Cap replacement suggestions for SFDR?

[Ptron]

However, the experiments I did with the polystyrene versus disc ceramics in the tone stack were entirely confined to just swapping different types of the same value caps on a separate occasion, and I swear I could tell a difference.

Not trying to be a jerk here, and I know that people who know a hell of lot more about amps than me tout the sonic differences in caps. I can't say that you didn't hear a difference due to caps, but I can say that using this method you can't know. I can think of a several factors off the top of my head, any single one of which would keep you from really knowing if the caps get any credit for tone change:

tolerence - the caps probably have a %10 or even %20 tolerence. Unless you precisely match them for actual capacitance, you have no way of knowing if any percieved change in tone is due to the cap having different actual values.

everything else that can change - can you guarentee that all the settingings on your amp are exactly the same? How about the guitar? How about your position in relation to the amp (can be a huge factor) or the amps position with the walls or surrounding furniture. How about line voltage? How about how warm the amp is?

aural memory - I don't know how good the average person's aural memory is but Like a lot of human senses, it's probably not very good with details.

Human psychology - The reason tests have to be double blind to be valid. Our perceptions are easily fooled. especially if we want or expect something to be a certain way. People get a little offended if you suggest they are vunerable to these things but we all are.

You'd have to eliminate all possible deviations before you really know. I know I can walk away from my amp for a while and without touching anything, feel differently about how it sounds when I come back.

[tubeswell]

Sure, can't deny all of that. Still there was a difference in what I heard; maybe the difference was in the tolerances? possibly. Could this even be attributable to excessive intermitent ear wax? Can't say for sure. The disc ceramics I put in were definitely brighter and better sounding to my ear than the polystyrene caps I put in the Tone Stack.

[David Vinson]

Maybe Audio Clips

Was considering making audio clips of the same lick with different guitars, to eliminate a variable & have the same exact signal to A/B changes to the amp. But part of making a lick *is* the way the amp works; so this approach may be meaningless (not to mention obsessive ;-)

I've got the parts in, and the caps are 20% tolerance. arrgh -- one control already out the window. Am going to replace the plate resistors first; and am taking pictures to document the process.

[Ptron]

Was considering making audio clips of the same lick with different guitars, to eliminate a variable & have the same exact signal to A/B changes to the amp. But part of making a lick *is* the way the amp works; so this approach may be meaningless (not to mention obsessive ;-)

I've done just that a few times when experimenting with changes.. I've even gone a step further and made recordings with identical settings/mic placement/yadda yadda (obviously this adds a whole 'nother batch of issues) so I could A/B the results without the downtime. Obsessive? Probably, but it's fun isn't it ? I don't mean to discourage anybody from experimenting, I'm just trying to point out some of the limitations as I see it...and why I think a lot of what you hear about this type of this or this brand of that has to be taken skeptically; a lot of it, maybe most of it, is complete bunk.

I've got the parts in, and the caps are 20% tolerance. arrgh -- one control already out the window. Am going to replace the plate resistors first; and am taking pictures to document the process.

If you have a scope and a signal gen you could match caps by putting them in an RC filter and finding the 3db down point.

[David Vinson]

Beat this thang to death

I hear you. I ruled out a recorded source for the reason you mention - when I said "part of making a lick *is* the way the amp works", I meant that the amp itself can shape the process of recording the test source, so it becomes circular.

So my planned fallback is to try to reproduce the exact same lick within different styles; all using the same settings between mods. I'll try this on multiple guitars, blah blah. The technique will the variable ;-)

Well, do ya think this topic has been beat up enough for now? Thanks to everyone for their input and experience! I'll come back with some results once I have them.

[Ptron]

Here's my method for this sort of test. although I've only tried it 2 or 3 times. It's a lot of work.

-Record the guitar di onto my computer. Usually a couple licks and a couple chords. Try to get the full dynamic range.
-Set up the amp where it's easy to move it and put it back in the exact spot.
-Use masking tape and sharpie to mark all control settings.
-Make sure the amp is settings are useful to the experiment, like if it's a change that affects the way the amp breaks up, make sure it's breaking up.
-I suppose I should check the line voltage and make sure it's the same when I do the post mod recording, but I've never done this.
-Set up my best condenser mic a foot or two away, it's important that it not get bumped.
-Set the recording levels then don't touch them.
-Record the pre-recording through the amp, watch out for feedback loops!
-Do the mod
-Again, don't bump anything. I actually measured the distance from the mic to the amp and made sure it was straight on and centered so if I did bump it I could at least get it close.
-Put everything the way it was before and re-record the pre-recording.
-A/B the clips, I loop small parts of the clip so I'm not comapring apples to oranges.

There are still a lot of problems with this. The recording process can add some big error margins. Plus, if you don't have nice flat monitors or a sensitive, flat mic you may not get accurate enough recordings to be useful. Also, as you pointed out, you don't get the reaction of the guitar with the amp but what are you gonna do?

Now if you can make the mod switchable you can eliminate a lot of this crap but you have to make sure your listening position(i.e. ears) is the same. It can make a big difference.

For all those reasons I won't go into again all this only gets you part way there. Never tried this but how about this:

-Put the clips randomly in a bunch of pairs some A/B, some A/A, some B/B
-Give the clips and clip pairs non descript names like 1A & 1B, 2A & 2B,etc., call them A and B even though they might be the same clip pasted over. The listener can't know what they're really listening to. Make sure you write down what they really are of course!
-Find a friend with good ears. Have them A/B the clips while you write down their responses.
-Don't tell them beforehand that some of the clips are the same (as long as you're not worried about them getting mad at you for tricking them )
-Compare to the real list and viola!

OR have a friend set up the clips for you so you don't know what they are. Then you can be the listener.

You could even get a third person involved and make the tests double blind. In a double blind test, neither the person administering or documenting the test, nor the subject of the test, know which is which.

If you have a switchable mod, you could put a dummy switch on the amp to allow for A/A-ing or B/B-ing. Then try not to let whoever's listening see which one you flipped.