[MacedonianHero]

Noted, no measurements were shown to definitively offer quantitative data, but still an interesting article:
 
http://www.stereophile.com/content/minnesota-audio-society-conducts-cable-comparison-tests-0
 

 

[mikeaj]

Anybody have any idea what the bottom graph means?  Are the four curves offset from one another for visibility, or are the signal levels actually different?  What's the deal with the green one?
 
As is usual, methodology could use some improvement, particularly with regards to only doing single blind testing (looks like the audience can see the people doing the switching, maybe? possibly the cables plug in with a different sound, or make different sounds when being handled?).  And as they point out, they didn't do any analysis of the data for significance.

 

[stv014]

Quote:

Anybody have any idea what the bottom graph means?  Are the four curves offset from one another for visibility, or are the signal levels actually different?  What's the deal with the green one?

 
The signal levels could be slightly different because of different cable impedances, especially with 4 Ohm loudspeakers. That would also explain the frequency response variations with the highest impedance cable, although it does not agree with the impedance vs. frequency characteristics described for the loudspeakers.

 

[mikeaj]

The signal levels could be slightly different because of different cable impedances, especially with 4 Ohm loudspeakers. That would also explain the frequency response variations with the highest impedance cable, although it does not agree with the impedance vs. frequency characteristics described for the loudspeakers.

I was thinking that too, and then...oh whoops I didn't see the (dB) after Amplitude on that graph the first time. The cables are off by a couple tenths of a dB from each other? That seems maybe plausible with a 4 ohm loudspeaker, though as you say (I noticed this earlier), it doesn't match what they say about the loudspeaker impedance over frequency.


and also, anybody honestly doing an experiment like this, and thinking for more than a couple minutes about what to do, probably would have tried an A vs. A (or B vs. B, etc.) test too...