Capacitor Test: Polyester Film vs. Paper-in-Oil (PIO) in Guitar Tone Circuit

Summary

Does capacitor kind (paper-in-oil -- PIO, polyester film, ceramic) matter in guitar tone circuits? If so, how much? A short test, downloadable samples, and my opinion below.

Reading time ~ 4 minute(s)

Background

Some people swear by paper-in-oil (PIO) capacitors for electric guitar tone circuits. They say all else being equal (same capacitance, similar tolerance — voltage is not relevant in guitar circuits), a PIO capacitor will sound different, meaning better. Words like “warmer”, “mellower”, “livelier” are being employed to describe the difference. The theoretical basis for this would be the dissipation factor, which obviously varies with capacitor “architecture”.

Not surprisingly, PIO capacitors are significantly more expensive. Some of the models and brands being tossed around in these discussions are the famous Bumble Bee, Angela/Jensen, Vitamin Q, and the cheaper, but well regarded Russian K40Y9.

Since I had to replace my factory stock .047µF tone capacitors because they are way too dark for a humbucker guitar, I decided to go for paper-in-oil, see if there’s any gain. I opted for the cheaper Russian K40Y9, of which I also got the .047µF, just for the purpose of this test, before moving on to a .022µF.

Hypothesis

Are paper-in-oil capacitors different/better sounding than factory stock polyester film capacitors, all else being equal?

Test Setup

The guitar used is a Dean Soltero Standard Special, equipped with Dean DMT USA covered humbuckers, Nostalgia neck, Baker Act bridge. This is Dean’s take on Gibson’s Les Paul model.

The awful stock wiring was converted to 50’s wiring. The volume potentiometer are stock 500K. The bridge tone pot is a Dimarzio 500K, push-pull. The neck tone pot is the stock 500K. All pots are audio taper.

The recording was performed direct, via the Hi-Z input of an Edirol UA-25EX interface, input gain knob at 12 o’clock. No amp sims, no effects applied. During all recording, the volume pots were kept at 10. The difference in volume between samples is due to the 50’s wiring, and it is the same between both sets (polyester and PIO).

The stock capacitors are polyester 2A473J:

Polyester Capacitor 2A473J

The new capacitors are Russian paper-in-oil (PIO) K40Y9:

Russian Paper-in-Oil (PIO) K40Y9 Capacitor

The important thing is to remember that the setup was identical in both situations, the only variable was the capacitor.

The Test

The following video is a thorough A/B test of polyester vs. PIO for various tone knob positions, first for the neck pickup, then for the bridge. I chose to roll off the tone in increments of 2, which appeared to be enough for the purpose of the comparison. The video is about 9 minutes long.

The polyester capacitor samples were recorded with the cap soldered in place. The PIO samples were done with the cap connected via test leads.

I played a simple G-A minor-F-G sequence, and a basic A minor Pentatonic lick, flatpicking.

I suggest actually not watching the video initially, but instead just listening to it in order to avoid confirmation bias (you’re convinced PIOs should sound better, and the samples will sound better to you) by reading the captions, try to see if you get any “a-ha” moments.

Discussion

My conclusion obviously applies only to the Russian K40Y9 PIO capacitor, as tested against the polyester film 2A473J.

They offer identical results, for all practical intents and purposes.

I have no direct experience with $15+ capacitors, and I think I’ll never have, judging by the test I’ve just done. I really doubt that, since there’s no practical difference in this test, there will be a significant one using a Vitamin Q, for instance.

I can definitely fool myself into hearing just a tad more harshness for the polyester in the bridge (where the Dimarzio pot is), but I have to try really, really hard, and it’s so minute it makes no practical difference whatsoever.

Some samples may seem to sound very slightly different, but not consistently, and not between corresponding samples (polyester vs PIO). That is attributable to my playing dynamic and getting in and out of tempo occasionally.

This guy’s results were one of the factors that determined me to try PIOs. However, looking back at them, I’d say his very audible differences (with the tone pot on 10!) is due to switching from modern to 50’s wiring, and upgrading to CTS pots, not to the capacitors. Instead of keeping all else equal, and just test with different caps, he compares two completely different guitars from the point of view of electronics! That test is therefore very misleading.

Perhaps there would have been a significant difference between really cheap, high-tolerance ceramic caps and PIO, but that difference would have been the same between said ceramics and the polyester ones.

Conclusion

As you can imagine by now, my opinion is that there is no real, usable difference between the two types of capacitor in the guitar tone circuit, all else being equal. None whatsoever.

Unless you have really crappy stock capacitors in your guitar, for significant improvements in tone you should probably look elsewhere first.

Download High-Quality Audio Samples (FLAC, circa 9MB .zip archive)

(4.00 / 10 votes)
Loading...

33 thoughts on “Capacitor Test: Polyester Film vs. Paper-in-Oil (PIO) in Guitar Tone Circuit

  1. Interesting comparison – i can hear differences but that could be due to differences in playing and strumming/different cap values (did you measure then and were they exactly the same?)

    Have to say the difference is very subtle- the PIO just seems to have a tad more depth and body and less “dead” than the poly cap (particularly notcable when dialed in more- the Poly cap is dead at 0 tone and kills off any usable tone at “0” wheras the PIO does not).

    But it is very subtle. Worth the difference? Well if its a case of an extra 5 or 6 pounds for a couple of caps -then YES certainly is worth it! Less than a guitar magazine at WH Smith.

    However, if its 100 quid plus, then NO! And Gibson should feel ashamed at selling its fake bumblebees at over $100 a pair – they are just cheap poly caps inside

    Would LOVE to have compared the FLAC audio samples but they are no longer available?

    Any chance of getting them re-loaded?

  2. Thanks for your comment.

    No, I did not measure capacitance, simply went with the manufacturer specs. The K40Y9 are rated at 10% deviation from nominal capacitance. I’m sure none of them are precisely .047µF, but close enough for all practical purposes.

    Would you be able to consistently tell which capacitor is being used in an A/B test, though? However, I agree, if you must upgrade your caps, why not go with PIOs, as long as the price remains decent (i.e., insignificant).

    Don’t start me with Gibson, please :)!

    I have updated the FLAC download link, please go ahead and have a listen.

    Regards!

  3. I do believe there’s “much” thing different. Especially, if you try smaller value of each caps (polyester and PIO), I tried 0.01uF (10nF). Maybe, just maybe, stock caps installed in guitar is just for customer easily notice when the knob is turned, and the stock caps is always too high in value, that’s too much in cutting high frequency. You will be able to hear the difference if there’s is not to much cut, and that’s about how a caps behavior in cutting and passing the frequencies.

  4. @marsen: I don’t quite understand your comment, but I think you mean there’s a significant difference for lower cap values? I would be surprised if there was, given my results. Again, we’re talking a significant (not hugely so, but definitely there) difference!

  5. Yes,that’s what I mean. The difference is much more in lower cap value.
    In your flac sound sample, to my ear, the guitar sound smoother with PIO caps.
    I’ve been using PIO because when I go with overdrive or crunch, the sound still have something when they are clean, more focus, and there’s some quality that I can’t really describe. BTW, I’m using 0.01uF; 200v PIO, with no-load 500k push-pull tone pot.

  6. @marsen: “the difference is much more in lower cap value” — that could be interesting to test, intuitively speaking, although I wouldn’t expect too much. I’m not denying there may be some very subtle nuances in tone from film to PIO, I just don’t think they are as significant as to justify paying too much for it (caps + tech fee if you don’t fo it yourself).

  7. Next time do a real comparision,
    take the guitar and put 2 outputs, 1 passing for a tone with 1 cap and the other 1 with the Other cap… then you can take the record at same time for a good comparision, If you plays each phrasse never will be exactly the same sound, for a fair test, anyway, you are playing very “dirty” sometimes you fail on the string strumm.
    THen do a circuit with 2 output for a fair and clean test. Is my opinion, I did many test with caps and wirings.

  8. I’m not very sure what you mean… I don’t think you can get more fair than what I’ve done (same everything – guitar, player, cable, etc., the only variable being the cap – how is that not fair?)

    Your setup is overkill, definitely. Also, if you use two pots, you are introducing a very important new variable in the equation, which is the other pot, with its variable tolerance. Now that would make the test completely unfair and useless.

    Further more, I am investigating a real, significant, usable difference, not some sort of an infinitesimal difference (which I am sure exists) that would make no difference whatsoever in real life.

    My “dirty” playing is the same for both caps, it shouldn’t make any difference. I’m investigating coloration of the tone by the cap, not volume or my playing accuracy.

    Next time let’s see your tests. Although if you use the setup you described, with two pots, I can assure you your test is deeply flawed.

  9. There are several issues relating to proper evaluation that haven’t been mentioned:

    – use a tube amp – a solid state amp doesn’t have sufficient dynamic range to project the varience you’re looking for.

    – increase the distance between the pickups and the strings to reduce string damping. Contrary to many opinions, close pickups are NOT good for complex tonality.

    – use strings that are heavy enough to actually have tone. Nothing lighter than .010’s – .012’s are much better.

    – PIO caps require a SIGNIFICANT break-in period – as much as 100 hours. To properly compare caps, each should be fully broken-in before comparison.

    A new PIO cap sounds very similar to a chicklet only without the mud. However, as it burns-in you can blatently hear it go through changes in response and loudness. Unfortunately, the break-in changes are not uniform, so the guitar sound will vary constantly until the cap becomes stable. When the rest of your system (strings, PU adjustment, tube amp, cable) is in shape for high performance there is absolutely no question about the superior tonal quality of a foil and paper in oil cap. It is not even close to subtle. There are string harmonics that are reminiscent of an acoustic guitar! When played through a distortion pedal the distortion is warm and creamy!

    I have been trying to figure out a way to pre-condition a PIO cap to break it in before installation, but I haven’t worked that out yet…

    1. Old post but totally agree. Who ever did this test didn’t really know much about capacitors. With the PIO caps the dynamic range is much wider and you should test it with a relatively clean/crunch sound. Play tender and then play harder and you will hear the difference clearly. And indeed PIO caps require a break in perioud of at least few hours.

      1. With the PIO caps the dynamic range is much wider

        What a hazarded assertion. Please explain how does the same impedance offered by different capacitor kinds alters the dynamic range? I assume you do understand dynamic range refers to volume…

        And indeed PIO caps require a break in perioud of at least few hours.

        That is utter, unproven bollocks, indeed. Please realize that PIO capacitors were being used across a whole range of applications, some of them mission critical. If “break-in” was a fact, it would follow that weapons systems or avionics applications would be left at the mercy of the “break-in” process… As if any sane engineering team would allow that to be the case. Alas, entertaining the idea of capacitor “break-in” is a luxury afforded only to idle tone fiends who know little about physics.

  10. I wish that was true. It could be, I have no way of knowing for sure because I could not find any relevant tests. Those should have the cap as the ONLY variable, like I did here, nothing else should be different.

    String gauge should not be a factor here, at all. Either there’s a significant difference between caps with an given set of strings, or there isn’t. There are players with HUGE tone and .008s (e.g., Reverend Billy Gibbons – actually uses .007s(!), Malmsteen, Brian May). With the proper settings, string gauge only affects the fatness and, most importantly, the player “feel”, more than the body of the tone.

    What does actually happen in the capacitor during the so called “break-in”?

    If you want to “break it in” (whatever that means), why not record a guitar strum, and play it over and over through a guitar tone circuit?

Leave a Reply

Your email address will not be published. Required fields are marked *

Allowed HTML tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

Subscribe to article comments   Subscribe to all comments