What do you believe regarding the sound of passive components?

[Reeltarded]

Frequencies above audible range do affect the outcome. Imagine if you had an eq setup for an arbitrary range above audible say.. 40-100kHz.

Phase shift. It's a thing. Dmn it is hard to talk about this without woo-woo words. The.. magical virtual space a machine makes (that I mentioned earlier) is as much to do with things outside the range of the machines operation.

When real record producers still existed.. you get more anecdotes than the internet can contain when your task is to audition every single thing, ever.. forever. Neve 10xx series has 20 different virtual spacial flavors. Much of this is because of ringing +/- above the audible range. I love machines.

[cdemike]

or 1980.. Picking caps by Walter Jung and Richard Marsh
people knew more 100 years ago.. today again we need lo learn basics..and again and again.......

https://milbert.com/Files/articles/Pick ... tors_1.pdf
https://milbert.com/Files/articles/Pick ... tors_2.pdf

This is exactly the kind of article I was thinking of in my post before. I think particularly useful to this discussion are Figure 9 (part 1) and Figure 14 (part 2).

Another good reference to add to the pile. Take note of the magnitudes of the differences shown in the charts, and the frequency scales, though. If a 1 or 2 ohm resistor were added in series with a cap, would that be noticed at audio frequencies?

I might be interpreting your post incorrectly, so feel free to correct me if I'm off-base, but Figure 9 appears to show much larger differences in the audio frequency range. Because the x-axis has a lower bound of 10kHz, we'd have to extrapolate the curve's behavior to the frequency range we'd be concerned with for guitar amplifiers. We'd have to make assumptions about the shapes of the curves, but both displayed commonly-used dielectrics (polypropylene and polyester) are displayed linearly with the given scale (i.e., have what appears to be nearly-perfect logarithmic behavior in the audio range; both x and y axes are logarithmic). There are minor differences in the displayed y-intercepts, which, as Martin pointed out, are very low values and unlikely to objectively produce audible differences. Because we can assume a reasonable degree of predictable logarithmic behavior extending the chart to, say 1kHz, polypropylene would have an equivalent impedance of approximately 12 ohms since the slope appears to be ~-1 if using the displayed logarithmic scales. Polyester seems appears to have a similar slope but for argument's sake I'm assuming a slope also of -1 (we'd have to bust out the graph paper for better numbers...). Because polyester has a very different y-intercept, however, it appears that its impedance at 1kHz would be approximately an order of magnitude different: ~1.2k ohms. Taken further (and this is truly getting into speculative territory since it requires an assumption of reasonably-ideal logarithmic behavior even further past the graph's displayed end point), at 100hZ the differences would still have an order of magnitude difference of 1.2k for polypropylene and 12k for polyester. While small, it certainly seems that this is getting into the range of audible, albeit subtle, differences. If we use the "shadow components" analogy I mentioned in my first post, I think it'd probably be audible to at least some players and listeners if a 12k grid stopper were placed in and out of a circuit.

What stuck out to me in Figure 14 was the differences in the y-axis scales: almost all of the graphs shown have miniscule differences in dissipation in the audio range, and the only major departure was polyester, which had significantly greater dissipation. However, those differences only emerged in the very upper ranges of the audio spectrum and are relatively small compared to, say, a capacitor's real-world capacitance vis-a-vis manufacturing tolerances. That might impact the response in the presence range, though, which again, if we're running with the "shadow components" analogy would be pretty subtle.

As Martin pointed out, most people here seem to agree that dielectrics impact the sound, so this may be beating a dead horse. I still think it's important to know there's good science to explain that, though, especially since I've read elsewhere while researching this topic people repeating the phrase "capacitance is capacitance," which is erroneous per these articles. I suspected frequency-dependent series impedance would play a role, so I'm somewhat relieved that I wasn't buying totally into some silly magical audio engineering at least regarding the capacitors. What the article doesn't address is whether there are audible differences between capacitors which share the same type of dielectric (i.e. do SoZos actually sound different than Mallory 150s?), and I suspect that is probably getting too in-the-weeds for most publications to have much interest. I imagine the differences may come down to variation in production techniques like differing composition of leads, properties of the specific formulations of dielectric used, degrees of material impurities, etc. That might be enough to result in an audible difference, especially if we're taking into consideration the broader context of a circuit. To me, one of the most interesting parts of the article were Figures 1b and 3: arrangements of capacitors in a circuit can introduce noticeable differences in distortion, though I am unsure how much of those observed differences in electrolytics could be cross-applied to film capacitors. It'd be interesting to know, for example, if the reduction in distortion observed with series capacitors would result in a difference audible between something like an Orange FAC control and individual coupling caps of equivalent value. It does make sense, though, that the same frequency-dependent series impedance would have different effects depending on how the circuit is designed, particularly since the relationship between the values of the components and their resultant values when placed in series or parallel is opposite between capacitors and resistors.

I could be totally off-base with the resistors, and I haven't done much experimenting there. After reading about the scale of differences between capacitors in the articles, I suspect the inductance differences will be small when comparing between new-production resistors of the same materials/construction method. I'm not familiar with the manufacture of resistors, so it may be possible that differences in industrial conditions, manufacturing techniques, market demand for certain characteristics, etc. would impact inductance readings between older resistors of the same type compared between manufacturers and compared against new production. Those same factors could impact newer resistors, too, though, so maybe there would be a distinction between cheap no-name CF resistors vs. a recognized high-quality CF like Vishays, for example. Another variable could be an imperfect frequency-dependent relationship between DCR and impedance. Again, though, I'm hesitant to draw a definitive conclusion until I see numbers showing a reason why there'd be an audible difference (setting aside the afore-mentioned proven, evidence-based inquiries that showed, for example, that CC resistors produce second-harmonic distortion at high voltages).

[martin manning]

I was cautioning against anyone pointing to a difference, any difference, and claiming that that is what they must be hearing.

Here's another reference on film caps from Vishay. The curves to look at are KT and KP, where per the first note:

Dielectrics according to IEC 60062:
KT = Polyethylene terephthalate (PET)
KP = Polypropylene (PP)

KT is polyester, M150, 6PS e.g., and those two generally are the extremes in the charts.

[martin manning]

Frequencies above audible range do affect the outcome...

I recall hearing R. Neave say (in an interview) that he thought mixer amps needed 100k BW, even though that wasn't generally accepted.

[Reeltarded]

Frequencies above audible range do affect the outcome...

I recall hearing R. Neave say (in an interview) that he thought mixer amps needed 100k BW, even though that wasn't generally accepted.

This is where the virtual space of the machine is. In a couple very nicely equipped rooms I have sat with someone on breaks silently enjoying the perception of a non-extant discrete space. Just turning a couple things on changes a depth.. sort of. It can even decide which of twelve identical mic amps is for this female vocalist using this particular mic that passes the *key test. Each mic amp has a distinct voice and idle space.

It's honestly spooky.

*Key Test:

We are checking for kinks in the extreme highs and into the ultra range. (phase shift at the point of capture is bad juju)

You take a janitor's key ring and hold it up high, repeatedly drop them and catch right in front of the mic grasping quickly to create the most energy. A signal chain with notching reveals itself immediately because there isn't a clean series of transients for each key. It smerges together and all definition is hopeless. (even if you would never notice) It is a mechanical model of electronic oscillation. It is mud, only 9 octaves up. heh

Like a Mesa Boogie Mk IIB.

If you think we are crazy, you should stay away from microphone doods.

[Reeltarded]

Oh man! I just realized where we need to look.

Doug Fearn. If anyone has more than simple insight into everything and everything else, it is him. His mic amps are mostly RCA tube manual designs at the core.

He picks parts for a real reason. There aren't many parts.