The video clearly shows a very small difference in roll-off which is shown before so no surprise.
That's not under debate. What's being questioned is the audibility of it.
When I do some calculations based on the amplitude and divisions of the scope image at 12kHz I see (c.a.) - 0.1dB attenuation by the Isolda and -0.3dB by the Monster cable.
Now below some calculations based on the given data of the cables:
load 8 Ohm
Isolda = 0.189 Ohm
Monster = 0.294 Ohm
Isolda:
8/(8+0.189) = 0.977x = -0.2dB attenuation in the load (assuming the total cable round trip = 0.189 Ohm) This seems likely as acc. to the specs on the website one would expect 0.12 Ohm for 7 meters.
So the attenuation of the loudspeaker signal due to resistive component in the Isolda cable is 0.2dB and when using the website - 0.13dB.
Monster:
8/(8+0.294) = 0.965x = -0.3dB attenuation in the load.
The speaker (in this case the dummy load) would thus sound 0.1dB (or almost 0.2dB) louder with the Isolda due to Ohmic causes.
Of course there is some small treble roll-off at 20kHz which is clearly seen as the differences get smaller in lower frequencies.
As there are no exact values it looks like the treble roll-off is about -0.1dB to 0.2dB more at 12kHz compared to 500Hz.
There is more:
The reason it shows an estimated difference of 2.5dB (can't resolve 5dB in this case) is because the DC resistance where you measure across has the same current (the one through the dummy load) but is in fact a different shunt resistor.
0.147/0.0945 works out as 1.5555 = 3.8dB
That would be the actual difference assuming the resistance of the cable is measured correctly.
In any case it is obvious we are talking about
different dB's
You talk about
dB differences between different shunts, and I talk about the
differences between the voltages across the load (which would normally be a speaker which one would listen to).
We would thus hear a 0.1dB level differences between the cables and not 3.8dB
Now... assuming the 12dB (x factor 4) difference you reported earlier and is shown in the video with the noise between the lower frequencies and the highest ones. The calculation is as follows (assuming perfect FR and 100% Ohmic Isolda).
At 160Hz the attenuation of the monstercable = -0.3dB but at 20kHz the 'seemingly increase of resistance due to reflections (not skin effect ?) would be 8/(8+(4x0.294)) = 0.87x = -1.2dB
So while the Isolda would have a flat FR the FR at the dummy load would be sloping downwards by 0.9dB.
That would be audible
assuming your 12dB difference in high frequencies across the shunt would really be there between 160Hz and 20kHz.
So
your 12dB is
my 0.9dB in reality.
But the rol-off is there. NO doubt about it.
You say the reason is reflections but do not hear you talk about the
skin effect.
From an online skin effect calculator found on the web I got the following results.
freq. factor resistance at that frequency
100Hz = fact 1.0 * 0.294 Ohm = 0.294 Ohm
1kHz = fact 1.05 * 0.294 Ohm = 0.3087 Ohm
10kHz = fact 2.24 * 0.294 Ohm = 0.658 Ohm
20kHz = fact 3.0 * 0.294 Ohm = 0.882 Ohm
100kHz = fact 6.5 * 0.294 Ohm = 1.911 Ohm
This will give factor 3 increase in treble at 20kHz due to the skin effect = 10dB more treble across the cable. This close to the estimated 12dB in your measurement ?
It is pointless to do this for the Isolda as the conductor is very thin so all current will flow around the outside anyway (1-nil for Isolda).
So a Monster cable of 7 meter length will have -0.3dB less output at 20kHz than at 1kHz due to skin effect alone and -0.9dB at 100kHz.
It appears to me that part of the 'roll-off' can at least be attributed to skin effect.
Also one has to realize that the scope amplitude measurements have an error as well.
In any case I am not really worried about 0.something treble roll-off in a 7 meter cable. It will be less in more reasonable lengths.
To me it is clear... there are very, very small differences between cables and not worth fussing about. Certainly not clearly audible and also not the root cause of the reported sound differences.
It is clear to me where the reported '12dB' comes from and can understand why you would want to show this.
In my opinion it would be more honest to speak about the consequences and level differences at the load and not in the cable as one then must 'convert' those numbers back to the real world.
To me it seems the biggest part may be skin effect but one can only prove that by using more spacing between the conductors in the cable and do the measurements again.
This could be done by laying 2 Isolda cables on top of each other, evenly spaced and using one wire in each cable.
That way one could prove the waveguide theory to be valid when the treble roll-off changes.
