful amps/pre for these super speakers. Tube electronics despite their positive sides, cut too much infra energy. At least I'm not yet ready to make that sacrifice. Maybe in the future when things settle.


Someone pls drop me a killer recommendation for the amp/pre/integrated😀. I have been eyeing some chi-fi as a bang for the buck. Dartzeel inspired amps, Accuphase clones, Shanling A3.2 Sanken magic transistor and H2 dominant very cleanly made beauty, Jungson stuff etc. I'm positive there are deals of the decade to be made.


Hungarian Yuichi Arai?
 
Those Fane's h3 values are very good. At wich drive voltage were those measurements taken ?

Regards

Charles

Much under 2,83V but inductance/motor related h3/h5 level does not change much at higher voltages at mid band (at bass yes), it's not like h2/h4 which are suspension etc. related and rise pretty monotonically rgds. drive voltage.

I watch mainly mid band h3. Shorting rings/low Le mainly affects h3 above the system resonance. Some might watch much more bass performance.
 
Last edited:
Size and coverage angles definitely make a difference in the listening experience.

If I were to design a synergy (inspired) system, it would probably consist of a big conical waveguide/Arai type of horn mounted inside a large J3-like horn.
The outer horn could be loaded with 18" or 15" woofers and 12" or 10" mids. 2 of the latter may breath through the mid/high horn.
 

Attachments

  • Large-Format_0003_CJ-1894-Glass-Front-View2.jpg
    Large-Format_0003_CJ-1894-Glass-Front-View2.jpg
    60.4 KB · Views: 484
Last edited:
Size and coverage angles definitely make a difference in the listening experience.

If I were to design a synergy (inspired) system, it would probably consist of a big conical waveguide/Arai type of horn mounted inside a large J3-like horn.
The outer horn could be loaded with 18" or 15" woofers and 12" or 10" mids. 2 of the latter may breath through the mid/high horn.

Definitely worth idea to fine tune. I have also played with nested horns on idea level. Colossus could also be said to be a nested design, but the wavequite of the woofers is 180deg😀 And a more brutal force "iron curtain" Sd-solution. Sd-scaling is non-elegant (and costly) but easy and it always works🙂 And the wooden parts' structural integrity is easy to achieve. With big horns similar acoustical deadness of every panel (like a thick and braced cabinet) is (also) a great challenge without them becoming unmovable objects🙂
 
Valid points!!

Indeed, your Colossus consists of a mid / high horn - for which controlled directivity is most important, nested in a 180° horn wall.
I like the modular design of it, and while not exactly portable, the individual cabs can be moved by 1 person.

A single Danley J3 weighs almost 200kgs. It seems quite complicated to design such a synergy horn in modular fashion.
 
After 1-2 weeks of full transistor signal path I changed it to whole tube setup.

Tube signal path is ultra-purist and consists of Audio Note Kits 5.1 non-oversampling DAC with modded I/V-transformer (Noguchi finemet) and modded SE-tube stage after it (with finemet opt). Pre is TVC (permalloy) + SE-tube tube stage (finemet OPT). The amp is capacitor coupled SRPP stage driving SE-UL amp with KT88 output tube (finemet opt also).

Listen to the differences (Im not saying they can be heard even 30% by recording the sound but still😉):

Transistor: Colossus - Sonique (Full Transistor) - YouTube
Tube: Colossus - Sonique (Full Tube) - YouTube

I think I might need to bi-amp the Colossus to the moon.😀
 
Last edited:
After 1-2 weeks of full transistor signal path I changed it to whole tube setup.

Tube signal path is ultra-purist and consists of Audio Note Kits 5.1 non-oversampling DAC with modded I/V-transformer (Noguchi finemet) and modded SE-tube stage after it (with finemet opt). Pre is TVC (permalloy) + SE-tube tube stage (finemet OPT). The amp is capacitor coupled SRPP stage driving SE-UL amp with KT88 output tube (finemet opt also).

Listen to the differences (Im not saying they can be heard even 30% by recording the sound but still😉):

Transistor: Colossus - Sonique (Full Transistor) - YouTube
Tube: Colossus - Sonique (Full Tube) - YouTube

I think I might need to bi-amp the Colossus to the moon.😀

Framing the transistor vs tube dilemma in such black-and-white terms is overly simplistic.
Try some Accuphase class A mosfet amps, and then we'll talk ;-)

Marco
 
Comparing ('framing') tube setups vs transistor setups according to some sort of protocol seems almost impossible.
I don't think it's Legis intention to draw 'valid' conclusions from the posted videos, but merely to demonstrate some differences in the rendering.

BTW, Accuphase - while topnotch - appears not to be the transistor amplifier (brand) to end all transistor amps, as far as measured performance is concerned.
 
Last edited:
The suspension of the 12" has gotten less lossy and Qms (in a box) has risen from 6,7 --> 10,3. The suspension also feels more sensitive and springy when pushing a driver.

Just like with the 15" woofer modules the impedance peak is almost as high as with a naked driver because I only line the box with felt and don't stuff it. Doesn't kill the kick.😛

Impedance of a single mid module at start and after ~1week. All four drivers connected (giving 8ohm nominal imp.), which share the same air space. They effectively measure like one single driver (sharp impedance peak), the shared air space evens out their minor differences nicely.

2c0x99L.jpg
 
Last edited:
The suspension of the 12" has gotten less lossy and Qms (in a box) has risen from 6,7 --> 10,3. The suspension also feels more sensitive and springy when pushing a driver.

Just like with the 15" woofer modules the impedance peak is almost as high as with a naked driver because I only line the box with felt and don't stuff it. Doesn't kill the kick.😛

Impedance of a single mid module at start and after ~1week. All four drivers connected (giving 8ohm nominal imp.), which share the same air space. They effectively measure like one single driver (sharp impedance peak), the shared air space evens out their minor differences nicely.


That impedance plot looks very good, much better than in the specsheet. It's an indication of low mechanical losses.
 
Last edited: