The 'Sound' of Capacitors..

[culturejam]

In low voltage stompboxes maybe you can`t hear the difference.

And this is what it's all about for me. If the differences are not particularly audible in stopmbox circuits, then it becomes academic for me. It's still interesting, but not something I'll worry about.


But I'm curious about this "cumulative" effect some of you lads have mentioned. Does this accumulation of tonal difference also apply to a circuit with only a few caps (like a Bazz Fuss or even Fuzz Face)?? Is it more pronounced with more simplistic circuits?

[mictester]

As well,

Just for the fun I looked up datasheets for WIMA capacitors, just as an example.
If there's no difference..then why am I looking at characteristics that display capacitance vs.... Frequency. And impulseresponse? Apparently something is happening there.

Yes, Dirk. Look carefully at the frequencies they're talking about. We're talking AUDIO frequencies here. Capacitor types really matter at RF (I know - I often design RF circuits professionally), but there's NO difference at audio frequencies.

Settle it for yourself:

Build a (say) tubescreamer twice. Once with generic polyester capacitors, and then again with other types of capacitors measured to have the same values. Next - enlist the help of a trustworthy friend to perform a proper "double-blind" test. You will NOT be able to hear the difference. Really!

[mictester]

I've found it to be a cumulative thing, as well, very generally speaking.

I've also found a few instances where putting a polarized cap in certain high gain portions of circuits will produce some really nasty results - clearly audible stuff - D.C. blocking was not effectively occurring. I've had similar experiences on just a few occasions with ceramics in some of the same locations. I wanted to believe that it was just my imagination, but when I hot swapped for another dielectric type (with capacitance being equal), I could confirm that it was an actual audible difference.

But for the most part, it's become a nonissue for me, as I tend to mix up whatever I have on hand, anyways.

I've told most folks that I prefer box films for the main reason that they are compact and make population very easy. I also like monolithics for tight high count layouts.

Perhaps the most fun of all for me throughout the years has been the "magic Muff capacitor experiment," where I've built the same exact Muff circuit and only swapped out the couplers for another dielectric or even just a different brand. So many folks claim that there's "preferable" types of caps that I just had to try them all for myself over the years. My conclusion is that there are zero audible differences between the caps in the Muff circuit (big surprise), unless there are some monster tolerance differences at key points in the circuit, namely the couplers before and after the Sustain pot, the pair that set the shelf for the clipping diodes, and the tone stack caps, mainly the treble cap. But none were deal breakers, or produced a "superior" tone.

I really like the saying that something is more than the sum of it's parts. It holds true in the case of pedals, at least to me. I'm more concerned with attention to circuit design than component composition most of the time.

Thank you! Some real sanity at last! The effects you're seeing in high gain or high impedance circuits (the lack of DC blocking, for example) is either the capacitor leaking DC (which will particularly show up at high impedances) or the effect of the capacitor charging up over time! One Boutique box I saw used 100µF capacitors for inter-stage coupling. The impedances were so high in that part of the circuit that there was a charging time from switch-on of a couple of hours! This was from a well-known solder-jockey (no real electronic knowledge) who'd obviously heard that he could get more bass through the circuit with larger value capacitors...

The microphonic characteristics of some capacitors are well known, and wouldn't be chosen for audio circuits for that reason (unless you're a clueless Boutiquer), but modern capacitors are very well manufactured and suffer much less from the ill effects that you describe.

[mictester]

- D.C. blocking was not effectively occurring. I've had similar experiences on just a few occasions with ceramics in some of the same locations. I wanted to believe that it was just my imagination, but when I hot swapped for another dielectric type (with capacitance being equal), I could confirm that it was an actual audible difference.

AFAIK a cap will always block DC totally unless the dielectric has broken down... there's no path for it to pass.

Electrolytics (in particular) can leak quite a lot - particularly large value ones.

[mictester]

I can definitely here a difference between 1uF mini box caps and 1uF 16 volt electro's either polarized or NP,(I think everyone can) in big muffs I've made and modded. I can also definitely here a difference between 100nF ceramic and 100nF mini box types. They are the most noticeable to me. In the pF range I cant tell (as far as I've ever noticed) a difference really between ceramic and brownies/greenies etc.

I don't like a lot of ceramic caps in big muffs. Just doesn't sound right if its all ceramic.

I haven't made a real big effort listening to cap types outside of the big muff and a fuzz of mine to see which sounds best to me so that's about all I can contribute on this subject.

I'm pretty sure you're just fooling yourself!

[mictester]

Sorry, Dirk, but NO!! It doesn't matter what the capacitor is made of - at audio frequencies, and inside the dielectric breakdown voltage, any two "100nF" capacitors will both measure the same and sound the same! The only percieved difference will be in the way they look!

Sorry but I disagree with you too. There must be a few of us here imagining things.

There definitely are! If the values are the same, and you're not approaching the working voltage of the device, there will be no difference whatsoever.

Spurious pseudo-science is for the Audiophools - we're more technical than that.

[bumblebee]

I can definitely here a difference between 1uF mini box caps and 1uF 16 volt electro's either polarized or NP,(I think everyone can) in big muffs I've made and modded. I can also definitely here a difference between 100nF ceramic and 100nF mini box types. They are the most noticeable to me. In the pF range I cant tell (as far as I've ever noticed) a difference really between ceramic and brownies/greenies etc.

I don't like a lot of ceramic caps in big muffs. Just doesn't sound right if its all ceramic.

I haven't made a real big effort listening to cap types outside of the big muff and a fuzz of mine to see which sounds best to me so that's about all I can contribute on this subject.

I'm pretty sure you're just fooling yourself!

I'm not pretty sure, I'm 100% sure I know what I'm talking about.
If you cant hear it that doesn't mean it doesn't exist.
Just like science cant explain how the universe came into being (the big bang is bullshit. everything from nothing....sure!) but that doesn't mean "god did it".

Am I to say 8 1uF electros sounds identical to 8 1UF box caps in the SAME pedal when it doesnt?

[Greg]

- D.C. blocking was not effectively occurring. I've had similar experiences on just a few occasions with ceramics in some of the same locations. I wanted to believe that it was just my imagination, but when I hot swapped for another dielectric type (with capacitance being equal), I could confirm that it was an actual audible difference.

AFAIK a cap will always block DC totally unless the dielectric has broken down... there's no path for it to pass.

Electrolytics (in particular) can leak quite a lot - particularly large value ones.

I understand there can be a leakage through the dielectric over time.. but enough to not block DC effectively in a circuit ?

[IvIark]

Spurious pseudo-science is for the Audiophools - we're more technical than that.

It's nothing to do with technicalities, I don't measure what I hear with a calculator. You're speaking as though the people who disagree with you have never tested anything like this for themselves, but somehow I doubt that's likely to be the case. How will ESR affect the final tone, or even the physical properties of the constituent materials? I just don't believe it's as cut and dry as you suggest.

One thing I do agree with you about is that I personally couldn't do a double blind test with cheapo caps and ultra high end and pick out the high end ones, but that's not because I don't believe there can be any audible differences, I just don't think the ultra high end cap will necessarily always sound the best to me. I've heard subtle differences but not been able to say that I specifically prefer one over another, but that didn't mean they sounded exactly the same, just that I had no preference. Maybe that's close enough for some people and they don't feel they have to bother, but some of us like the experimentation and choosing components that we feel give the best results.

[Greg]

Settle it for yourself:

Build a (say) tubescreamer twice. Once with generic polyester capacitors, and then again with other types of capacitors measured to have the same values. Next - enlist the help of a trustworthy friend to perform a proper "double-blind" test. You will NOT be able to hear the difference. Really!

This really is the best way to test the theory I guess.. but a lot of trouble seeing you would need to try a number of different cap types... and of course you'd need to also match the resistors, pots, etc.
The differences, if any, would be subtle - and it's not scientific because it still depends on the hearing capabilities of the testers which can vary enormously, but it's much better than testing the same pedal after changes ... even the same pedal can sound good one day and not so good another... not because it changes, but because we do.

There's been some great responses.. this is always a bit of a controversial subject, because you can't tell someone else that what they're hearing isn't real... nor can you totally prove it.

Like culturejam, I'm mainly interested in the effect of caps in pedal circuits rather than higher voltage circuits at this time.

One thing I do agree with you about is that I personally couldn't do a double blind test with cheapo caps and ultra high end and pick out the high end ones, but that's not because I don't believe there can be any audible differences, I just don't think the ultra high end cap will necessarily always sound the best to me. I've heard subtle differences but not been able to say that I specifically prefer one over another, but that didn't mean they sounded exactly the same, just that I had no preference. Maybe that's close enough for some people and they don't feel they have to bother, but some of us like the experimentation and choosing components that we feel give the best results.

For the purpose of a test like this, your personal preferences wouldn't matter.. just whether you could consistently pick one from another.

[jessej]

I can hear the difference, it is quite obvious. Claiming otherwise is just pure ignorance.

Here's a quick list of some factors to consider why 100nF measures the same, but sounds completely different.

http://en.wikipedia.org/wiki/Types_of_c ... _tan-delta
http://en.wikipedia.org/wiki/Hysteresis
http://en.wikipedia.org/wiki/Ferroelectric
http://en.wikipedia.org/wiki/Electrical_dipole_moment
http://en.wikipedia.org/wiki/Slew_rate

I'm sure someone who has an degree or education can tell you more, this is just stuff I read up on myself during the short time Ive been into electronics (hobby), not ever studying maths etc...

You can measure your 32" TV and it will be 32", but your resolution can vary, so can your contrast and luminosity and brilliance and saturation. Or are you claiming all 32" TVs have the same picture quality because they are all 32" TVs when you measured them???

Just measure up an electrolytic and a ceramic cap with "the same value", socket your board and test. If you can't hear the difference in transients and overall oomph throughout the dynamic range (soft to loud), then there's possibly something wrong with your hearing (or your amp/speakers??).

[IvIark]

For the purpose of a test like this, your personal preferences wouldn't matter.. just whether you could consistently pick one from another.

Wouldn't matter to whom? The test is just that, a test. My choice of what caps I use is exactly down to my personal preference at any given time with any given circuit and nothing else.

[bumblebee]

I own 12 big muffs, I have modified a great deal of them,. In particular the Sovtek ones. I know I replaced all 100nF caps with 1uF electros (NP) and hated the sound of it(not because of the value as I put 1uF's in ALL my muffs), I swapped them all for 1uF box caps and loved the sound of it. There was a huge and very obvious difference in sound in a total of 5 sovtek Big Muffs I have done this to. In each case I tried electro's first as I always seem to have more of them on hand than box caps.

I haven't got a fucking clue why it sounds different but I know it does in 5 examples I've experimented with. I'm like that, I don't know why things do what they do in big muffs (well I do know some of the "whys" about muffs as I'm very familiar with them) (or any pedal really), I just know what to change and to what values to use to get a specific result in various pedals, the why has never really bothered me....until recently, I have become more interested in the "why" part of it.

However, just because I cant explain why this or that happens, doesn't mean it doesn't happen, all that means is I lack the technical knowledge to explain why it happens.
This is why I will eventually get my EE degree!

[IvIark]

This really is the best way to test the theory I guess.. but a lot of trouble seeing you would need to try a number of different cap types... and of course you'd need to also match the resistors, pots, etc.

That could be a pain which is why I suggested something low part count like a booster where you could use a rotary switch to swap out the input and output caps. It'd be easier than exactly part matching 30 odd components. With a 3 position 4 pole it'd give you 3 matched pairs of caps to test against each other.

[Greg]

For the purpose of a test like this, your personal preferences wouldn't matter.. just whether you could consistently pick one from another.

Wouldn't matter to whom? The test is just that, a test. My choice of what caps I use is exactly down to my personal preference at any given time with any given circuit and nothing else.

Of course.. I was referring to the test that was proposed to see if a person could detect one cap type from another.

You said you couldn't do the test because of personal preferences.
The test doesn't affect, (and shouldn't be affected by), what you like to build with or what you like to hear.
As soon as you bring those things in, the test isn't effective.

[MoonWatcher]

I can also definitely here a difference between 100nF ceramic and 100nF mini box types. They are the most noticeable to me.

I could not, if I made sure that they all spec'ed to within 5% of that value. It wasn't until the tolerances got out of whack in critical parts of the circuit (before and after Sustain pot, for example) before it was audibly noticeable. I'm talking about one cap being around 85nF and another being 140nF. And if I matched any two different dielectrics within these different off target points, they again sounded the same.

The only part of the Muff circuit that I think is absolutely debatable in regards to caps would be the pair tied into the clipping diodes, since they set the clipping shelf. I seem to be a bit overly sensitive to these - undistorted lows versus clipped highs can sound rather detached to me personally, especially with the small 47nF triangle spec caps. Different dielectrics seem to behave differently, no doubt because of the relationship between the diodes and the dielectric.

[bumblebee]

Yeah 47nF is waaaay to small a value here, a minimum should be 220nF and 1uF is ideal IMO. 47nF results in boomyness, often mistaken for bassyness. 220nF is ok-ish but you can sometimes still hear a little bit of that boomy sound coming thry playing chords on the low string if tuned in anything below F, like E to A etc.

[oldgravity]

Am I to say 8 1uF electros sounds identical to 8 1UF box caps in the SAME pedal when it doesnt?

Again it sounds to me like you're just going by the nominal value. Did you actually measure to make sure these 8 "1uf" caps were the same precise value as the other "1uf" caps? If not, it's pretty obvious why they sound different. Electrolytics commonly have a tolerance of +or- 20%, which means that your "1uf" box cap could be as much as 30% difference in capacitance from a 1uf electro. Enough that the corner frequency could be 2/3 of an octave different. Do that 8 times throughout a circuit, and it's not surprising that it sounds different.

There is an excellent reason to use box caps instead of electros, which is that they won't break down in 10 years and need to be replaced.

Just like science cant explain how the universe came into being (the big bang is bullshit. everything from nothing....sure!)

Aha... these comments are helpful because they help me understand where you're coming from, in terms of epistemology, and your personal philosophy of "knowledge". It's kind of like I don't need to argue about ontogeny recapitulation with somebody who thinks the Earth is 6000 years old (not saying that you believe that). Obviously there are much deeper disconnects in our thinking than with regard to capacitors.

[mictester]

Am I to say 8 1uF electros sounds identical to 8 1UF box caps in the SAME pedal when it doesnt?

They DO - as long as the capacitor values are the same.

[IvIark]

You said you couldn't do the test because of personal preferences.

I didn't say that at all, I'm more than happy to do a test. In fact I'm ordering some 100nF electrolytics for that very reason. You quoted me from a post I wrote to mictester which was about my personal preference and general feelings on the matter. They aren't going to change because they're based on my own experiences.