[krismusic]

I am pretty convinced that cables in an audio system simply have to be fit for purpose. However. It occurs to me that headphone cable may be an exception. As it provides a direct link between the source and transducer. Maybe the material used does affect the character of the sound. Is there any scientific truth in this idea?

 

[ab initio]

No. It takes hundreds of meters of cable before the effects from cables' electrical characteristics can become large enough to make a difference.

See http://en.m.wikipedia.org/wiki/RLC_circuit
You can estimate how long cables must be to put the corner frequency anywhere near the audio band

See cable specs for numbers, e.g., mogami lists electrical characteristics


Cheers

 

[cjl]

If the headphone cable's impedance at all audio frequencies is much, much less than the impedance of the headphones themselves (preferably by at least a couple orders of magnitude), then it will not make a difference to the sound. Fortunately, this means pretty much every cable ever made for headphones will not affect the sound (since they pretty much all will meet that requirement).
 
(On the other hand, if you're driving your BA IEMs with half a kilometer of 20AWG wire, or 50m of 30AWG, or even 5m of 40AWG, you will probably be able to hear a substantial difference in the sound)

 

[krismusic]

That was my gut feeling. Just that someone is claiming differences and I thought there might be some logic in the particular case of headphone cables.

 

[Steve Eddy]

No. It takes hundreds of meters of cable before the effects from cables' electrical characteristics can become large enough to make a difference.

Oh no, you can have audible effects with far less than hundreds of meters.

se

 

[HideousPride]

I have a couple aftermarket cables for various headphones and IEMs I purchased, but did so because the stock cable had one of the following problems:

1. Not the right length for my needs
2. Inflexible and/or prone to tangling
3. Frayed or damaged from use

I personally didn't hear any difference when it came to just changing out the cables regardless of the material used (bronze, silver, etc), but your enjoyment of the headphone can greatly increase if you have a cable that isn't causing you grief for one or more of the reasons listed above. And hey, if for some reason you do find the headphone seems to sound better in some way, all the better. Whether or not it's a placebo effect isn't much of a concern so long as you don't try to press your viewpoint on others.

 

[ab initio]

Oh no, you can have audible effects with far less than hundreds of meters.

se

I guess you could use 40 gauge or smaller and muck things up in shorter cables.

Got any links or citations on calculating the transfer function for cables from their impedance per length given the load impedance?

My quick RC and RL estimate for 22 awg cable on 100 ohm headphones says it takes 100 m or so before audio frequencies are approached in the effects. Maybe i oversimplified, but i dont expect an orders of magnitude error.

Either way, its the materials the op asked about so it would take a darn long cable before the difference between silver and copper would be audible.

Cheers

 

[superjawes]

You might be able to push audible effects into a very sensitive, low impedance IEM.

Otherwise, I did some resistance calculations yesterday just out of curiosity. Changing a 3 meter, 28 AWG, copper cable to a 3 meter, 24 AWG, silver cable yields a resistance difference of abpit 0.1 Ohms. I think you'll find much more than that in the tolerances of other components.

 

[SilverEars]

The cable is like adding a impedance adapter after the amp, but very very tiny impedance.  If it's like .1, there is not reason the impedance will affect the SQ.  I mean, you have iphones and DAPs with many times larger output impedance.

 

[Steve Eddy]

My quick RC and RL estimate for 22 awg cable on 100 ohm headphones says it takes 100 m or so before audio frequencies are approached in the effects. Maybe i oversimplified, but i dont expect an orders of magnitude error.

First, you said "hundreds of meters," second, the subject is "headphone cables," not interconnects. The vast majority of headphones are not orthos. So now run your numbers assuming a reactive load that does not have a flat impedance and calculate the frequency response errors.

Either way, its the materials the op asked about so it would take a darn long cable before the difference between silver and copper would be audible.

True enough.

se

 

[ab initio]

First, you said "hundreds of meters,"

You're right, I might have over exaggerated a bit!

 

 second, the subject is "headphone cables," not interconnects.

This is an important distinction to make. Interconnects are for voltage bridging  a relatively low impedance source (~ 75 Ohms?) to a high impedance input (~ 10 kOhms?)---here, the only real electrical aspect of any sane length of interconnect that one needs to worry about is noise rejection.
 
 

The vast majority of headphones are not orthos. So now run your numbers assuming a reactive load that does not have a flat impedance and calculate the frequency response errors.

I'm not really sure how this significantly changes the results. Are there headphones whose impedance changes by an order of magnitude over the audible range?
 
Dynamic headphones should be modeled as an inductor in series with a resistor, right? Ignoring electostatics, are there any drivers that exhibit significant capacitance?
 
A simple model for a headphone cable would be a series resistance and inductance with a parallel capacitance. Am I missing anything? Since we aren't worried about anything over 1 MHz, I'm going to ignore transmission line effects. Is there anything that I'm missing?
 
Cheers

 

[SilverEars]

  You're right, I might have over exaggerated a bit!

 
This is an important distinction to make. Interconnects are for voltage bridging  a relatively low impedance source (~ 75 Ohms?) to a high impedance input (~ 10 kOhms?)---here, the only real electrical aspect of any sane length of interconnect that one needs to worry about is noise rejection.
 
 
I'm not really sure how this significantly changes the results. Are there headphones whose impedance changes by an order of magnitude over the audible range?
 
Dynamic headphones should be modeled as an inductor in series with a resistor, right? Ignoring electostatics, are there any drivers that exhibit significant capacitance?
 
A simple model for a headphone cable would be a series resistance and inductance with a parallel capacitance. Am I missing anything? Since we aren't worried about anything over 1 MHz, I'm going to ignore transmission line effects. Is there anything that I'm missing?
 
Cheers

 

 

 

[ab initio]

   

Spoiler: Warning: Spoiler!

 

 

 
 
A factor of 2 isn't even close to an order of magnitude (i.e., factor of 10).
 
Cheers

 

[cjl]

   
 
A factor of 2 isn't even close to an order of magnitude (i.e., factor of 10).
 
Cheers

How about these:
 
http://www.innerfidelity.com/images/GradoGR8.pdf
http://www.innerfidelity.com/images/KlipschX10.pdf
http://www.innerfidelity.com/images/ShureSE535.pdf (OK, so this is only a factor of 4 or 5, but still...)
http://www.innerfidelity.com/images/UETF10.pdf
http://www.innerfidelity.com/images/WestoneUM1.pdf

 

[superjawes]

Isn't this a cable discussion? When it comes to driver impedance, having variable effects in the audible range is not only expected, but it can also be the goal.