Objectivist or Subjectivist? Give Me a Break

[DonH50]

@valkyrie: Capacitors have parasitic resistance and inductance, and also exhibit varying degrees of hysteresis (DA) and sensitivity to voltage, temperature, etc.

As Chris said, cables can be modeled as distributed RLCG networks, plus other terms to model dispersion, hysteresis and such, but by and large their parasitics are inconsequential at audio frequencies (unless they are very large, which is not the case for the vast majority).

"Settling time" depends upon cable characteristics plus the source and load.

I am not sure I followed the speaker cable description, it did not quite make sense to me...

A number of short threads in the ToC linked in my previous post (bottom of the previous page) have pretty pictures of how cables (mis)behave.

 

[j_j]

J_J,
However how would we measure this for a cable? The graph depicted is measured via FFT of the resultant sound from the speaker drivers (some kind of microphone/ADC/computer combination) - if we are going to do this for a cable - and hope to get the "cable" response I would think it would be swamped by the information from the speaker drivers and speaker crossover.

Capture the amplifier output vs. the speaker inputs. Assuming you do the math right, you can determine the impulse response due to the cable only, at least within your amplifier bandwidth.

 

[Orb]

There is something wrong with what's being reported here, because at least one ABX test has identified #22 speaker cables at a fairly short length.

And, of course, .1dB, if it's broadband, is getting up to the level of audibility. You will not hear "loudness" differences, rather they will appear as "quality" differences.

Typically matching is required to .2dB, but all you need is .1 up and .1 down to create an audible situation, albeit one that does not sound like a loudness change, but rather a quality change.

This raises an interesting point because many amps and CD players/DAC will have .1 +- difference somewhere over the 20hz-20khz frequency range, and usually at multiple points, this is compounded that source-output impedance/noise-distortion/filters/etc varies across the frequency range as well (apreciate this variation will also depend upon the type whether digital/amp/etc).
Matching exactly at 1khz should not be seen as arbitarily level matching loudness -just adding to the discussion.

That said I agree there must be level matching otherwise it does present what JJ says, but in reality any truly valid test would need to be in a completely enclosed/controlled environment; meaning bespoke scientific equipment to control and simulate all those variables.
Personally I feel any such test measurement also should be a more complex mathematical note rather than a single tone (more like a specific major chord that also has a specific-defined attack and decay) and analysed as a true envelope with time/amplitude/harmonics - but this is just me although I know some others in the audio world with scientific backgrounds have a similar thought (not naming and dragging them into said discussion as there is no satisfactory answer for all sides to the current discussion and so we have now 174pages).

I would like to see an ABX double blind controlled environment test using various music where two identical sources only differentiate between one having zero distortion and the other introduced 1% distortion maximum (low order and not skewed to high order) using different music each time and hit statistical 85%+ with 15 attempts; I appreciate some testing has been done to 4% (and this was not same environment) but curious just how well anyone will pass 1% in that specific ABX environment.
My money would be on zero across a group and not just a single listener(could be eating humble pie ), but technically it should be audible.
Anyway different point but valid to the objective listening.

And so expanding upon the last point but coming back to 0.1db being noticable.
JJ, has any DBT randomised ABX be done with two identical sources apart from one having an introduced 0.1db loudness and the listener identifying that source 85% ?
I appreciate it is more complex than this because we could be talking consistent 0.1db or peaks/dips over the FR of 20hz-20khz and the results may differ, anyway interested if this is a controlled ABX test you know about.
Cheers
Orb

 

[Robh3606]

Thanks for your patience - the light came on. For everybody, at least everybody who is interested, what J_J is referring to are the same "impulse" response measurements that John Atkinson takes in his speaker measurements for Stereophile reviews.

Hello Valkyrie

That's actually the step response which is derived from the impulse. Here is a graph showing both. The impulse is black and the step is red. With the step you can clearly see driver polarity and time offset.

Rob

 

[Groucho]

So is the verdict that a cable (a) improves the sound by imparting some character, or that maybe it 'corrects' the defective output of the source, or (b) that the cable is a necessary evil? In either case, what is it that distinguishes the $2000 cable from the $20 cable?

To me, the whole discussion is absurd, because it doesn't attempt to answer these questions. If we decided (a) then we could quantify what was good about it and duplicate it reliably rather than by the current lottery that supposedly depends on source impedance etc. (But could it really turn out that deviation from zero distortion/flat frequency & phase response was better? I don't believe it).

If it was (b) then we could suggest that everyone uses the shortest cables possible for an instant 'night-and-day' transformation of their system. Maybe a shorter $20 cable would sound as good as a longer $2000 cable etc. Maybe more careful design of the circuitry could eliminate the effect of the cable anyway e.g. lowering source impedance, applying theoretically optimal termination.

 

[Robh3606]

Just for fun here is what an inexpensive interconnect cable looks like measured on a CLIO test set-up. It is looped between the output and the input port. The cable is a no name you could purchase at Radio Shack for a couple of bucks. The first is the impulse response the second is frequency response and phase.

You get what you would expect.

Rob

 

[Groucho]

Great stuff. Do you have any multi-thousand dollar cables lying around to compare it with..?

 

[Robh3606]

Do you have any multi-thousand dollar cables lying around to compare it with..? .

Hello Groucho

No

The "best" I own are Straight Wire Symphony 2's. Hardly going to break the bank with those. It looks the same, very little difference at all measuring a 2 meter interconnect. Almost the same exact phase change above 1K even with different construction.

Rob

 

[j_j]

JJ, has any DBT randomised ABX be done with two identical sources apart from one having an introduced 0.1db loudness and the listener identifying that source 85% ?
I appreciate it is more complex than this because we could be talking consistent 0.1db or peaks/dips over the FR of 20hz-20khz and the results may differ, anyway interested if this is a controlled ABX test you know about.
Cheers
Orb

Not quite that simple, but a signal detection test has demonstrated audibility of that kind of level difference in a simple signal. This is not quite the question you asked, but it is a DBT cognate that is constructed specifically to be extremely sensitive and self-training. This was done as part of a level-roving experiment that also involved delay (or not) between the stimuli. This kind of level difference is ONLY detectable with clickless, instant switching. If you put 500ms between signals, you're at more like 2dB.

Hence the requirements on tests to be windowed switching, clickless, etc.

A scientific test has to be DBT or some cognate (i.e. computer administered SBT), but that's just the start. It also has to be well designed, use good training, provide feedback, have positive and negative controls, and a lot more things.

 

[Ethan Winer]

So is the verdict that a cable (a) improves the sound by imparting some character, or that maybe it 'corrects' the defective output of the source, or (b) that the cable is a necessary evil? In either case, what is it that distinguishes the $2000 cable from the $20 cable?

a) I hope not!

b) I hope not!

Other than "special" cables that really do aim to alter the sound, the difference between a $2,000 wire and a $20 wire is $1,980.

Maybe a shorter $20 cable would sound as good as a longer $2000 cable

Yes, if the capacitance of the expensive wire was substantially less than the cheaper wire, and capacitance was even a factor.

Maybe more careful design of the circuitry could eliminate the effect of the cable anyway e.g. lowering source impedance

Now you got it! Properly designed gear is capable of driving typical lengths of wire. Six feet seems long enough for most setups to get from any device to any other device. Then again, my subwoofer is about 12 feet away from my receiver, but even that isn't a problem. Especially since a response to 20 KHz is hardly needed. Though distortion might possibly increase with inadequate devices driving very long lengths. It's not only the output impedance, but also the amount of current the output can provide.

--Ethan

 

[j_j]

Properly designed gear is capable of driving typical lengths of wire. Six feet seems long enough for most setups to get from any device to any other device.

I've had occasion to use 20 foot RCA cables in some cases (yes, that's annoying, count on it). I did find that some consumer equipment was a bit shy of current drive and so I found some slope limiting (yes, with a scope, once there was something in evidence audibly).

You can argue that it shouldn't be so, and I'd agree.

 

[microstrip]

Just for fun here is what an inexpensive interconnect cable looks like measured on a CLIO test set-up. It is looped between the output and the input port. The cable is a no name you could purchase at Radio Shack for a couple of bucks. The first is the impulse response the second is frequency response and phase.

You get what you would expect.

Rob

And what else should we expect from this test?
Perhaps you can repeat it with a cable costing only a few cents.

 

[Robh3606]

And what else should we expect from this test?
Perhaps you can repeat it with a cable costing only a few cents

Hello Micro

That is what the test shows. That cable is as cheap as you can get

Rob

 

[Phelonious Ponk]

Not sure what you're saying, micro. If you're a believer in expensive cables, wouldn't you expect some signal loss or degradation from this test? If this is what you would expect from a Radio shack cable, what is it that cable believers expect from expensive ones?

Tim

 

[Ethan Winer]

You can argue that it shouldn't be so, and I'd agree.

Yes, that's what I'm arguing.

BTW, one free way to overcome an inadequate RCA output is to take the signal from a headphone output if that's available. It's possible that there will be slightly more noise and distortion than from an RCA output, but it's not automatic. Especially if the headphone amp is not loaded down by headphones.

--Ethan

 

[j_j]

Yes, that's what I'm arguing.

BTW, one free way to overcome an inadequate RCA output is to take the signal from a headphone output if that's available. It's possible that there will be slightly more noise and distortion than from an RCA output, but it's not automatic. Especially if the headphone amp is not loaded down by headphones.

--Ethan

Or cheap*** electrolytic dc blocking.

No, that should never happen.

That and 2.14 will get you a Venti drip.

 

[Atmasphere]

Yes, that's what I'm arguing.

BTW, one free way to overcome an inadequate RCA output is to take the signal from a headphone output if that's available. It's possible that there will be slightly more noise and distortion than from an RCA output, but it's not automatic. Especially if the headphone amp is not loaded down by headphones.

--Ethan

I find that a lot of headphone outputs are almost completely terrible. Noise! If the headphone amp involves an output transformer, the headphones or a similar resistance should be in place else the transformer will express the inter-winding capacitance rather than the turns ratio.

 

[Ethan Winer]

I find that a lot of headphone outputs are almost completely terrible. Noise!

Some headphones outputs suck, but many do not. Some are the exact same signal as the line output, wired in parallel with the RCA jacks! It depends entirely on the quality of the output amplifier.

--Ethan

 

[Atmasphere]

You mean 'preamp', right? Usually if there is an headphone output on something like a receiver or integrated amp, there is a 100 ohm resistor in series with the output, as it is derived from the loudspeaker outputs.

 

[cjfrbw]

a) I hope not!

b) I hope not!

Other than "special" cables that really do aim to alter the sound, the difference between a $2,000 wire and a $20 wire is $1,980.



Yes, if the capacitance of the expensive wire was substantially less than the cheaper wire, and capacitance was even a factor.



Now you got it! Properly designed gear is capable of driving typical lengths of wire. Six feet seems long enough for most setups to get from any device to any other device. Then again, my subwoofer is about 12 feet away from my receiver, but even that isn't a problem. Especially since a response to 20 KHz is hardly needed. Though distortion might possibly increase with inadequate devices driving very long lengths. It's not only the output impedance, but also the amount of current the output can provide.

--Ethan

I liked an article written by a prominent British sound architect, can't remember who, who builds surround sound recording studios in Britain and Europe. In typical dry manner, he said, "I don't object to expensive cables at all. I am sure most don't sound any worse than oxygen free copper."