For the last time, cables DO matter!

[preload]

Welcome "Lost-Boy"!

No idea what you're referencing

 

[kemmler3D]

Davis et al, JAES 1991;39:461-468.

Less than half a decibel at 20khz per 10 feet isn't worth losing sleep over IMO. Also, not sure what the specs of these cables would be. Presumably something kinda crappy? These would be the worst-measuring cables I've seen on the internet, actually.

 

[Axo1989]

Less than half a decibel at 20khz per 10 feet isn't worth losing sleep over IMO. ...

It's a small difference for sure. But we insist on 0.1 dB matching for best practice controlled/blind testing, so that variation from 2 kHz up (per the graph) is certainly within scope of differences we prefer to eliminate.

 

[julian_hughes]

It's a small difference for sure. But we insist on 0.1 dB matching for best practice controlled/blind testing, so that variation from 2 kHz up (per the graph) is certainly within scope of differences we prefer to eliminate.

0.1 dB? Urrgh. Slackers.

 

[kemmler3D]

It's a small difference for sure. But we insist on 0.1 dB matching for best practice controlled/blind testing, so that variation from 2 kHz up (per the graph) is certainly within scope of differences we prefer to eliminate.

Fair enough, but I'm still curious what cables this graph was generated from. Even Amazon Basics CCA does a lot better than this at 3m. I know cables can matter in theory (this graph shows swapping cables could, in certain situations, be audible), but in real life in 2022 you have to go pretty damn far out of your way to find cables bad enough to produce this effect. For $0.20 per foot (roughly) you get -0.01dB at 20khz and -0.25dB insertion loss @ 3m.

Time to be really serious even though this thread was a joke: PSA: If you are currently using speaker wire that is significantly worse than 14ga CCA, please consider spending $3-5 to upgrade.

e: And full disclosure, I use cables made out of canare 4s11 because 1) it's nice to use cables that are the exact length you want and 2) the cable itself has a very nice feel and 3) overkill is comforting for irrational reasons, and 4) it's fun to make cables. But I very much doubt they perform any better than the 20 feet of random brown lamp cord I gave away when I sold some extra speakers on Facebook Marketplace a couple weeks ago.

 

[dartinbout]

No idea what you're referencing

People, who believe cables make a difference, live in the fantasy world of Peter Pan.

 

[julian_hughes]

People, who believe cables make a difference, live in the fantasy world of Peter Pan.

Try running your hi-fi with cables that are just an inch too short. Q.E.D.

I'm here all week.

 

[preload]

People, who believe cables make a difference, live in the fantasy world of Peter Pan.

I live in a world of science. Are you suggesting that 100ft of 22ga zip cord won't result in potentially audible differences in frequency response?

 

[SuicideSquid]

Less than half a decibel at 20khz per 10 feet isn't worth losing sleep over IMO. Also, not sure what the specs of these cables would be. Presumably something kinda crappy? These would be the worst-measuring cables I've seen on the internet, actually.

I, for one, find a random graph from 30 years ago with no context or additional information exceptionally compelling.

 

[julian_hughes]

I, for one, find a random graph from 30 years ago with no context or additional information exceptionally compelling.

Personally with cables I never settle for only one graph. After all, the left and right might be different.

 

[kemmler3D]

I live in a world of science. Are you suggesting that 100ft of 22ga zip cord won't result in potentially audible differences in frequency response?

Assuming the effect is additive / linear... probably it won't. 10 or even 100x a -0.01dB dip at 20khz is really hard for most people to hear.

However, I am still strongly in the camp that people need cables, and not no cables, to hear the sound.

 

[posvibes]

The signal jumping the synaptic gap may have the harmonics sheared off.

Aaahhh.

 

[preload]

I, for one, find a random graph from 30 years ago with no context or additional information exceptionally compelling.

You're welcome to read the paper and decide for yourself. JAES is a peer-review journal. Also, the date doesn't matter, since the last time I checked, our understanding of LCR hasn't changed. But I'm not going to convince anyone who isn't open to the idea that the output impedance of the amplifier, the speaker wire, and the loudspeaker form a circuit that can be easily modeled based on their LCR parameters. In fact, I know there are people on this board with the software and know-how to do it. It's not rocket science, and it's all verifiable.

 

[posvibes]

When does wire become cable? How many wires does a cable make? More than one apparently.

Ahh grasshopper, it is often said that one wire does not a cable make, but two wires are often crossed and so we speak false when we should speak Kippel.

 

[preload]

Assuming the effect is additive / linear... probably it won't. 10 or even 100x a -0.01dB dip at 20khz is really hard for most people to hear.

However, I am still strongly in the camp that people need cables, and not no cables, to hear the sound.

No, the effect is not a decrease in volume. The effect of series impedance combined with the output impedance of the amplifier and the reactive load of the loudspeaker/crossover (with a varying impedance vs frequency curve) will results in >1dB deviations in the FR. 200 ft (100ft x 2) of 22GA wire comes out to about 3.3 ohms of resistance.

Hopefully this illustrates how the speaker cable CAN make a difference. Now, if we're talking 12GA and 10ft, then the series resistance is <0.2 ohms, and the deviations are a lot smaller.

Here is the measured frequency response of a loudspeaker when adding in 1 ohm of series resistance (green) and 2 ohms of series resistance (black) from Roger Russell's site:

Speaker Wire

 

[SuicideSquid]

You're welcome to read the paper and decide for yourself. JAES is a peer-review journal. Also, the date doesn't matter, since the last time I checked, our understanding of LCR hasn't changed. But I'm not going to convince anyone who isn't open to the idea that the output impedance of the amplifier, the speaker wire, and the loudspeaker form a circuit that can be easily modeled based on their LCR parameters. In fact, I know there are people on this board with the software and know-how to do it. It's not rocket science, and it's all verifiable.

Of course it is, and of course it can. That's not the question. The question is "under normal use cases is this audible, or even measurable?"

If we're running power over 22 gauge cable or 12 gauge cable at 30 feet, we can probably measure a difference in the audible range, though it'll likely be small. Is that a normal use case? No. Are you going to find an audible difference between bog-standard 14 gauge cable, and $1,000-per-foot 12 gauge cable over a six foot run? Absolutely not. Do analog RCA cables or XLR cables or digital cables ever create an audible difference unless they're defective? Nope.

You don't need you wave your hands and shout "UM ACTUALLY!" when someone says "cables don't matter", because there's some esoteric, one-in-a-million scenario where cables actually matter. It's always implied in those statements that we're talking about normal distances and typical use cases.

 

[julian_hughes]

Let's all get real. You don't need graphs to measure cable. You only need a tape measure.

 

[kemmler3D]

No, the effect is not a decrease in volume. The effect of series impedance combined with the output impedance of the amplifier and the reactive load of the loudspeaker/crossover (with a varying impedance vs frequency curve) will results in >1dB deviations in the FR. 200 ft (100ft x 2) of 22GA wire comes out to about 3.3 ohms of resistance.

Hopefully this illustrates how the speaker cable CAN make a difference. Now, if we're talking 12GA and 10ft, then the series resistance is <0.2 ohms, and the deviations are a lot smaller.

Here is the measured frequency response of a loudspeaker when adding in 1 ohm of series resistance (green) and 2 ohms of series resistance (black) from Roger Russell's site:
View attachment 235046

Speaker Wire

Wouldn't this actually be more like the change in frequency response of an amplifier when played through a certain loudspeaker?

Anyway, if you are using 100ft runs of 22ga zipcord, I would say you have probably already made so many mistakes in your setup that cable is going to be the least of your worries.

 

[julian_hughes]

Wouldn't this actually be more like the change in frequency response of an amplifier when played through a certain loudspeaker?

Anyway, if you are using 100ft runs of 22ga zipcord, I would say you have probably already made so many mistakes in your setup that cable is going to be the least of your worries.

Double that. It'll all be bi-wired. Twice as good.

 

[preload]

Here's more from the Davis paper, this time showing the interaction of the power amp (which has an output impedance), various cables, and the loudspeaker. #7 is a standard 12GA zipcord. #10 is a braided Kimber Kable (14GA equiv). #2 is jumper cables for a car (7AWG). #12 is 18AWG lamp cord.
A 0.25-0.5dB difference with a bandwidth of 1 octave is potentially audible. That being said, the difference between the non-exotic cables is probably <0.1dB.