I am an audiophile, but not an audiophool. I don't buy magic elixirs and oxygen-free copper, especially not in mains cords and speaker wire. The difference between capacitor sounds and amplifier sounds, tubes and transistors, digital vs. analog, have to have some basis in physics, and be demonstrable in a blind A/B comparison, or it doesn't exist. If a blind A/B difference can be detected, then we need a scientific hypothesis, and an explanation with reproducible results, or it's mere superstition.
There are some differences that are so gross, so obnoxious that you don't need a blind test to detect. We all know that. It sometimes happens when a component fails, and all of a sudden there's a hum that wasn't there before, or some kind of distortion that is just painfully obvious, and it goes away when you fix the problem.
Case in point: I have a pair of JBL 500 bookshelf speakers that I purchased for my office about 15 years ago. They're not what I would call hi-fi, but they are adequate for use with my computer sound system, playing the various bleeps and boops, and sometimes an AAC or MP3 here and there. They sounded OK for the past decade or so, but lately they have just started sounding obnoxious. I simply could not stand to listen to music through them. I would have to turn them off. "Obnoxious" was definitely the operative word here. And it wasn't just while I was thinking about it. Background music was distracting me from my other work. It wasn't subliminal.
I was able to rule out the source and most of the amplifier by switching to my AKG K-501 headphones. I still wondered if the amplifier was behaving badly, maybe oscillating, or under-biased with nasty crossover distortion (I usually listen at low volume). But no, the amplifier bench tested well within spec, and all the voltages and currents were normal. It was definitely the speakers -- or my hearing. But no, the headphones sounded OK. It was the speakers. But how? I've never experienced speakers going bad before.
I finally popped the back off the speakers, and at one glance, I had my suspicions. The crossover. It is the simplest kind of crossover that would possibly work, using the cheapest components possible: one (ferrite core) inductor and one (electrolytic) capacitor. The capacitors had gone bad in both channels. They were five microfarad bipolar electrolytics, and if their values changed, if they shorted in one direction, they could become quite non-linear. And of course, a change in value would make the crossover operate at the wrong frequency, and hence, change the frequency response and distortion characteristics of the speakers. I had to replace the caps.
So, what to replace them with? Electrolytics? Well no, but not with audiophool caps either. Electrolytics can deteriorate over time. Film caps generally do not. Five microfarad film caps are available for building crossovers. They're more expensive, and they're physically larger than electrolytics, which is why JBL didn't use them on entry-level bookshelf speakers, which outlived their warranty by some 14 years anyway.
So I paid my three dollars each, and put them in. The main problem was, they wouldn't fit on the crossover circuit board. I had to wire them on, and then use hot glue to secure them in place so they wouldn't vibrate and rattle around. But I got them installed.
The speakers sound much better now, thank you very much. Not obnoxious. I can work once again, undistracted, with music playing in the background. There is definitely a difference when a component is malfunctioning or operating out of spec. Capacitors can have a "sound", especially when they're used as a frequency-discriminating element in a filter, such as a speaker crossover, where voltages are gyrating all over the place. Bypass and coupling capacitors most likely do not have a "sound", unless they're inadequately specified, or if they're malfunctioning.
