The 'Sound' of Capacitors..

[roseblood11]

Why, do the red ones have more haunting mids?

I have no idea. But I married her anyway

[deltafred]

I tired changing from both film to both ceramic (all within a couple percent of each other). Still don't hear a difference. Must be bad ears. Because I really want to hear the difference.

There was a post over on TalkBass (yea, yea, I admit it I'm a bass player at heart) where a luthier recorded some clips using his USA Standard Jazz bass then removed the neck, bridge, strings, pickups and controls and stuck them on a piece of timber (lumber - US) from his garage and recorded the same clips. He posted 3 clips and had a poll to identify them. the number who got it right was pretty low.

The TalkBass Wood Police got their panties in a real bunch over it citing all sorts of reasons why it was invalid. One guy even wanted to discount all the votes of people who couldn't tell the difference because they couldn't be bothered to guess.

[RnFR]

best post in this thread, yet.

[CodeMonk]

... I keep failing to see a way to demonstrate some palpable difference between cap types.

Yeah, I know how to use a breadboard, despite what you might think. What I can't figure out is how to switch 4 or 5 or 6 (or more) caps at one time. One I manually swap one cap in a circuit (especially a Muff), I can't hear a big difference. I can't even hear a medium difference with one swap.

So please, tell me, how do I swap them all at once on the breadboard? Surely, all the caps are important, or else you and many others wouldn't bother to use the caps you use. I just don't know how you can effectively, realistically test one dielectric type vs another using a breadboard. But man, I really do want to know how, because it sure would be a handy skill to have.

See RG Keens article on the CD4053.

[culturejam]

See RG Keens article on the CD4053.

Yeah, but then I'll have "IC noise" getting in there, and that'll skew the listening.

[CodeMonk]

See RG Keens article on the CD4053.

Yeah, but then I'll have "IC noise" getting in there, and that'll skew the listening.

True.
But if you did that for all the cap switching, wouldn't that make the noise not matter so much?
You get the same "noise" regardless of what caps are used.
I dunno, just a thought.

[Shiny_Beast]

mictester, you obviously have access to the internet. You should try downloading and reading a capacitor data sheet. The producers of capacitors don't agree with your "facts".

Two things:

1. As a professional electronics engineer I've read plenty of capacitor datasheets. The manufacturers DO agree with me, and their devices work just the same with sinewaves, squarewaves, triangular waves or any other waveshape or combination of waveshapes you can throw at them. The TYPE of signal is immaterial.

2. The basic laws of physics and electronics cannot be denied - even by the illustrious "Skreddy".

I really think some basic electronics theory is needed by the majority of these numbskulls....

Pretty much everybody disagrees with this point, except some electronic engineers looking at specs that probably weren't designed with passing an audio wave in mind.

I've read a bunch of technical posts/articles on the net that clearly imply how even the way a cap was wrapped at the factory may effect how it translates a sign wave. I don't know enough science to comment, but it seems more intuitive to me to think that all the specifics of a capacitor can effect it's tonal qualities than to beleive that all that matters is the capacitence rating.

You engineers see a couple specs, you want to be down to earth scientific people, so you dismiss anything that isn't on your sheet, basically take the high road and call everyone else an idiot cause they don't believe your spec sheet tells the whole picture. Meanwhile these are people that are using their ears, which is pretty material since we are talking about passing an audio signal not measuring dialectic whatever...

[diagrammatiks]

mictester, you obviously have access to the internet. You should try downloading and reading a capacitor data sheet. The producers of capacitors don't agree with your "facts".

Two things:

1. As a professional electronics engineer I've read plenty of capacitor datasheets. The manufacturers DO agree with me, and their devices work just the same with sinewaves, squarewaves, triangular waves or any other waveshape or combination of waveshapes you can throw at them. The TYPE of signal is immaterial.

2. The basic laws of physics and electronics cannot be denied - even by the illustrious "Skreddy".

I really think some basic electronics theory is needed by the majority of these numbskulls....

Pretty much everybody disagrees with this point, except some electronic engineers looking at specs that probably weren't designed with passing an audio wave in mind.

I've read a bunch of technical posts/articles on the net that clearly imply how even the way a cap was wrapped at the factory may effect how it translates a sign wave. I don't know enough science to comment, but it seems more intuitive to me to think that all the specifics of a capacitor can effect it's tonal qualities than to beleive that all that matters is the capacitence rating.

You engineers see a couple specs, you want to be down to earth scientific people, so you dismiss anything that isn't on your sheet, basically take the high road and call everyone else an idiot cause they don't believe your spec sheet tells the whole picture. Meanwhile these are people that are using their ears, which is pretty material since we are talking about passing an audio signal not measuring dialectic whatever...

hey man don't lump all engineers together with mictester.

i can't wait till mictester tells us about his hollow earth/polypropylene capacitors tool of bankers conspiracy theory.

since if all caps are the same...you'd only ever need one dielectric right?

[earthtonesaudio]

"If the listening test is repeatable, you're doing it wrong."
-Dr. Emmett "Doc" Brown

[MoonWatcher]

since if all caps are the same...you'd only ever need one dielectric right?

This is kind of the other pole to what I was thinking. If just one type of specific dielectric application was "audibly different in the right way," why would we use any other type? Why does it always end with, "just try it and you will find out?"

Have we all been just blindly using any cap, or is the difference negligible? Does it require anything scientific just to figure this out?

For the record, I've used every type of dielectric that I could fit into many different Big Muff builds, and any differences that I've heard I could not in good faith account to the capacitors. Even so, the differences were never better or worse per se, because they were so small. And this was with none of them spec'ed to make everything otherwise equal, they had different transistor types, etc.

It would be cool to hear a profound difference, especially one that would make me decide to go with one dielectric over others. If I'm not hearing something that is clearly there, I guess I should be thankful, since it makes cap selection easier - just grab the value that I need, and concentrate on other things.

I've come to the conclusion that a specific type of circuit would need to be created to over-emphasize the audibility. But that wouldn't be an existing pedal that anyone would use.

Lots of talk about why it should sound different. It never seems to get many steps beyond that.

[Greg]

Saying you can hear a difference when you know what type of cap is in there is worthless.. the human brain for notorious at fooling us into thinking we hear what we expect to hear (a difference.. an improvement).
It's merely opinion as opposed to fact.

Likewise, test equipment can show characteristics that cannot be heard, and things can be heard that cannot be tested.

If you can consistently pick the cap type in a true blind test when switching between types, (and the values are exactly the same), then you can hear the difference.
How many have actually done this though ?

[kleuck]

Something i just remembered : when i compared cables capacitances and sound, two cables with the same capacitance sounded slightly different in highs, one was a little "harsher" than the other.
On the french forum, other people compared the same cables, and got the same conclusion.
It was the Sommer Spirit vs Cordial 122, the Cordial being more "aggressive"
Not a huge difference, but a true and repeatable one.

[FiveseveN]

Did they know which one was "harsher" beforehand? Can you measure a change in impedance in the high-end?

[kleuck]

Did they know which one was "harsher" beforehand? Can you measure a change in impedance in the high-end?

They knew, as we were comparing on a dedicated thread.
Did not measure "impedance in the Hi-end" only capacacitance -and resistance, but just to be sure the cables were properly built.
Anyway, with a serie resistance of 0,000003 ohms in a 1M ohms amp...

For example -i did not have the Sommer Spirit for the first recording, but my preferred old cable (don't know the brand) gave similar results (after that i ordered Cordial 175 and Sommer Spirit, did live comparison, but no samples as i would have need to re-do with all the cables, some i gave to my sons)

Old Hi-End Cable total capacitance 490 pf

Cordial122 total capacitance 570 pf

The Cordial should be the less bright, the smoother : it's not the case.

[FiveseveN]

All I can hear is a slightly lower cutoff frequency, which is exactly what one would expect: http://buildyourguitar.com/resources/lemme/ (see the part about varying capacitance--Figure 8 ).

I dunno, maybe something got lost in translation. Describing a subjective sensory experience in words is hard enough, I guess it's even harder in a foreign language.

[kleuck]

I know this link : i post it very often (and this one) to explain what is really the interaction between PU & Cable ; even here for the sake's of our preferred "professional electronics engineer"
But don't focus on the average bandpass, only on the highs, if your PC is connected to a good Hi-Fi amp, you will hear that the Cordial sounds brighter and harsher, which is theoretically not to expect with a higher capacitance (BTW, the Cordial has a very low capacitance, but mine is a little longer than my old cable : 5m30 for the old cable, and 6m15 for the Cordial).
Not a huge difference, but the kind you can expect from two different caps technologies, to me, here, it's almost like polystyrene cap vs ceramic when used in a Treble Booster (filter of course)
Actually, i'm SURE the best effect to really compare caps IS a Treble booster.
Must say that i did not change anything between recording, nor did i turn off the amp : just swap cables.
These samples are not the only tests i made of course, i just had to make some, but i compared a lot before.

I dunno, maybe something got lost in translation. Describing a subjective sensory experience in words is hard enough, I guess it's even harder in a foreign language.

Ho yes, very hard when it comes to subtle things, and we do are talking subtle things.

[Nocentelli]

If you can consistently pick the cap type in a true blind test when switching between types, (and the values are exactly the same), then you can hear the difference. How many have actually done this though ?

Let's do this!

Set-up a simple fuzz e.g. BMP with agreed values and a fixed gain and tonestack, play a 10 second riff via a looper and swap out a single coupling cap between 100n poly, ceramic and 0.1u electro. Record five passes of each, then chop the file up to scramble the order of the loops, and we'll vote on which out of the 15 are which cap.

[kleuck]

In pedals, i did not really hear any difference (except for the TBs) -mine have always a minimum numbers of gain stages (2 max at the moment)- but i'm pretty sure that a BM with only polystyrene caps in the feedback loops will sound very differently from à BM with only ceramic caps in the NFBs.

Saying you can hear a difference when you know what type of cap is in there is worthless.. the human brain for notorious at fooling us into thinking we hear what we expect to hear (a difference.. an improvement).
It's merely opinion as opposed to fact.
Likewise, test equipment can show characteristics that cannot be heard, and things can be heard that cannot be tested.

Not really true, because test equipment can only show you what you are seeking -among the things it can measure, and real world equipment is not always able to measure a lot of things.
Ears (and brain of course) on the contrary do not have to measure one and only thing at a time ; and more important, they ARE the final "customers" of our sound chains, so, the final judges after all.
When i play, it's for my ears, not for my DMM.

[deltafred]

since if all caps are the same...you'd only ever need one dielectric right?

Right - but only if caps were used for just audio, then we would need a very limited number of types.

Audio is almost a spin off as it covers the very lowest bit of the frequency spectrum of most caps. Most of their differing properties - able to operate at high frequencies, high voltages/current/AC, self healing, etc are not utilised.

From when I was at college I cannot think of a single cap type that was quoted as 'good for audio', most were aimed at radio frequency applications or AC phase shift and power factor correction.

Back then mojo hadn't been invented and snake oil was just a gleam in a young marketing man's eye so manufacturers data was all we had to go on for selecting dielectric types.

[Diamond]

Okay, just did a very unscientific test with some caps.

I built up a Bazz Fuss on my breadboard. MPSA18, 1N914, 100n film output cap (measured 102.2 nF), 100K resistor off the collector (measured 99.7K), emitter grounded.

I used 5 different caps for the 100n input cap and measured them with my Atlas LCR. All of them were rated for 100v or less. Here's what I tested:

• FILM - 101.81 nF (cheap-ass green chicklet)
• FILM - 100.64 nF (pricier Panasonic ECQ)
• CERAMIC - 101.00 (multilayer ceramic, AVX)
• FILM - 101.4 nF (WIMA box cap)
• CERAMIC - 99.86 (cheap-ass disc cap)

Played the same basic cheesy 80s metal lick a few times with each cap. Went through all the caps 3-4 times.

The results? Well, I can't hear a difference at all between these caps, at least in the function of the input cap of a simple Bazz circuit. And I REALLY wanted to. Badly. It would be so much more awesome for me if I could hear a difference.

I will be trying this same test with other circuits to satisfy my yearning to actually hear some difference that is meaningful.


I did notice that the cheap-ass ceramics were mostly all way under the rated value. The first couple I tested were around 70nF! None of them were over 100nF, and I had to test about 50 of them to find one that was within 1% of the rated value. The film caps were mostly over value by a little. So for my money, I'll probably be getting multi-layer ceramics, as they seem to be a LOT more consistent and they are much smaller.

Fun stuff!

(and by the way, I am continually amazed at how awesome the Bazz Fuss sounds for so few parts)

Great post!