DI's can do a lot more than just impedance matching. In fact, there are so many different kinds of DI boxes, they can all impart their own sound upon whatever is plugged into them.

Guitars typically need to see a load of no lower than 1,000,000 ohms. Direct inputs on most inexpensive audio boards are generally not quite that high. Most of those are in the order of around 10,000 ohms. That will definitely damp the sound of an electric guitar. When the input impedance is higher, so as not to "load down" the pickups on the guitar, the sound will appear to be more open with better articulation & low-end.

Some active DI's, which require battery or microphone " Phantom Power" generally have an FET input transistor. FET transistors have extremely high input impedances and so, mimic the same input impedance of a tube guitar amplifier. If you want to take a guitar direct, you should utilize an active DI box, plugged into the microphone preamp & not the line input.

Many DI boxes available utilizing input transformer. This transformer has its own pros and cons as well. Typically, these transformers don't have an input impedance any higher than 50,000 ohms. That's far too low for guitar and will change the character of the sound of the guitar. I've actually used plenty of those myself but mostly on bass. Back in the day, we didn't have active DI boxes and so, there was no other alternative. This transformer DI boxes would be appropriate to connect to the output of your guitar amplifier directly. That would allow your guitar amplifier to properly load your guitar while utilizing the amplifier as just a big fat active DI. But because your amplifier does not have a balanced output nor do you want to take a powered output into a line much less a microphone preamp. That's a big no no. And so then the transformer direct box serves to convert your amplifier output to a balanced microphone input. Some direct boxes are actually designed to be connected to an output of a guitar amplifier or have a special switch to allow that provision.

The clipping bass you mention means, simply, you ain't setting levels correctly. Meters don't show everything. That's what you listen.

Your test is also somewhat flawed but is not flawed in that, the audible differences are obviously apparent to you. A lot of that has to do with improper gain staging in your process. And while you've indicated your examples to be direct injection or no direct injection, none of these are from a microphone. So they're all direct injection. Hearing the difference between circuitry, design concept, amplifiers operating at different gain levels. So unfortunately, your test is not quite a scientific one. As your test was not performed with the full knowledge of how things work. So your test had errors Directly Injected into the test.

Your observations about the difference in sound, the different ways you plug in your guitar already demonstrate that you understand what you are listening to. You are more than likely hearing the effects of loading down the guitar pickup, additional circuitry, uncertain gain staging. And it sure don't sound the way you want it. So those were important observations on your part. Welcome to the land of audio engineering.

And so now, your question is... What? You are merely making observations. You want to know what's gooder or better than gooder? What if I said a book? Of course those don't make much sound on their own without the help of some human intervention of some sort. Recording guitar is the same way. You don't plug your guitar into your recording device and instantly sound like Jimi Hendrix. Not even Jimi Hendrix would sound like Jimi Hendrix that way.

So what kind of microphone were using on the non-direct injection?
Mx. Remy Ann David

No microphone, or the only microphone I have now is an SM57, which don't sound good on acoustic (or I'm micing it wrong, or don't have a good room sound). Well, the purpose of the experiment was to explore the need for direct boxes between instruments on stage in a live setting. So what your saying is to use an active DI box on a guitar or bass, then a microphone preamp before plugging into a soundboard? Let's see, I need an ohm resistance of 1 million for guitar, so an active DI will do. Then a preamp which boosts the signal to line level for the engineer. Well, I just cannot get into those electrical engineering books, haha. Thanks for your expert advice.

Can someone explain impedance to me? I know the tech definition and how to hook things up right, but I've never understood "how" it "changes" the sound.

It seems to me that we should be aiming for zero impedance all around for maximum efficiency, but this is obviously not the case.

Impedance has to do with being able to fully transfer energy. It is expressed in ohms but cannot necessarily be read by an ohm meter. That's because it's not a resistive expression but an inductive indicator.

Plus, with each type of DI be it active or passive, each one most definitely has its place. For instance, a typical direct box that utilizes a transformer will have a 50,000 ohm primary that you plug your guitar into. Depending upon the input windings ratio, your output windings ratio can be higher or lower. It is generally described in terms such as 50:1, 10:1. This means that there is 50 times more wraps around the iron core then the output windings has. This allows one to not only match impedance for greatest energy transfer but also allows one to step down or step up level. Transformers can actually be used as passive amplifiers. It would be a passive amplifier if you were to reverse your input and output. So, if you had a 1:50, you'd be creating approximately output voltages 50 times higher than the input voltage. However, your input current would be greater than your output current. So you can raise the voltage and lower the current or lower the current and raise the voltage. So you can also use them as attenuating pads and create a microphone input level that is compatible with a microphone input when fed from a line output. A line output is too high in voltage for a microphone input. Current is not a huge factor in these applications as we are not talking about powering loudspeakers were current is the bottom line. That's why those are only 4-16 ohms. They load the amplifier down heavily but the amplifier is designed to provide highest current which equates to highest power and that's where you want Watts not little itty bitty voltages. So, transformer direct boxes are Bi (you can fill in the rest there)

Active DI's are only one-way devices and so are quite hetero. The most basic of direct DI boxes utilize a single FET transistor. The input to the FET transistor is approximately 2,000,000 ohms. Whereas its output is quite low and its output impedance. This actually drops the level but by virtue of its very low impedance, it matches up well to a 150 ohm (typical) microphone input while functioning from the 48 V back fed phantom power from the console/preamp to make the unit function. Those cannot be used bidirectionally like transformers can be since those are completely passive coils of wire. Unfortunately, most transformers for audio purposes cannot be practically made with a 1,000,000 ohm input impedance.

In answer to your question about musicians on stage live: You misinterpreted the use of the DI in live applications. If you use an active DI, that's all you need to plug into the soundboard. If you use a passive DI, that's all you need to plug into the soundboard. The soundboard has a microphone preamp. So there is no necessity for an outboard microphone preamp because it's already in the soundboard when you plug the DI into it.

Yes, in a sense zero impedance represents a direct short. However, out there in high power amplifier land, some of those amplifiers can be operated with a direct short. So if you pile on 8, 8 ohm speakers in parallel, you'll get 1 ohm which is virtually a short. And some amplifiers are designed for that kind of impedance abuse. Others just blow up. And we're not talking about the transfer of power when dealing with low-level devices such as guitars, keyboards, microphones, we are dealing with voltages, very low power, very low current. This is why we all had to take DC 101, quite literally. You have to be able to understand DC circuit theory before you can even understand impedance. And in the land of transistorized amplifiers, there is the flow of holes which actually goes in reverse of the voltage. Yup, the flow of holes. Now this is been explained to me time and time again and I don't understand it because I'm not an electrical engineer creating and designs circuitry from the silicon up. However that doesn't stop me from taking somebody's already good sounding op amp and designing an entire console around it. Because we're not talking power, we're talking voltages. This is all a little hard to comprehend when you are not a tech head.

Both have their place in every studio and one should have both
Mx. Remy Ann David

Impedance is the ratio between the amplitudes of a sinus AC voltage and a sinus AC current. It is also indicating the amount of phase shift between them.
Iow, wire coils like in speakers are driven by AC which causes an induktiv resistance that produces a phase shift between voltage and current which is heavily frequency depending.
That is called impedance ( not resistance ). To prevent reflections that alter the frequency response it is advisable to match the impedances as good as possible.

Coils or chokes and capacitors produce phase shifts between current and voltage.
Coils: high frequencies = high impedance & vice versa = DC is a short
Capacitors: High Fs = low impendance & vice versa = DC is blocked.
This is rather difficult stuff to learn, because it is imaginary and invisible... and there is math and piles of formulas involved...
like

Zero impedance is a short for the connected devices ( e.g. amps) that feed AC signal to the load on the other end, because the real DC resistance of e.g. cables and speakers is rather low. If there is a coil or condenser ( even parallel leads ) in an AC signal path there is always impedance, as well.

OK, so a mismatch in impedance changes the sound because different frequencies cross the mismatch in unequal amounts, thus introducing distortion into the original waveform.

Not just distortion, you may even see a different slew rate for some frequencies causing them to arrive at a different time than others. It kinda smudges up things when your fundamental is being paired with an even order harmonic that isnt anywhere close to you note.

Yeah, that's exactly what BigK is saying. Most people think of phase shift as being 0° or 180° but the fact is, everything flies through these wires going at different speeds versus frequencies. This is the essence of how old-school inductor/capacitor equalizers work as well. It's all done with phase shift and that's why many old equalizers are "inductor/capacitor" . An inductor is basically just one side/1/2 of a transformer. And it's the inconsistencies of how fast frequencies versus timing is in phase timing through the transformer that gives each different manufacturers transformer a unique sound all their own. And the core be it iron, iron/nickel, pure nickel because of its magnetic properties will also affect how fast these little frequencies are all spinning around, so to speak. And although you may have described it as " introducing distortion into the original waveform" is technically correct but not accurate. It is the phase shift that helps to give each transformer if it's unique quality. This is a distortion that is musically relevant as opposed to wanting to be avoided because you think it's distortion. Many of the finest preamps manufactured utilize high-quality input transformers & output transformers aesthetically for their unique auditory qualities. Distortion perhaps to a person that thinks that means it's distorted. Which it ain't.

I'll never be anemic
Mx. Remy Ann David

This is sooo cool. Buddha was right, there is a universe in every atom.

OK - slew rate- I know that one. It's how fast the signal can change. Peavey amps have a slower slew rate than Crown or QSC, hence the are "mushier".

Thanks all!
-Johntodd

While most operational amplifiers are higher slew rate today, they haven't always been. Some folks even think the API 2520 is fast. It ain't. It only makes a difference when you try to make crappy electronics perform like professional electronics. Then you hear the difference but that's not directly attributable strictly to slew rate because everybody knows what API sounds like. A lot have described the sound as " fast" but in all actuality it's because quality transformers were utilized. The circuit was still a good design and is to this day even if it is slow in slew rate. It makes a bigger difference with cheap consoles that don't include output buffer transistors from the IC chip op amp like in real professional equipment i.e. Neve, Auditronics, plenty others. Those are designed so well, nothing gets his chance to slew.

Slew you later, bye!
Mx. Remy Ann David

Good stuff here so far... you can make it even simpler

start with a simple idea.... Something you know

what is a filter?

Think EQ....
how does an eq effect the sound.?

Think of impedance as a filter it has resistance, inductance and capacitance is varying amounts.

In simple terms:

resistance impedes all frequencies, the higher the resistance the more it impedes the flow of electrons. Ohms Law Current = Voltage/ Resistance directly applies to perfect resistors. (Although there is no perfect resistor out there)....

Capacitance impedes more selectively. As frequencies increase the impedance of a capacitor decreases. Zc(impedance of Cap) = 1/(2*pi*frequency*Capacitance value)

Inductance also impedes selectively. As frequencies increase the impedance of a inductor increases. ZL(impedance of inductor) = 2*pi*frequency* Inductance value)

A simple filter is an RC filter. It has a resistance and capacitance. Together they pass certain frequencies to pass and attenuate others.

if you make a rc filter like this

http://en.wikipedia.org/wiki/File:RC_Series_Filter_(with_V%26I_Labels).svg

you have made a low pass filter.

It attenuates frequencies above a calculate-able threshold (often known as a 3dB frequency)

Now your starting down the rabbit hole, how much further you go is up to you. There is a whole world down here.

Like bugs & worms, snakes, electrically speaking...

I love the 60 Hz Concertoone. NOT.
Mx. Remy Ann David

I wonder what my voice would sound like if I ran it through an induction cooker?

https://secure.wikimedia.org/wikipedia/en/wiki/Induction_cooker

No, really. What would that big coil do?

You will get a mighty humm.
And you'll sound like a delicate Princess singing raw Shanties after sipping a gallon of high octane rum.
ROFL.....

LMArseO!

I think the induction cooker would work great? Here is what you do: you cut a nice sounding clean track. Once recorded, you back feed that track out to a tube guitar amplifier. You then take the coil from the induction cooker in place across the output of the amplifier to further load down the amplifier & its speaker. You crank the amplifier for the proper amount of drive distortion along with the additional load from the induction coil. Ill placed a microphone on the guitar cabinet and rerecord that to a fresh track. I'll do anything to get any kind of unique sound I can out of anything that's lying around.

I utilized a blown up power supply across the input of an 1176 to get just the kind of guitar sound I wanted. It wasn't a full way rectifier it wasn't a bridge rectifier it was a three-quarter rectum fire which provided just the right kind of fuzz.

So yeah, gimme' that induction smoker!
Mx. Remy Ann David

LOL!