Tone cap survey (SE Michigan)

David Collins

Senior Member
Joined
Dec 20, 2009
Messages
1,066
Reaction score
798
A recent discussion here spurred me to get back to setting up a tone cap survey I've been meaning to do for a while. Since there seems to be quite a good number of capacitor aficionados here I figured it as good a place as any to start.

Here's the most recent test box -

capbox1.jpg


capbox2.jpg


The testing procedure is quite simple. We take a guitar, disconnect the tone cap, and hook up the alligator clips of the test box in its place. Take one tone cap specimen of a particular type (like a NOS PIO) to be tested, and measure its value. Then we go through a box of extra caps of another type (cheap standard ceramics or polyesters) and measure values until we find three of that type which match the test specimen to within a given tolerance (a factor yet to be determined).

Then each cap is marked with a number, 1-4. An assistant not participating in the survey will draw cards to randomly determine which number cap will be installed in which lettered position of the switch box. Then the listening tests will begin.

These tests may consist of a single player playing the guitar with the aid of an assistant to switch the capacitors in real time as they play. Or they may include a group of listeners with a single player. In either case, the listeners'/player's task will be to try and recognize which position contains the capacitor type different from the other three. If done in a group this will have to be done silently by marking survey cards (no sharing answers ;)).

After the first test is completed the four caps will be removed and randomly drawn and reinstalled again, then the test repeated at least two, perhaps three more times.

Now here is where it gets cloudy. There are so many variations beyond this simple format the testing could take. What type of guitar and pickups? What type of amp, settings, playing style? What tolerances should be deemed acceptable to ensure value errors are not effecting the goal of comparing material types?

It's possible that I may have to answer some of these questions before going too far in surveying for material type alone. Identifying where a reasonable difference limen (just noticeable difference) may lie in value could be easily tested with this rig as well, starting with three perfectly matched caps (or as perfectly as possible) compared to one that is 2% off, then moving up to 5%, or 10%, until we find a difference which can be reliably identified as such. Likewise, this type of testing may prove useful to identify a rig/style/setting combination which makes identification of differences most easily noticeable. Of course then we get in to issues such as time such testing would take, or listener fatigue from prolonged complicated test sessions, etc.

So beyond that, I suppose I'm bringing this up mainly for two reasons. First is to invite any who may like to participate in such a survey to do so. I will bring the test box and a bin of caps in to the music store in Ann Arbor I am now working through, so that availability of the test can be as flexible as possible. I would still have to be present to set up the test, so appointments will have to be scheduled, but I should be fairly flexible here. Likewise, if there are ever any group gatherings in the area I may be interested in stopping in for a quick survey.

Second reason is of course to invite reasonable and constructive criticisms of the test method. Once parameters like cap value tolerances or guitar/amp settings can be refined, it seems to me a near bullet-proof system to compare the effects on tone of different cap construction/materials. Still, bullets will no doubt fly, so I'd like to get some ideas of what directions they may be coming from. The more concerns are aired, the more the test methods can be refined to make sure these concerns are addressed, and hopefully make the results all the more reliably and trustworthy.

And just to be clear, I'm not trying to stir any pots here, prove any points, or prefer any outcome. Just doing some honest testing, impartial to the direction of results, so long as they are as honest and objective as possible.

So let me know what you think, whether you want to come and give the survey a shot, or have any suggestions on how the methods may be improved.
 

5F6-A

Senior Member
Joined
Aug 16, 2007
Messages
10,256
Reaction score
2,662
really cool idea... :thumb: still, be aware that the construction of the box and type of wire used in it will affect the results. IMO, the only method that provides trustworthy information about tone caps is actually installing the caps IN the guitar. Other options will be adding capacitance and other subtle yet real sonic abnormalities to the tone to some degree. Just keep that in mind when you evaluate your results. :)
 

David Collins

Senior Member
Joined
Dec 20, 2009
Messages
1,066
Reaction score
798
Very good points to discuss. To address each individually -


really cool idea... :thumb: still, be aware that the construction of the box and type of wire used in it will affect the results. IMO, the only method that provides trustworthy information about tone caps is actually installing the caps IN the guitar. Other options will be adding capacitance and other subtle yet real sonic abnormalities to the tone to some degree. Just keep that in mind when you evaluate your results. :)

No doubt there will be some who may feel that the only way to truly appraise a capacitor's effect would be in the context of traditional installation. If we leave the specific factors of test interference vague and ambiguous however, it's a slippery slope between this vs something like a psychic saying they couldn't perform well because there was bad or hostile energy in the room. To avoid falling in to that trap, let's try to break that down to clear, specific points of potential interference.

One you mention is capacitance of the box, which is a very real source of potential interference. I hesitated on whether to use a foot or so of Elixer cable (lowest capacitance shielded cable I have available), or unshielded leads from the box to the guitar, balancing the potential of minor noise interference vs added capacitance which may effect the results. Though I still don't feel that the extra 20pf or so of the shielded cable would interfere, I chose unshielded leads to better appease objections from others who may feel differently.

Now I can't provide measurements of what the capacitance of a typical Les Paul circuit would be, but if you count all the hardware, shielded wire, pickup assemblies, etc, and disconnected the caps and unhooked the coils from their leads, I've no doubt you would read the capacitance of the system at least in hundreds of picofarads, if not up in to the nanofarad range. Here we are adding 35pf to that (what this test rig measures at), and keeping this constant through the testing.

With all this in mind, if one were to argue that the added capacitance could perhaps interfere with the results, then it would be very hard to argue how the change in capacitors may be nullified at some arbitrary tipping point of say, washed out at 480pf total circuit capacitance but clearly audible at 445pf. This would mean expecting a change in an LP junior, but the added capacitance of extra controls and wire in an LP Standard would likely make the change inaudible. Or that cap changes may not be audible if you have a shielded cavity due to the added capacitance, or pickup covers, or an 11 foot wire as opposed to a 10.

So I feel comfortable in writing off the very slight added capacitance as a safe constant, well short of any threshold of change with any potential to obscure differences between caps if they are to be heard in standard application.

Regarding issues like "type of wire", it is set up with 22 gauge cloth pushback wire (Fender style), ranging 1-2" from each spring clip to the switch. The lead from the switch to the clips is approximately 2' of 18 gauge unshielded wire. The total resistance as far as I can read is about 0.2Ω, or roughly the equivalent of 2" of the coil wire on your pickups. So again, given that this equivalent of adding 2" or coil wire to one of the pickup bobbins is a constant held across testing of all caps, I do not feel reason to believe it would in any way lessen or obscure any tonal changes a capacitor may impart.

Box construction - that's a tougher one. I know I've heard many credit microphonic properties of capacitors with effect on tone, so if this is your belief then removing them from the instrument is certainly a valid issue. Of course capacitors inside a control cavity are not generally in direct contact with the wood, nor have I heard many claim that their particular qualities can only be heard when wedged tight up against the wood. Furthermore, this box is far from shock-mount isolated and acoustically insulated. I have little doubt that it will vibrate a great deal in close proximity to an amplifier, and if there is any microphonic effect on the caps it should certainly come through here. While not the exact same vibrations as within in a guitar cavity, if caps are prone to influence from this factor it should be at least as apparent in this box as it would in a control cavity.

I have done many tests of caps mounted inside guitars, and have many reasons why I have decided the external box to provide for more reliable testing, but getting in to all that is another several paragraphs. I hope the specific concerns about potential interference have been reasonably satisfied.

http://www.mylespaul.com/forums/tonefreaks/166428-mojotone-capacitor-shootout.html

4 different capacitors, hard wired inside the guitar, same amp & eq settings, same guitar licks ;)

I did see those videos, and commend you for taking the time to do them. I do however feel that blind testing is absolutely key. If you had the time, and an assistant to help, I think a much better test could have been done by having your assistant install caps randomly, out of sight, and of the same measured value if it were specifically cap material you were testing.

I think it would be great if you could take those four caps, and make four more videos. This time have an assistant number the caps, shuffle and draw from deck of cards 1-4, install that cap, then bring it to you to for the video tests. Repeat four times, keeping it random (meaning the same cap might be installed several times, while others not used at all). Most importantly of all, neither you as the player nor any viewers can know which cap is being played or listened to at any time. Have your assistant seal the answers in an envelope, and do not open it until the videos have been posted and a good number of people have given their answers to identify which cap was which.

No doubt the differences between .047μf vs .022μf will be noticeable in the lower range, so anyone should have 50/50 odds of getting a right answer as to cap type from there. Of course they would have to be measured as well to make sure that any changes that were heard were not due solely to value differences. If they measure significantly off from each other, then I suppose such a test would be rather pointless.

Now I'm not actually asking you to do this (though it would be interesting), as I know it would be a lot of work. That's why I'm doing a test such as I am. Very well controlled, blind and effectively eliminating potential subjective interference for both players and listeners.
 

David Collins

Senior Member
Joined
Dec 20, 2009
Messages
1,066
Reaction score
798
Basic points I'm interested in -

If you find capacitor material type to have a significant effect on tone, and you have tried many different capacitors and heard dramatic changes affected:

A) Do you expect that you would be able to identify a difference between one type of cap and another in a "1 in 4 odd-one-out" test, if switching between the caps could be done in real time, listening in person, and the switching mechanism had no significant audible effect on the tone circuit?

If "Yes" then:

B) Do you feel that this particular test setup may have such a significant effect on the overall circuit as to outweigh and override any level of influence a cap may have. Remember, any changes are to the circuit on the whole will be consistently present for all caps tested.

If "Yes" to B, then:

C) What are the specific, reasonable, non-ambiguous factors which most concern you regarding effect on signal and interference with goals in testing?

And remember, it you feel this test method flawed and unreliable, you may also have to say that any of the tests posted on forums and videos with alligator clip cap changes, or a cable over 10' long, or a guitar that is shielded, or one that has more than one switch in it, that these people are just hearing things because the effects of cap changes would be outweighed by these factors in the same way as they would by this test box.



Sooo, who wants to take the survey, what type of guitar and rig do you feel can make the differences most apparent, and what types of caps do you think should yield the most significant difference from one another?
 

Raz59

Senior Member
Joined
Sep 8, 2009
Messages
1,009
Reaction score
28
To start -
What do you mean by significant? Because the notion of significance can be subjective - if you're serious as a heartbeat about your guitar then any difference, even if close to none, would be significant.

But to answer your questions, let's assume I am serious about guitar and I do find significant effect/dramatic change on tone by changing capacitor material:

A) Yes, I do expect to be able to identify 1 type of cap out of other 3. And it would specifically be current production ceramic caps (tiny tanned balls) versus any other;

B) I feel this test is good enough for me to be able to accuse the difference. There was another test done with alligator clips and the difference I mentioned in A was there;

C) My only concern is really the human factor. You might unconsciously change where you pluck the strings (closer to neck or closer to bridge), change the angle of your picking or the intensity of the picking and the listener might associate that sonic difference to one cap over the other.
So to do a lick always the same way would be essential, but even then you might a string with a bit of your nail and create a interesting harmonic...

If you find capacitor material type to have a significant effect on tone, and you have tried many different capacitors and heard dramatic changes affected:
A) Do you expect that you would be able to identify a difference between one type of cap and another in a "1 in 4 odd-one-out" test, if switching between the caps could be done in real time, listening in person, and the switching mechanism had no significant audible effect on the tone circuit?

If "Yes" then:
B) Do you feel that this particular test setup may have such a significant effect on the overall circuit as to outweigh and override any level of influence a cap may have. Remember, any changes are to the circuit on the whole will be consistently present for all caps tested.

If "Yes" to B, then:
C) What are the specific, reasonable, non-ambiguous factors which most concern you regarding effect on signal and interference with goals in testing?

And remember, it you feel this test method flawed and unreliable, you may also have to say that any of the tests posted on forums and videos with alligator clip cap changes, or a cable over 10' long, or a guitar that is shielded, or one that has more than one switch in it, that these people are just hearing things because the effects of cap changes would be outweighed by these factors in the same way as they would by this test box.
 

David Collins

Senior Member
Joined
Dec 20, 2009
Messages
1,066
Reaction score
798
I suppose my use(s) of the word significant is indeed in need of clarification, so thank you for bringing that up.

In relation the the possible effects of the test unit itself, by significant I mean creating a change so great as to largely or completely obscure any effects a change in tone cap material may deliver.

Example: if the test box added a capacitance of .01μf to the circuit, and/or had a resistance of 10 or 20 kΩ, then I think it could be easily argued to have such a potentially significant effect on the circuit that even if the capacitors did sound different in normal installation, their effects may not be heard in this context.

My standards for declaring the effects of the test box as insignificant in this context is not meant to imply it will not affect any change to the circuit whatsoever. Rather, its resistance and capacitance do not change the circuit to any degree outside of normal variables such as cavity shielding, harness layout, or cable length. It is simply to point out that if this factor obscured cap material effects, then one would also have to conclude that shielding a cavity, or adding two feet of cable, or having pickup covers would be even more likely to obscure their effects. Adding two feet of cable may change the sound slightly, but not many would consider this change significant enough to completely obscure effects they've heard from changing caps. Using this box in place of a standard capacitor will be less significant than two extra feet of cable.

In terms of effects of capacitor type, significant of course means enough of a change to make it clearly distinguishable from the other caps. Then of course there's statistical significance when looking at survey results, but that's where I pass it off to my wife for review.

A) That's great, and that's what I'm looking for. Only problem is, I'm looking for survey volunteers such as yourself just a bit closer to southeast Michigan. :) Regarding cheap reference capacitor type, I kind of knew most would say modern ceramics, but was kind of hoping that cheap polyesters or maybe monolithic ceramics might pass for acceptable. Modern ceramics just suck for tolerances. Trying to find 3 ceramics to match one PIO might easily involve going through a bag of 50 to get even close. Plus, they often seem terribly heat sensitive. Pick them up them up with your fingers, and you can watch the value change from your body temperature.

I'll try and find some decent tolerance ceramic discs to match a PIO, but I may fall back at least on monolithic ceramics - they seem much more reliable in terms of value in my experience.

B) I would love to find some willing survey participants in the area who feel the same way.

C) Human factor is indeed a very big issue - huge one I would say. This is exactly why I prefer to go to such great lengths to ensure comparisons are done in real time, switching while a person is playing, and that all testing is done blind.
 

David Collins

Senior Member
Joined
Dec 20, 2009
Messages
1,066
Reaction score
798
So no takers in the area interested in a survey? I could gather some local musicians for a survey I'm sure, but was hoping to find some enthusiasts from the forums in my area who may be interested.

It may take time, as this is a small side project I'm trying to squeeze in to my schedule as time allows, but I'll be sure to keep folks here posted. Perhaps a video survey would be nice, but between compressed audio and unknown speaker systems, I don't know if I could comfortably rely on video survey results as being ideal.

If anyone near Ann Arbor feels interested in taking the test though, just let me know. My box-o-caps and I will be around.
 

Sct13

Platinum Supporting Member
Joined
Sep 17, 2011
Messages
21,804
Reaction score
29,188
Interested, I am at worrk at the moment, let me think about the test mehods you propose.

I am not sure it needs to be complex, blind yes but not complex. adding and subtracting capacitance (of the system) is almost too precise.

The subjectivenes of the listener and the variance of the player is a huge variable. (as stated in this post already)

Time for "Mythbusters"
 

David Collins

Senior Member
Joined
Dec 20, 2009
Messages
1,066
Reaction score
798
Interested, I am at worrk at the moment, let me think about the test mehods you propose.

I am not sure it needs to be complex, blind yes but not complex. adding and subtracting capacitance (of the system) is almost too precise.

The subjectivenes of the listener and the variance of the player is a huge variable. (as stated in this post already)

Time for "Mythbusters"

Thanks for giving it consideration.

Just to be clear, I'm trying my best to keep the test as simple as possible. "Here's a switch, one position is different from the other three, you try and name which one". It's only the explanations or arguments of the details which can get complicated, such as whether the introduction of the test box may interfere with the circuit in a way outside of normal parameters (which I believe I have substantiated that it shouldn't).

Any concerns or suggestions as to how the testing may best be executed or improved though are more than welcome. I've gone through several iterations of this type of test and feel this one is reasonably simplified and refined, but am always open to improvements.

I'd love to do a "ToneBusters" video series.... :D
 

Mookakian

Senior Member
Joined
Jan 18, 2010
Messages
6,601
Reaction score
2,862
Im all for it, no harm done in more testing clips and it may inspire my own testing of something I have not considered. If you do get around to it i can definately say it will be appreciated by quite a few.

Dont be thrown by the queries, just means peope are listening and eager ;)
 

Latest Threads



Top