Paper in oil caps

[midwayfair]

I've thought of a test, now.

Find two identical mojo and non-mojo parts, as per MFD, but not as per mfg or possibly even not as same material. The capacitances are identical. Insert a switch into the guitar to select for treble bleed of either one or the other. Turn the guitar as low as possible in volume, and turn any intermediary pedals, the gain, and the volume of the amp up to 10. Record that. See if there's a noticeable audible difference, and get some scope data on it. To strum the guitar strings and eliminate "user inaccuracy/not playing right in both tests", get one of those auto-strummer motors that someone posted on these forums before, so that the motor provides a consistent test behavior.

If we can scope out measurements of different frequency responses, like the guy micing his amp and graphically representing the outputs over the frequency ranges, and if we can or can NOT hear a difference before and after pulling back the curtain on the wizard a la the graphic charts in side by side comparisons, then we should be able to consider the matter settled once and for all.

All that has to be done is put the test parts at the quietest part possible in the signal chain (low volume guitar), and amplify it over and over with a number of units, to be sure that any differences are clearly audible to even the person with the most Tinny of ears.

This is almost a good experiment.

You also need a random-ish method of alternating sound samples in under 2 s with an A/B button (or hand raising) for the test subject to press to identify each sound sample, with a large enough sample size to eliminate chance. You could ask a person, "did these sound different" and get a "yes" if the person doesn't want to give the impression that they couldn't hear something they were supposed to.

I'm not convinced that an auto-strummer sufficiently isolates the capacitor as the only change in each sound sample, but a large enough sample size should make that a wash. But since I'm not trained in research methodologies, I couldn't tell you how large a sample size per test subject is sufficient. You'd need preliminary scope data to determine how close the the auto-strummer can get.

Another problem: What is the test subject comparing to? You can tell them, "This is Sound A and this is Sound B" but they can't remember a baseline sound for more than 2 seconds, so any choice beyond the first switch will give no comparison for them to decide on pressing A or B. You need a way to run a double blind test and obtain a sample size that can rule out subject bias or chance while dealing with the human brain's inability to remember sound for longer than 2 seconds. The best I can come up with is: Every test is always "Sound A" first. The subject must press a button when Sound B is different. Do that several times over multiple sessions (to eliminate ear fatigue) with multiple people and you should have a decent set of data that can be testing in repeat experiments. That would be a robust experiment.

[induction]

Slightly more robust, I think, would be to prepare several pedals with nothing in them but coupling caps hooked to dpdt stompswitches.  Some have two identical-value caps being switched, others have only one, and the switch does nothing.  In some of the pedals with two caps, both caps will be the same type, in others, different types.  Instead of just listening, let the test subject play through the pedals, switching at will. Then they state whether they think the caps are being switched or not. Answers are simple yes/no.

You generally want experiments to be as simple as possible, and I think playing instead of recording removes a lot of extra complications.  Youtube would be much worse because of the way they compress the audio.

The statistics wouldn't be hard once the numbers are in place. The test is designed to see if anyone tested can tell cap types apart, even if they can't identify which are which.  If you try to extrapolate the results to a population distribution, it gets a little harder, but if you just want to see if anyone can pass, sample size isn't a big concern.

The downside is you can't be both the subject and the person who prepares the boxes, which means it takes at least two people to do the test, and it can't be done remotely.

[wavley]

You'll need buttons for all of these things if you truly want an audiophile quality blind test: "deeper lows and higher highs", "it was as if a veil was lifted", "more air around the performers", "enhanced micro-dynamics", "three-dimensional soundstage", "reach-out-and-touch the performers", "preserves the musical flow with a fabulous sense of body and an extremely tuneful midrange.", "Pitch certainty was superb.", "fast bass", "The lows are incredibly lush and defined", "Hateful tizzyness in the top end", "There was an absence of a sweet midrange colouration", "Transients increased at the expense of midrange texture and air", "The deep fundamentals leave you gasping for breath", and my personal favorite "The lack of multiple crystals should prevent inter-crystal diode effects".

And this...

To be completely honest, I don't care if the guy makes up stuff about his gear or not when it's your average interview, tell people that your amp has caps filled with dragon's blood for all I care or don't talk about your gear at all, but when it's a serious nuts and bolts talking shop with a peer interview at least keep it realistic.

I want some "formed" dragon's blood caps in my amp. How much do they cost?!  

$100 each   I build them in the old Mallory toilet paper tube cap bodies so that they look 100% vintage, some of the leads are clipped a bit short and please ignore the bulging because it's hard to tame dragon's blood.  They will give your amp glassy highs, haunting mids, chewy low-mids, bell like attack, and a huge tight bottom end.

edit: I would also like to add this nugget of me being a jerk from the same thread

"You're being interviewed by a guy that can tell the difference between a Pre-CBS Fender and Vietnam era knock off guitar and you know it, but you still keep up the charade.  That would be like if I got famous and one of you guys interviewed me and I insisted that my fuzz face with two extra knobs built on cardboard had extra special mojo because the tone comes from the cardboard, and more specifically from Vintage Ikea cardboard from the Lagan countertop and while I was making it I had 2001 a space odyssey on betamax playing on a 1964 tube tv while druids danced around my living room drinking the dew of the first spring day.  When you know full well that electrons just don't work that way."

[gritz]

You'll need buttons for all of these things if you truly want an audiophile quality blind test: "deeper lows and higher highs", "it was as if a veil was lifted", "more air around the performers", "enhanced micro-dynamics", "three-dimensional soundstage", "reach-out-and-touch the performers", "preserves the musical flow with a fabulous sense of body and an extremely tuneful midrange.", "Pitch certainty was superb.", "fast bass", "The lows are incredibly lush and defined", "Hateful tizzyness in the top end", "There was an absence of a sweet midrange colouration", "Transients increased at the expense of midrange texture and air", "The deep fundamentals leave you gasping for breath", and my personal favorite "The lack of multiple crystals should prevent inter-crystal diode effects".

And this...

To be completely honest, I don't care if the guy makes up stuff about his gear or not when it's your average interview, tell people that your amp has caps filled with dragon's blood for all I care or don't talk about your gear at all, but when it's a serious nuts and bolts talking shop with a peer interview at least keep it realistic.

I want some "formed" dragon's blood caps in my amp. How much do they cost?!  

$100 each   I build them in the old Mallory toilet paper tube cap bodies so that they look 100% vintage, some of the leads are clipped a bit short and please ignore the bulging because it's hard to tame dragon's blood.  They will give your amp glassy highs, haunting mids, chewy low-mids, bell like attack, and a huge tight bottom end.

edit: I would also like to add this nugget of me being a jerk from the same thread

"You're being interviewed by a guy that can tell the difference between a Pre-CBS Fender and Vietnam era knock off guitar and you know it, but you still keep up the charade.  That would be like if I got famous and one of you guys interviewed me and I insisted that my fuzz face with two extra knobs built on cardboard had extra special mojo because the tone comes from the cardboard, and more specifically from Vintage Ikea cardboard from the Lagan countertop and while I was making it I had 2001 a space odyssey on betamax playing on a 1964 tube tv while druids danced around my living room drinking the dew of the first spring day.  When you know full well that electrons just don't work that way."

Post of the week!

I see that the "son of noise gate" thread has just been locked - a few days after the parent thread got deleted. I guess that this is just further evidence that the World Of Pedals revolves around just two questions:

1) Which capacitor will make me sound like Hendrix?

2) How do I bias a transistor? And btw which transistor will make me sound like Hendrix?

Ok, that's three questions, but you get the point. It's getting old.

Regarding listening tests - it's tough to prepare a meaningful double blind test for internet consumption - and what's the point anyway? People on the two extremes of the argument tend not to be motivated by facts, but by faith:

"I want to believe"

"I don't want to believe what you believe"

It's not a new conflict and the "reasonable majority" who look for research from reliable sources, read datasheets and select components for pragmatic reasons + personal experience and the advice of people who actually know what they're talking about tend to just walk away becausethey know that if they interject they'll get grief from the extremists on both sides. These are people that communities can't afford to lose because they aren't controlled by dogma. Seriously.

If some people spent as much time practicing guitar as they did arguing about stuff that may not make any tangible difference then perhaps their tone would improve tangibly! And perhaps we would see some original ideas rather than the same old fifty-year-old gruel reheated.

As you were.

[intripped]

premise 1: english is not my native language, so please excuse me if i express myself in a bad way
premise 2: i'm not an electronic engineer - actually i know electronics very little

why instead of trying to organize a consistent listening test which could provide reliable results, don't you just begin with some measuring around and see if there is actually any difference between two (or more) capacitors, with same capacitance and different dielectric\technology?

i'm thinking about a procedure using white noise and a spectrum analyzer, in order to obtain directly and easily comparable diagrams.
As you surely know, white noise is already used to make such measuring: you inject white noise at the input of a device and plot the frequency response at the output
doing this we could actually see the frequency responses of different capacitors, and we could also define if: there are differences - these differences are theoretically audible

i know, this is kind of a poor measure, because it doesn't consider the transients and the real dynamic situation, but anyway it could be interesting and maybe revelatory: i've never seen something like this properly done before

... and only as a subsequent step introduce a listening test, with all the subjective and unpredictable variables that this procedure inevitably provides.

[mistahead]

I believe that the argument had, somewhere up there, descended into the idea that:

'scope measurements indicating a different waveform impact / secondary component value, in a range which is audible for a given application, where two components use different tech/methods to reach the same primary component values are not indicative of what the real world outcome will be.

Science is leaking.

[Jdansti]

I was going to try to get caught up on the thread, but I don't want to make my headache worse. Without reading the last two pages, I'll just say that I agree with everyone. 

[mistahead]

Still the golden rule will always prevail - does it sound better in the build in front of you in the rig you're playing through, with the folks you're playing with, to the folks you're playing to?

If yes - well that is it.

If no... get new folks to play with and to.

[induction]

I believe that the argument had, somewhere up there, descended into the idea that:

'scope measurements indicating a different waveform impact / secondary component value, in a range which is audible for a given application, where two components use different tech/methods to reach the same primary component values are not indicative of what the real world outcome will be.

Science is leaking.

Pure logic does not tell us that differences between caps seen in an oscilloscope that fall outside of the range of operation of the part and circuit in question can cause an audible difference in a pedal.

Obvious strawman is obvious.

Oscilloscopes can tell you if two waveforms are different, but they can't tell you if anyone can hear that difference. For example, an oscilloscope can tell you that one waveform is inverted relative to another, but I doubt anyone could tell that by listening. The only way to know whether a difference is audible to someone is a hearing test. The question is itself subjective, but the test can be performed objectively.

In science, experimental data trumps theory, assuming the experiments were performed rigorously. If the question is about what you can hear, the answer is given by a hearing test. In any case, we have no physical theory that delimits what can be heard by humans. If we want to make predictions, we use guidelines that are based on hearing tests, not on physical theory.

Theory exists to describe and explain phenomena. In this case, the phenomenon is a human hearing a difference between two caps. Before using theory to explain how this happens, we need to know whether it actually happens at all. Otherwise, there's no phenomenon to explain.

I don't have a horse in this race. When I say I'm skeptical, I'm only saying that I haven't seen any evidence of the phenomenon in question, and therefore don't have a positive belief in its existence. That is the default position in science. It doesn't mean I have a positive belief in its nonexistence, though such a belief could be developed if I search for it and fail to find any evidence of it. All it takes is one counterexample to demonstrate that it exists, and I would modify my position. (That's how science works.) That counterexample will have to take the form of a hearing test.

I don't understand why this bugs you. Maybe you prefer your arguments between true believers and absolute deniers who make up their minds in advance of any evidence. That's not how science works.

Science is doing just fine, thanks.

[mistahead]

I believe that the argument had, somewhere up there, descended into the idea that:

'scope measurements indicating a different waveform impact / secondary component value, in a range which is audible for a given application, where two components use different tech/methods to reach the same primary component values are not indicative of what the real world outcome will be.

Science is leaking.

Pure logic does not tell us that differences between caps seen in an oscilloscope that fall outside of the range of operation of the part and circuit in question can cause an audible difference in a pedal.

Obvious strawman is obvious.

Oscilloscopes can tell you if two waveforms are different, but they can't tell you if anyone can hear that difference. For example, an oscilloscope can tell you that one waveform is inverted relative to another, but I doubt anyone could tell that by listening. The only way to know whether a difference is audible to someone is a hearing test. The question is itself subjective, but the test can be performed objectively.

In science, experimental data trumps theory, assuming the experiments were performed rigorously. If the question is about what you can hear, the answer is given by a hearing test. In any case, we have no physical theory that delimits what can be heard by humans. If we want to make predictions, we use guidelines that are based on hearing tests, not on physical theory.

Theory exists to describe and explain phenomena. In this case, the phenomenon is a human hearing a difference between two caps. Before using theory to explain how this happens, we need to know whether it actually happens at all. Otherwise, there's no phenomenon to explain.

I don't have a horse in this race. When I say I'm skeptical, I'm only saying that I haven't seen any evidence of the phenomenon in question, and therefore don't have a positive belief in its existence. That is the default position in science. It doesn't mean I have a positive belief in its nonexistence, though such a belief could be developed if I search for it and fail to find any evidence of it. All it takes is one counterexample to demonstrate that it exists, and I would modify my position. (That's how science works.) That counterexample will have to take the form of a hearing test.

I don't understand why this bugs you. Maybe you prefer your arguments between true believers and absolute deniers who make up their minds in advance of any evidence. That's not how science works.

Science is doing just fine, thanks.

Are you arguing with me over one of the points we agree upon?

Don't start lecturing on science son.

[FiveseveN]

But it isn't at all clear what you do agree on. Someone seems to be playing the "it takes more faith to be skeptical" game.
Allow me to rephrase the "should it be audible" question in a way that I hope to give a new perspective on your argument: Do you agree that we have test instruments (scopes, spectrum analyzers) that outperform the auditory system of any human (in bandwidth, dynamic range, distortion, noise etc.)?

[gritz]

But it isn't at all clear what you do agree on. Someone seems to be playing the "it takes more faith to be skeptical" game.
Allow me to rephrase the "should it be audible" question in a way that I hope to give a new perspective on your argument: Do you agree that we have test instruments (scopes, spectrum analyzers) that outperform the auditory system of any human (in bandwidth, dynamic range, distortion, noise etc.)?

It's possible to measure lots of stuff that is absolutely inaudible to the human ear - the most obvious one being volume differences of much less than 1 dB. Unfortunately, simply knowing that two signals are different is enough to change our perception of them. I've fooled myself like that plenty of times! That's why it's important that all tests are best done under controlled supervision, where people like me can't try to cheat - just putting samples out on the interwebz is open to all kinds of abuse. It's also possible that the person preparing a comparison demo of two different bits of hardware might modify their playing (even subconciously) according to which piece of hardware they're playing through. The power of suggestion is strong. A better test might be to pre-record guitar samples and then reamp each piece of hardware with the same samples. This does however make it easier for people with analysis hardware to spot (perhaps inaudible) differences, which is why tests should be supervised. The particular hardware under test might "prefer" being plugged into a real guitar too (as classic fuzzes and wahs tend to be).

These threads are generally a bit flawed in that people end up arguing about vanishingly small differences that are only perhaps audible to a very select few (assuming in a particular case that there is any difference at all), whilst mojo-hounds often claim night-and-day differences. They can't all be right, yet it's treated by many as a black and white argument. People who sit on the fence and say "prove it" get stick from extremists on both sides, 'cos it's basically a religious argument. "I do / do not believe. Ergo you are a heathen savage / slave to dogma." *Delete as applicable.

And without putting people in MRI scanners, or giving 'em brain implants how can anybody be certain of what another person actually does "hear" (perceive)?

As far as I'm concerned people can believe what they like as long as they don't mislead newbies. But they do.

Edit: one very good source of information on human hearing / perception is research regarding hearing aids. There's a lot of it available on the web, but it's deep stuff. Psychoacoustics is complicated, but it's not magic!

[wavley]

http://www.rane.com/pi14.html

[gritz]

http://www.rane.com/pi14.html

That is the coolest thing I have ever seen - do you have a verified schematic?

It reminds me a little of this software gem: