Can someone explain what caps are?

I hate to turn this into an ad, but if you have questions like this you really should have my Audio Expert book.

This is a pretty deep subject, so I'll hit only the high points: Capacitors serve a number of purposes in electronic gear, from determining the frequency of an EQ, to filtering hum out of the AC mains in a power supply. The main point here is to distinguish caps that are in the signal path, versus caps that are not in the signal path. There are many types of caps - Mylar, disk ceramic, electrolytic, polystyrene - and each is better for some purposes than others. Ceramic caps are excellent for preventing ultrasonic oscillation as part of a power supply, but you'd never pass audio through one because they add distortion. Polystyrene caps have low distortion, but they're made from metal strips that are rolled up to be smaller, which adds a series inductance to their capacitance. And so forth.

Edit: I'll add that "upgrading" capacitors is rarely useful or needed, though electrolytic caps do deteriorate after 30 or 40 years and so might need replacing.

--Ethan

They're essentially a differentiator. If you put DC across it you get 0 because the difference between any 2 instances between a DC signal is 0. At higher frequencies you get a bigger and bigger voltage (a lower resistance, in other words) as we see the rate of change increasing. This means you can make a circuit sensitive to frequency and using resistors you can generate very specific time constants for filters, oscillators, envelope followers and more.

However as Ethan's clearly highlighted, they're flawed and depending on the job some are built better than others. In general distortion increases with frequency but using the right caps you're talking of a difference between 0.001% and 0.0009%. In general the distortion is below the noise floor, but as caps go bad they can start doing weird stuff.

It's no mistake that the most hi fi sounding gear use great electronics. capacitor upgrades are sometimes very useful, and the differences can certainly be heard, especially if you can mod an electrolytic capacitor out of the signal path, if not, a higher quality, quieter operating electrolytic like an elna silmic ll are nice, Electrons make mechanical vibration/noise as they do there work, the better caps(no matter their purpose) damp, and don't allow this vibro/noise through as much. I'm no expert on electronics, but i've heard the differences caps can make with my own ears, i'm a diy madman, and i've subbed all kinds of components for so called better components, some make diffs, some don't, but generally, better electronics make better sounding audio gear, especially from a distortion/noise perspective, color pieces are a whole different thing, but even then, most of those sought after pieces have quality parts in at least part of the circuit.

BTW, a blind A/B test on the net for something like this is basically useless IMO, it's about familiarity with a piece, and being able to recognize a change for the better or worse.
listening to a stock Soundcraft delta 200 input strip, compared to a Jim Williams modded strip(mostly electro cap elimination/opamp upgrade, and optimizing), is the difference between listening to an AM radio, and going to a Symphony!

There are definitely better guys to speak of this than me, J Steiger, J Williams , Svart?

Hope this helps a little
T

I think the water analogy is pretty easy to understand. Voltage is the pressure of the water and current is the flow rate of the water coming out of a pump. Both can be affected by obstructions in the pipe (resistance/impedance) or reservoirs (capacitors).

Say your water pressure/flow is erratic and pulsing. You can add a large reservoir to your system and as the erratic pressure/flow pushes water into the reservoir, the amount of bulk water in the reservoir will buffer the pulses and the pressure/flow out of the other side will be much more steady. This is essentially how bulk decoupling capacitors work.

In filter theory, all decoupling caps form little reservoirs and will steady small pulses that come down the power rails. The resistance in series with the power rails helps this by lessening the pulse's power that the cap has to deal with, but it also can affect the voltage and current by restricting the amount of power that can flow to the part that is using that power. We call that end point the "sink".

In math terms, it's the same as creating an RC filter: 1/(2pi(RC)). The theory above can also be thought of in another way, where current/voltage transients are thought of in the frequency or time domains and the decoupling is based on specifically forming a filter to "short" those frequencies to ground, thereby filtering the "noise" out. It achieves the same thing as the reservoir theory, but it's more based on math. The math way is more specific to certain issues with noise issues in designs, however "rule of thumb" types of decoupling have been determined to be mostly effective.

Now, that was the idea behind decoupling of power rails and the like. The other side is frequently called "AC coupling" or "DC blocking" or even simply, "audio path caps" or something like that.

These serve a different purpose but in a similar way. A lot of active parts will have some amount of DC voltage or currents going into or out of them. The common source for this in audio is the Op Amp. Along with amplifying the audio signal, which is primarily AC signals, the opamp will amplify it's own input currents along with noise from it's output being coupled into it's input through feedback and such. Most of the time this is not a huge concern even though folks will try to sell you snake oil fixes for these things.. But sometimes you have enough DC coming out(called DC offset) of the part that the next part will amplify that DC and you'll lose headroom because the output of that next stage is offset by the amplified amount of the DC offset from the previous stage. This might not be a problem, but loud AC signals will cause the AC signal which has been offset, to clip earlier in the direction that the whole signal has been offset. DC blocking does just that, blocks DC offsets between stages.

AC coupling/DC blocking works as it does by forming a High Pass Filter above the point figured in the equation 1/(2pi(RC)), same as any filter, but the R portion is the output and series impedance of the system and the C is the coupling cap. This frequency determines the lowest frequency that will pass through the series capacitor. DC is essentially thought of as a frequency of ZERO.

While folks will swear that different caps sound different, it's only partially true, especially for decoupling caps. Most of the time with decoupling caps, manufacturers will put the least amount of cheapest caps they can in the box. So when "upgrading" caps, what you are really doing is overcoming the lack of quality decoupling. Once you reach a certain point, any potential problems will be resolved and there is no more performance to be gained from changing decoupling caps. Beyond this, it's just confirmation bias of the person trying to justify the purchase and "hearing" the difference.

Now, DC blocking caps are another issue. this is the place where people will swear and jump up and down and get red in the face to try to badger you into submission in order to get you to buy their "upgrades". Again, a quality cap is rarely used by the manufacturer here, unless you are buying quality gear. Changing the caps out to something higher quality will create some degree of improvement, but over a certain point it's futile. As was mentioned above, the difference in distortion on "quality" caps is negligible although there are plenty who intend to sell you their services who will swear to have proof that this isn't the case, yet they never really produce the proof, rather they just produce customer testimonials from folks who are, again, suffering from confirmation bias. Buying a nice, name brand, high temperature, low ESR cap will serve most people fine, mainly because most of the gear that really needs upgrading is the low to mid level gear which usually has a performance limited design anyway and you could have bought good gear that doesn't need upgrades by the time you've done all your mods.

Hope this answers some questions.

That's a high-end microphone. Why do you think anything could be done to improve the sound? At that price I'd expect the manufacturer to have done everything necessary to ensure the highest quality!

--Ethan

?? I don't own a CV4, so i'm not speaking from an owners perspective, but I think the CV4 is a mid level microphone until modded, listen to the cv4 samples thread(on this site), that utilize Shannon's capsule mod, the difference is remarkable! I also think the reality is, leaving $2 off the production costs of a mic that sells 10,000 or so units?, puts a lot of money in someones pocket no? I believe this probably leaves potential for "mod" improvements.

I've had in my hands, a stock apex 460, and jim williams's modded apex 460(jims personal mic) he left the stock chinese capsule and transformer(not sure about the tube), neither of these mics are mine, so you can exclude "conformation bias" from the equation, the electronically modded mic made the stock sound like a toy, plain and simple..so?

Yeah - I wouldn't consider the stock CV4 to be a high end microphone...

Marlee Matlin and Helen Keller could have heard the difference in my Ghost once the caps were upgraded to Nichicon poly's.

Well, yeah...It sounds fantastic...What's wrong with making it even better? Shannon's mod had nothing to do with the electronics.

Maybe some bitchy, badly worded comments would help. Great idea! 

I'm not sure i fully understand the last few posts , but i believe that manufacture's with an interest in making a profit(all of them lol!), set a real world price point for their product, if they exceed that price point, they won't get the unit sales numbers to make the investment worthwhile, if they spend too much on production costs/materials, they will loose their profit margin, hence making the investment not worthwhile. It's a compromise, and one of the first things to be compromised in the manufacture of audio gear, is the quality of the electronics. I get what svart is saying, you can take it too far, and i have lol!, i've done mods that left me scratching my head  , and i've done mods that are shockingly good   and everywhere between...so i have personally concluded, the right upgrade on the right piece can be as obvious as cowboys quote above lol!

good luck
T

ahh.. here we go again, of course the tests are never good enough. I told you guys when we were discussing cable differences, there is only one true answer to this, trust your ears first. And again, improvements can be revealed over time, in many different ways, and after that, they can become obvious. So blind tests have their own bias. There isn't any test that duplicate the experience of listening to music at your studio, (where you know the system, and listen in many different ways) over long periods of time. I think you only need a few seconds to tell if you hear an improvement, or if you like something or not.

Read Malcolm Gladwell's "Blink".

Ethan, if you haven't been in search of the microphone holy grail, I can understand how it might seem strange to hear of a $1500 microphone that needs upgrades. What's happening is there are a few microphones that are unquestionably better than others, an original vintage C12 is one of them. These mics can cost $5,000 to $15,000 if you're lucky enough to find one. What we're all mainly trying to do is come as close to that ideal as we can, so if a $600 upgrade to a $1,500 mic gets us that much closer to the real deal, we figure it's worth it.

Perhaps a beter description might be some mics are first class, some world class, so we are often trying to get our first class mics just over the hump to world class at the best price possible.