Thanks for posting that link Urizen, super moderator. Pretty similar to what many of us have been writing right here on AK for years now. Will this change anything? Extremely doubtful. In fact, I'm wonder if any facts or opinions shared on the Internet has the ability to change anyone's minds? If so, probably pretty rare when it does.
Alas, although I classify this article as a "good one" (since it aligns with my own experiences, measurements, and opinions), I really have to question if the author actually reads what he writes? No, make that, actually thinks about what he writes? Let me explain...
He starts out by saying that the DC component (resistance) has a bigger effect than L and C, so the AC components (L & C) can be ignored. Well, actually he says that "normal speaker cable doesn't have significant capacitance or inductance," and then proceeds to ignore them. Then he mentions some Toole/Olive research which suggests "midrange peaks of just 0.3 dB were audible" and sets that as a limit to his chart of calculated max lengths for various gauges of speaker cable with a particular hypothetical speaker. Then, he makes real-world measurements, followed by a short discussion of these measurements. Right off, he dismisses the differences between the two 12ga cables and focuses on the performance of the 24ga cable. I find the difference of the two 12ga cables to be of extreme interest. They had a max reported difference of 0.4dB, and Toole/Olive state that 0.3dB difference is audible, but he chooses to ignore it. Now, I do realize Toole/Olive stated a low-Q resonance was audible at 0.3dB difference, and that the reported frequency range of the difference was less than an octave, which isn't high-Q, but it may not be low-Q. But, since these are both 12ga cables, the DCR is going to be the same, so the L and C differences of the cables are responsible for this difference in measured performance. Yet, he chooses to proclaim L&C aren't significant. What drives this point even further home is that the length of cable used for this test is 20 feet, and his chart says that a 12ga cable can be 47.23 feet long and not impact the sound more than 0.3dB for his calculated example. Had he used a 47 foot cable, the measured differences could be as high as 0.8 or 0.9dB. His data suggests to me that L and C are indeed worthy of our attention.
I applaud his efforts to make measurements of the actual acoustic output of his speakers. He goes to some length to describe the hoops he had to jump through to come up with good data. From my own personal experiences, I can agree that getting repeatable acoustic measurements to fractional dB levels is NOT an easy task. The thing is, he didn't have to put himself through such agony. He didn't have to move computers out of the room, do repetition for averaging, make sure he didn't move the speakers or mics when changing cables, or worry about typical environmental noises. He could have measured the signal electrically at the speaker terminals to arrive at the same information. The losses imparted by ALL cables are rather easily measured.
All my nit-picky comments aside, I still think it is a decent article and I sincerely hope that it sparks at least a glimmer of wonder in the reader's minds that perhaps cabling is indeed an important aspect of a high performance hi-fi.
