box vs film caps (if that even makes sense)

[induction]

Drey wrote

6% Distortion at 10 Hz is a lot. With distortion circuits and fuzz boxes maybe thats not even that bad but with other clean effects like chorus boxes I think there's definitely a difference. Guitars usually don't go down to 10 hz but a low e is about 82 Hz.

What interests me is if you really can hear the difference. I need to build 2 chorus pedals or compressors with different caps and let my ears decide if I hear any major sound degradation

Don't forget that frequency is inherently logarithmic with pitch. 10 Hz is 3 octaves below the low E, and one full octave below the range of human hearing. The same test at 100 Hz (guitar amp outputs range from roughly 100 Hz to 10 kHz) showed only 0.25% harmonic distortion. This will be undetectable to humans in any guitar signal, whether it's fuzzed to death or crystal clean.

Two of the video's conclusions are pretty specious, in my opinion:
1. Capacitor composition definitely affects distortion.
(Fair enough.)
2. The type of distortion is discordant odd-order.
(It's only discordant if you can hear it.)
3. Ceramic capacitors are unsuitable for inter-stage coupling caps.
(Far from demonstrated. In fact, the tests shown in the video demonstrate that in the test circuit, the distortion measured in the audio range is too low in amplitude to hear.)
 

[Beaker]

Agreed. Those distortion figures are not bad at all in the frequency range we want them for.

I'm sure they are not ideal in all applications. However I personally am convinced they work well in most pedal designs.

I said they were quieter than standard ceramics - which implied that they were not as quiet as poly caps. However, they sound pretty quiet to me.

Same with the part about microphonics. I thought it was kind of common knowledge that caps can go microphonic - especially electrolytics.

I'm not saying that any of those facts are wrong or even misleading Drey, and it's always good to have all the info, but I really don't think that what we are using them for counts as "Critical Applications".

Pedal builders have a long tradition of using components that actively makes use of their flaws (transistor leakage for example) to their advantage.

Others deliberately mis-use components to get results, like Devi, DBA etc.

Thanks for contributing to the debate Drey.

[kirshman]

I have built 2 small clones from the layout on this site.  One with greenies and one with mlcc's.  To my ear they sound exactly the same, and the one with mlccs was definitely easier to build because of their size :-).  I know by ear is subjective, but mine tells me they worked great in this circuit.

[Frank_NH]

Re caps and composition:  What probably makes more of difference in sound between two identical builds is the fact that most caps have a 5% (J) or 10% (K) value tolerance.  This certainly can affect filters and blocking caps, though the impact is would likely be very hard to hear, in general, unless a cap was way off from its stated value.  I generally look for 5% box caps for my builds, but have used 10% with no problem.

I often measure my capacitors with my DVM and have noticed that pF range ceramic caps can vary quite a lot.  It may be that measuring small capacitance values is difficult for the average consumer-grade DVM...  

[Beaker]

Like Kirshman, I have re-built several pedals, including compressors and high gain pedals like Superfuzzes where originally I used poly film and/or poly box, and now use multilayer ceramics.

I have noticed three things:

#1. They are much easier to build wth mulilayers.

#2. They look nicer and tidier with multilayers.

#3 They do not sound any different - at least not that my ears are capable of noticing.

[ham_phist]

Hey all, this is a lot of good info. Can you clarify tho- mlcc is multi layer ceramic capacitor? Are these typically cheaper? I noticed that the greenies are cheaper than the gray boxes on tayda, so I went with those this time. But when I think ceramic, I'm thinking the tiny ones in pf.

I have other questions about how some of these work in circuits, but I need to gather them into some sort of intelligent and organized form before posting.

[nocentelli]

Drey wrote

6% Distortion at 10 Hz is a lot.

What interests me is if you really can hear the difference.

Unlikely. I've tried blind testing a simple clean boost with ceramic, polybox and electrolytic 100n caps for input and output and could not tell the difference. Maybe I have cloth ears. Maybe any effect is magnified if you used a more complex circuit with a string of "inferior" caps in series, but then multiple switching becomes problematic. If you build identical complex circuits with only the cap material as a variable and switched between them, you might notice a difference but you would have to be certain that ALL other components were identical in value/spec to draw a valid conclusion.

[IvIark]

kirshman wrote

I have built 2 small clones from the layout on this site.  One with greenies and one with mlcc's.  To my ear they sound exactly the same, and the one with mlccs was definitely easier to build because of their size :-).  I know by ear is subjective, but mine tells me they worked great in this circuit.

I have always thought that there are some subtle audible difference between SOME cap types, but in a clean circuit I've never noticed any difference unless the value varied enough to audibly change the frequency of a filter.  Any differences that I've ever experienced has been in dirt pedals where I think that character of the distortion can alter in some way.  Never anything night and day and always subtle, and really never in a way that I could pick one as better than the other, but different none the less.  In the real world I don't think the difference is anything worth bothering about and you could always get significantly more difference by changing the value or the gain of a transistor maybe.

I will still build things with some caps simply because they look cool, because it's nice when your build is visually appealing and there's no doubt some components do look cool, and that's fine by me as long as no one expects me to pay £15 for an "audiophile" electrolytic.  But I'd always use MLC first for the majority of builds because the results sound great, they're cheap as chips, I have gazillions, and they're small so easy to build with.  I do enjoy building with box caps too though, particularly in something with uniform 5mm pitch like PCB builds.

Thinking about it, I'm not too sure what real world form cap distortion actually takes in audio circuits, but maybe for the things we use them for it's not necessarily a negative thing anyway.  I'm quite fond of distortion  and it could be that the many non-linearities in caps can be a good thing in something inherently lofi like a guitar dirt pedal.  It's not something anyone should worry too much about though, just use what you can get cheaply and in quantity I reckon.