Audible difference in high-end capacitors? - ABX samples

[lashto]

Sorry guys, you are mostly chasing ghosts here, with your suggestions. Do some serious work, measurements, discuss results. There might be something to discuss, but not based purely on second guessing.

View attachment 319210 View attachment 319212 View attachment 319216 View attachment 319218 View attachment 319219

+1

And measuring the FR alone is probably not going to show much, if anything. Almost every amp/dac is FR-linear nowadays and they still sound different (sometimes).

For example, the HD measurements above do show a clear difference. Audible or not it's hard to say, but at least it's visible and measurable. Can you also measure distortion @ctrl ?

And maybe some of the caps can add a sort of compression-like effect which is somehow measurable?!

 

[voodooless]

Can you expand on that or give a source that explains it in layman’s terms?

I always thought capacitors were like batteries - they take energy in, store it, and then release it, they just can’t release slowly over time like batteries can.

I thought microphones work like speakers in reverse - instead of converting electrical energy into the motion of a diaphragm to make sound, they convert motion of a diaphragm excited by external sound waves and convert it into electrical energy.

See https://en.wikipedia.org/wiki/Electret_microphone

Basically what happens in these caps, is that the vibration of sound causes the material to flex and contract, creating larger and smaller capacitance. But it can also work the other way around, depending on the capacitor construction.

 

[fpitas]

Also, maybe some of the caps can add a sort of compression-like effect which is somehow measurable?!

I doubt that. Moderm polypropylene DF is almost unmeasurably low. But if you're worried, use pulse-rated industrial caps with screw terminals, not bizarre "audiophile" crap.

 

[wwenze]

Can you expand on that or give a source that explains it in layman’s terms?

I always thought capacitors were like batteries - they take energy in, store it, and then release it, they just can’t release slowly over time like batteries can.

I thought microphones work like speakers in reverse - instead of converting electrical energy into the motion of a diaphragm to make sound, they convert motion of a diaphragm excited by external sound waves and convert it into electrical energy.

Capacitors are two plates that are very close to each other

Two plates that contain opposite charges will attract

(Oh yea and when they do that their capacitance value also increase)

 

[lashto]

I doubt that. Moderm polypropylene DF is almost unmeasurably low. But if you're worried, use pulse-rated industrial caps with screw terminals, not bizarre "audiophile" crap.

I can barely solder a pair of cables so I am not worried there . Just curious.

I am however a bit worried that this thread is going like "we did one measurement, found absolutely no difference and therefore all those caps are/sound the same". That may very well be the case, but if you want real proof, go on and measure everything that may be relevant. And even measure some things that may not seem relevant.

Yes, that would be a lot more effort, but if it's worth doing... otherwise it's just another wasted measurement/thread.

 

[MAB]

Can you expand on that or give a source that explains it in layman’s terms?

I always thought capacitors were like batteries - they take energy in, store it, and then release it, they just can’t release slowly over time like batteries can.

I thought microphones work like speakers in reverse - instead of converting electrical energy into the motion of a diaphragm to make sound, they convert motion of a diaphragm excited by external sound waves and convert it into electrical energy.

Capacitors are two metal plates, separated by a dielectric (insulator). The capacitance C is proportional to the area A of the metal plates divided by the separation d (thickness of the dielectric that separates them.
C = εA/d
If the plates vibrate so that d varies, then the capacitance varies!
The other thing about capacitors is they store charge Q proportional to the applied voltage V.
Q = CV
So if the capacitance varies but no charge has been added or removed from the capacitor, a voltage will appear on the capacitance terminals equivalent to the deformation of the dielectric just like a microphone!

I made a capacitor out of two pieces of foil and a layer of backing parchment as insulator, wrapped around a wooden dowel for stability. It was tounge in cheek, but I actually am trying to illustrate how darn simple these are, and how little mystery there is. I wrapped my hack-a-cap in tape to keep it reasonably stable, but for sure it is probably a bit microphonic. Funny, the consistency of how the films are stacked is of large importance in the microphonic behavior, but all we hear about are the supposedly esoteric materials (copper foil, is that really esoteric???). That being said, I doubt that capacitor microphonics are a significant source of distortion.

 

[fpitas]

I can barely solder a pair of cables so I am not worried there . Just curious.

I am however a bit worried that this thread is going like "we did one measurement, found absolutely no difference and therefore all those caps are the same". That may very well be the case, but if you want real proof, go on and measure everything that may be relevant. And even measure some things that may not seem relevant.

Yes, that would be a lot more effort, but if it's worth doing... otherwise it's just another wasted measurement/thread.

Fair enough. Yet I have never once seen people test industrial grade polypropylene for this use, even though, 1. They are fully documented in datasheets and tested to meet specifications, and 2. They are much more likely to be well-made than audiophile stuff.

 

[fpitas]

Fair enough. Yet I have never once seen people test industrial grade polypropylene for this use, even though, 1. They are fully documented in datasheets and tested to meet specifications, and 2. They are much more likely to be well-made than audiophile stuff.

Oh and, 3. The price is very reasonable.

 

[pkane]

Yes, that would be a lot more effort, but if it's worth doing... otherwise it's just another wasted measurement/thread.

Why go to the trouble? Because some audiophiles believe in magic capacitors that sound better?

 

[Psycher1]

Possible to do some kind of "waterfall" test? I'd be curious to see how capacitors holding on to energy might be relevant. I believe the term is Capacitor Discharge Time Constant, which describes how different capacitors hold energy for differing periods of time when asked to discharge. I understand it can be whole seconds sometimes. This seems to me to not affect the frequency response, more, a time lagging effect.

Here, I feel I know just enough to be dangerous - this probably has no bearing at all, but I'd be curious to see it explored either way.

 

[uwotm8]

As attachments you can find three files starting with file names like "01-Samples_FC... .zip". These are the files with the "Fast Car" track sample recordings, one file for each capacitor.
In the "samples_PN.zip" file are the three pink noise recordings for the three capacitors.

Ok let me be an audiofool here. This is almost unlistenable - the first and strongest impression is music played by tweeter itself and I can't evaluate such samples properly, just can't get rid of "omg that's just totally wrong sound".

To properly compare these samples, they have to be put into editor and then goes "track A minus track B" substraction and then we'll hear what remains.

As mentioned above, I'm too for "Mythbuster-like" method - trying to find that extremely bad situation when it works. In our case, there can both old EL caps, thin wires and ferrite-core coils be used at the same time. Would be nice to do that to complete speaker, not just tweeter.

 

[voodooless]

This is almost unlistenable - the first and strongest impression is music played by tweeter itself and I can't evaluate such samples properly, just can't get rid of "omg that's just totally wrong sound".

That is exactly what it is:

As promised, this time there are short audio samples of each capacitor (in a simple crossover) recorded with a microphone via tweeter (test setup see below).

 

[uwotm8]

That is exactly what it is:

Yes, I know. But that doesn't make it work (for me, maybe works for somebody else).

 

[thewas]

I simply did not expect how loud some capacitors played the music.

Opens a new high end audio USP marketing chance, loudspeakers made only from audiophile caps, the disturbing and distorting rest left out.

 

[BDWoody]

Opens a new high end audio USP marketing chance, loudspeakers made only from audiophile caps, the disturbing and distorting rest left out.

Maybe a line of IEMs...or IECs?

 

[pkane]

To properly compare these samples, they have to be put into editor and then goes "track A minus track B" substraction and then we'll hear what remains.

Sure, here's the difference file: https://app.box.com/s/xcgtrzqj1jgh2z8z76i48h0bue1ssnl5

Listen at your normal listening levels -- let us know what you hear. This is high-pass filtered at 1000Hz.

 

[Spkrdctr]

I would suspect that if you sent the test setup to Danny he could have made it sound amazing. He has the masters touch.....

 

[phoenixdogfan]

5) Audio recording samples

As already said, the recordings were not made in a studio, so I can not exclude that the recordings are possibly distinguishable by background noise (street noise, children's cough - daughter with cold, TV sound,...) or editing mistakes - if you notice something like that, please report right here in the thread, so that I can possibly exchange the files.

During the recordings the microphone was about 2cm (0.8'') away from the tweeter. The maximum sound pressure at the mic was about 106-123dB. The used mic, an iSEMcon EMX-7150, shows >143dB 3% THD and around 125dB 0.2% THD.
View attachment 319177 View attachment 319180 Source: Production Partner 6/2012 review


During the ABX test in foobar I explicitly did not look for possible errors in the recordings, but tried to hear a difference between the capacitors in an "honest way".

Of course, I myself was extremely curious about the results. In particular, how the ancient film capacitors that I myself had bought ten years ago as bagged goods (at that time you could buy a bag of various bipolar film capacitors for a few Euros), would perform.
We expect, after the reviews of the high-end capacitor quoted in the opening post and the strongly presented statements and opinions of individuals in forums and YouTube videos that the difference will be staggering.

Therefore, in an ABX blind test, all ten of ten (10/10) trials should be correctly identified.
In other ABX tests, for example, I had 8/10 correct, but this turned out to be a coincidence after further attempts. Therefore, in order not to commit self-deception it should be 10/10 or two times in a row 9/10 correct answers.

All recordings are in lossless flac audio format.
As attachments you can find three files starting with file names like "01-Samples_FC... .zip". These are the files with the "Fast Car" track sample recordings, one file for each capacitor.
In the "samples_PN.zip" file are the three pink noise recordings for the three capacitors.

My own ABX results look like this:
Alumen-Z-Cap versus Old-Cap for pink noise recording
View attachment 319175

Alumen-Z-Cap versus Standard-Cap for "Fast Car" recording
View attachment 319176

As one can easily see, the different film capacitors were completely transparent and indistinguishable to me. So for me there is no reason to spend fifty times or more the price for high-end capacitors, compared to normal standard film capacitors.

You maybe just put Danny out of business.

 

[thewas]

You maybe just put Danny out of business.

Unfortunately people like him and most of his fans don't care about facts, you can bomb them with clear evidence and facts and they will just ignore them, but such work is good for third readers who are still open to learning.

 

[antcollinet]

I can barely solder a pair of cables so I am not worried there . Just curious.

I am however a bit worried that this thread is going like "we did one measurement, found absolutely no difference and therefore all those caps are/sound the same". That may very well be the case, but if you want real proof, go on and measure everything that may be relevant. And even measure some things that may not seem relevant.

Yes, that would be a lot more effort, but if it's worth doing... otherwise it's just another wasted measurement/thread.

Back to the audiophiles asking science to prove a negative "test everything or you don't know"

The burden of proof is other way around. If someone believes that an expensive cap makes a difference, they only have to test that one cap to prove the difference really exists.

Yet, amazingly, no-one ever does it.